Wingless-type Mammary Tumor Virus Integration Site Family, Member 5A (Wnt5a) Regulates Human Immunodeficiency Virus Type 1 (HIV-1) Envelope Glycoprotein 120 (gp120)-induced Expression of Pro-Inflammatory Cytokines via the Ca²⁺/Calmodulin-dependent Protein Kinase II (CaMKII) and c-Jun N-terminal Kinase (JNK) Signaling Pathways

Background: HIV-1 infection causes chronic neuroinflammation in the central nervous system (CNS).

Results: The spinal cytokine up-regulation induced by HIV-1 gp120 protein depends on Wnt5a/CaMKII and/or Wnt5a/JNK pathways.

Conclusion: gp120 stimulates cytokine expression in the spinal cord dorsal horn by activating Wnt5a signaling.

Significance: The finding reveals Wnt signaling-mediated novel mechanisms by which HIV-1 may cause neuroinflammation.

Abstract

Chronic expression of pro-inflammatory cytokines critically contributes to the pathogenesis of HIV-associated neurological disorders (HANDs), but the host mechanism that regulates the HIV-induced cytokine expression in the CNS remains elusive. Here, we present evidence for a crucial role of Wnt5a signaling in the expression of pro-inflammatory cytokines in the spinal cord induced by a major HIV-envelope protein, gp120. Wnt5a is mainly expressed in spinal neurons, and rapidly up-regulated by intrathecal injection (i.t.) of gp120. We show that inhibition of Wnt5a by specific antagonists blocks gp120-induced up-regulation of IL-1β, IL-6, and TNF-α in the spinal cord. Conversely, injection (i.t.) of purified recombinant Wnt5a stimulates the expression of these cytokines. To elucidate the role of the Wnt5a-regulated signaling pathways in gp120-induced cytokine expression, we have focused on CaMKII and JNKs, the well characterized down-stream targets of Wnt5a signaling. We find that Wnt5a is required for gp120 to activate CaMKII and JNK signaling. Furthermore, we demonstrate that the Wnt5a/CaMKII pathway is critical for the gp120-induced expression of IL-1β, whereas the Wnt5a/JNK pathway is for TNF-α expression. Meanwhile, the expression of IL-6 is co-regulated by both pathways. These results collectively suggest that Wnt5a signaling cascades play a crucial role in the regulation of gp120-induced expression of pro-inflammatory cytokines in the CNS.

Introduction

After infecting the central nervous system (CNS),³ the human immunodeficiency virus (HIV-1) often causes chronic neuroinflammation (1). The expression of pro-inflammatory cytokines, a molecular hallmark of chronic neuroinflammation, critically contributes to the pathogenesis of various HIV-associated neurological disorders (HANDs) (2–4). However, the mechanism that underlies the HIV-related cytokine expression in the CNS is still poorly understood. Gp120 is a major HIV-1 neurotoxin that is implicated in various HIV-associated neurological complications by causing neuronal injury (4–6). Among proposed mechanisms, one possible way for gp120 to cause neurotoxicity is to induce neuroinflammation (7). Previous studies have shown that gp120 in the CNS up-regulates pro-inflammatory mediator or regulators, including cytokines (8–13), arachidonic acid (14), and matrix metalloproteinases (15). However, the mechanism by which gp120 causes neuroinflammation is less understood. Prior studies revealed that gp120 can activate spinal cord microglia (16–18) and astrocytes (8, 10, 19). In the spinal cord, gp120 up-regulates pro-inflammatory cytokines such as tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6) in the activated glia (8–10, 18). Gp120-induced pro-inflammatory cytokines in the spinal cord critically contributes to the development of chronic pain (10). However, the molecular processes that couple gp120 induction to the spinal cytokine up-regulation is unclear.

Wnt5a is the prototypic ligand for the non-canonical Wnt signaling pathways, including the planar cell polarity (PCP) pathway and the Wnt-Ca²⁺ pathway (20–22). The Wnt-PCP signaling is mediated by small GTPase (Rho and Rac) and JNK activation. The activation of Wnt-Ca²⁺ pathway causes intracellular release of Ca²⁺, which in turn activates protein kinase C (PKC) and calcium/calmodulin-dependent protein kinase II (CaMKII). Spatio-temporally regulated Wnt5a signaling is critical for cell motility and adhesion, while aberrant in this signaling has been linked to tumorigenesis and cancer invasiveness (20, 23–26). In hippocampal neurons, Wnt5a is known to regulate the differentiation and plasticity of glutamatergic and GABAergic synapses (27, 28) and NMDA receptor-mediated synaptic transmission (29). Recent works indicated that Wnt5a regulates inflammation (30, 31). Wnt5a is up-regulated in macrophages and dendritic cells in response to microbial stimulation (32). Wnt5a may regulate the initiation of systemic inflammation (33), in an autocrine or paracrine fashion to stimulate the production of inflammatory cytokines (31, 34, 35). Wnt5a overexpression was observed in inflammatory diseases such as rheumatoid arthritis (36), psoriasis (37), chronic hepatitis and cirrhosis (38), atherosclerosis (39), sepsis (33), ulcerative colitis and Crohn's disease (40), insulin resistance and associated metabolic diseases (34). These findings suggest a role in regulation of systemic inflammation. However, the potential involvement of Wnt5a in CNS neuroinflammation has not been investigated.

In this work, we report for the first time an important role of Wnt5a signaling in the regulation of gp120-evoked neuroinflammation in the spinal cord. Our results show that the Wnt5a expression in the spinal cord is up-regulated by gp120. Importantly, Wnt5a is critical for gp120 to induce IL-1β, IL-6, and TNF-α, and exogenous Wnt5a evokes the expression of these cytokines. Furthermore, we also elucidated that gp120 activated the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways and that Wnt5a/CaMKII and Wnt5a/JNK pathways differentially regulate the expression of IL-1β, IL-6, and TNF-α. Taken together, our studies have identified novel signaling cascades through which HIV gp120 elicits neuroinflammation in the CNS.

EXPERIMENTAL PROCEDURES

Animals

All experimental procedures were approved by the Institutional Animal Care and Use Committee at the University of Texas Medical Branch (Protocol 0904031) and are in accord with National Institutes of Health guidelines. Young adult C57BL/6 mice (2–3 months old; 20–25 g body weight) were purchased from Harlan Labs. All mice were acclimated for at least 7 days before any experimental procedures. Mice were housed in groups of four to five in plastic cages with standard bedding and free access to food and water under 23 ± 3 °C temperature control and 12/12 h light-dark cycle.

Drugs

Recombinant HIV-1bal envelope glycoprotein gp120 (1 mg/ml; product 4961; lot 38 110003; National Institutes of Health, Germantown, MD) were aliquoted on ice (3 μl/tube) and stored at −80 °C. Immediately before intrathecal injection (i.t.), aliquots were thawed and diluted to final concentrations of 0.02, 2, or 20 ng/μl in vehicle (filtered (0.2 μm pore filters) 0.1% bovine serum albumin (BSA; Sigma-Aldrich) in sterile phosphate-buffered saline (PBS; pH 7.4; Invitrogen, Grand Island, NY), stored at −80 °C, and thawed before use). Aliquots were kept on ice and discarded after 1 h. Box5 (0.2 mg/tube; purity 98%; Storkbio Ltd, Tallinn, Estonia) and mouse sFRP3 (cat 592-FR/CF; R&D Systems, Minneapolis, MN) were stored at −20 °C and dissolved in PBS immediately before administration. Recombinant Wnt-5a (Wnt5a; cat. 645-WN/CF; R&D Systems) was dissolved in sterile PBS (100 ng/μl), aliquoted (4 μl/tube), and stored at −20 °C. Immediately before administration, wnt5a was thawed and diluted to a final concentration of 4 ng/μl with sterile PBS. KN-93 (KN; cat K1385; lot 035K11791; Sigma-Aldrich) was dissolved in sterile PBS (5 μg/μl) and stored at 4 °C. SP600125 (SP; cat. 420119; lot D00099883; Calbiochem, EMD Biosciences, La Jolla, CA) was dissolved to a final concentration of 2 μg/μl in sterile filtered dimethyl sulfoxide (DMSO; cat. D2650; lot RNBB0992; Sigma-Aldrich) and stored at −20 °C. dl-Stearoylcarnitine chloride (ST; cat. S2381; Sigma-Aldrich) was freshly dissolved in sterile DMSO (4 μg/μl) before use.

Mouse Neuron Culture and Transfection

Primary cultures of cortical neurons were prepared as previously described (41). Briefly, the neurons was cultured in 12-well plates coated with poly-d-lysine containing Neurobasal medium supplemented with 10% B27 and 0.5 mm l-glutamine for 10–14 days before use. For siRNA transfection, Wnt5a siRNA (target sequence GCAGAUGUAGCCUGUAAGU) (42) and control non-matching oligonucleotides were synthesized and transfected using GenMute siRNA Transfection Reagent (SignaGen Laboratories, Ijamsvile, MD) according to the manufacturer's instructions. Two days post-transfection, the cells were collected for analysis.

Proliferation and Differentiation of Human Neural Stem Cells (hNSCs)

All materials were purchased from Sigma except otherwise specified. Briefly, hNSCs were cultured as neurospheres in a basal medium DMEM:F12 (3:1, Invitrogen), supplemented with N2 (5 μg/ml bovine insulin, 100 μg/ml human transferrin, 100 μm putrescine, 20 nm progesterone, 30 nm sodium selenite) (43), 15 mm HEPES, 1.5% glucose, 2 mm l-glutamine, and 1× penicillin/streptomycin, 20 ng/ml epidermal growth factor (EGF, R&D System, Minneapolis, MN), 20 ng/ml bFGF (R&D System), 2.5 μg/ml heparin, and 10 ng/ml leukemia inhibitor factor (LIF, Chemicon, Temecula, CA). Expanded neurospheres were dissociated into single cells once every 10 days with 0.025% trypsin.

For in vitro differentiation, 3-day spheres of hNSCs were seeded at a density of 1 × 10⁵/cm² onto 12-well plates, which were precoated with 0.01% poly-d-lysine and 1 μg/cm² mouse laminin (Invitrogen). Cells were incubated with ELL media containing 20 ng/ml EGF, 10 ng/ml LIF, and 1 μg/ml laminin at 37 °C with 8.5% CO₂. After 4 days, the priming medium was removed, and equal volumes of differentiation medium were added to each well. Differentiation medium consisted of the basic medium described above supplemented with B27 (20 μl/ml) (GIBCO). Cells in differentiation media were incubated at 37 °C with 5% CO₂ for 10 days. Consequently, about 50% cells were type III β-tubulin-positive neurons (44).

Direct Transcutaneous Intrathecal Injection (i.t.)

For direct transcutaneous intrathecal injection of drugs, a modified version (45, 46) of the original method (47) was used. Mice were anesthetized with isoflurane (2% during the induction phase and 1.5% during the maintenance phase) with a flow of O₂, placed in the prone position, and hair on the caudal back was clipped. The experimenter's left thumb and middle finger securely held the caudal paralumbar region, just cranial to both iliac crests, and the left index finger palpated the tip of sixth lumbar (L6) spinous process, the highest point of the vertebral column, to guide a 1.5-inch long 30-gauge hypodermic needle connected to a 10-μl Hamilton syringe. The needle was inserted to the sixth lumbar spinous process with the beveled side facing downward at a 45° angle with respect to the vertebral column in a coronal direction. A sudden lateral movement of mice tail signified penetration of the needle tip into the intervertebral space between the fifth and sixth lumbar vertebrae. Solution (5 μl) was injected slowly (1 μl/s). Then the needle was held in this position for 10 s before slow removal to avoid any spillage of solution.

SDH Tissue Collection

Mice were anesthetized with isoflurane and perfused through the aorta with cold PBS for ∼1 min. The L4–6 lumbosacral spinal cord segments were removed quickly and the dorsal halves were dissected on an ice-chilled plate. The collected dorsal spinal tissues were stored at −80 °C until Western blotting analysis.

Western Blotting Analysis

The SDH tissues were homogenized in 300 μl of RIPA lysis buffer (50 mm Tris-HCl, pH 7.2, 150 mm NaCl, 0.1% SDS, 1% Triton X-100, 0.5% sodium deoxycholate, 1 mm EDTA, 1% protease inhibitor mixture (cat. P8340; Sigma-Aldrich), 1% phosphatase inhibitor 2/3 (cat. P5726/P0044; Sigma-Aldrich)). The lysates were shook on ice for 30 min followed by centrifugation at 4 °C for 5 min (12,000 × g). The supernatants were collected, and protein concentration was determined using BCA protein assay kit (product 23227; Pierce). Protein was heated in SDS-PAGE sample buffer (95 °C; 10 min), and loaded on 10% or 12% SDS-polyacrylamide gel (20 μg of total proteins per well). After separation, protein was transferred electrophoretically to nitrocellulose membranes (0.45 μm). The primary antibodies used included: rabbit anti-Wnt5a (0.5 μg/ml; cat. 72583; abcam), rabbit anti-Wnt5a (1 μg/ml; cat. 2530; CST), goat anti-IL-1β (0.2 μg/ml; cat. AF-401-NA; R&D systems), rat-anti-IL-6 (0.5 μg/ml; cat. 554400; BD Biosciences), goat-anti-TNF-α (0.1 μg/ml; cat. AF-410-NA; R&D systems), rabbit anti-p-αCaMKII (Thr-286) (1:1000; cat. 3361; Cell Signaling Technology (CST)), rabbit anti-CaMKII (1:1000; cat. 3362; CST), rabbit anti-p-JNK (Thr-183/Tyr-185) (1:1000; cat. 9251; CST), rabbit anti-JNK (1:1000; cat. 9252; CST), rabbit anti-p-PKC (pan) (Thr-514) (1:1000; cat. 9379; CST), and rabbit anti-β-actin (1:1000; cat. 4967; CST) antibodies. Horseradish peroxidase-conjugated antibodies (1:10000; Bio-Rad) were used as secondary antibody. Immunoblotting signals were detected using an enhanced chemiluminescence detection system (product 34080; Thermo Scientific). The intensity of non-saturated bands on Western blots was quantified by densitometry analysis (NIH ImageJ). β-Actin was included as loading controls.

Fluorescent Immunostaining

Euthanized mice were transcardially perfused with ice-cold PBS, followed by 4 °C fixative solution (4% paraformaldehyde (PFA) in PBS). Lumbar spinal cord segments were dissected, post-fixed in 4% PFA for 3 h at 4 °C, cryoprotected in 30% sucrose/PBS for 24 h at 4 °C, and embedded in optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA). Frozen sections (10 μm) were prepared for indirect fluorescent immunostaining. Sections were rinsed with PBS (5 min × 1) to remove OCT compound, blocked for 1 h at room temperature in 5% BSA/0.3% Triton X-100/PBS, incubated overnight with primary antibodies in 2.5% BSA/0.3% Triton X-100/PBS at 4 °C, and then incubated with fluorescent secondary antibodies at room temperature for 1 h. DAPI staining was performed to visualize cell nuclei before mounting. Images were captured using a confocal microscope (Nikon). The primary antibodies used included: rabbit anti-Wnt5a (2.5 μg/ml; cat. 72583; abcam), mouse anti-NeuN (5 μg/ml; cat. MAB377; Millipore), mouse anti-GFAP (1:1000; cat. MAB360; Millipore), and rat anti-mouse CD11b (5 μg/ml; cat. MCA74GA; AbD Serotec) antibodies. Cy3-conjugated donkey anti-rabbit, FITC-conjugated donkey-anti-mouse, and FITC-conjugated donkey-anti-rat secondary antibodies were from Jackson ImmunoResearch and used at a concentration of 3.75 μg/ml. NIH ImageJ was used for quantitative measurement of fluorescent signals.

Statistical Analysis

Quantitative data were expressed in means ± S.E. Statistical analysis was performed with Prism software (GraphPad). We used the Student's two-tailed t test for statistical comparison between two groups, and one-way ANOVA analyses with a Bonferroni post-hoc test for comparison among more than two groups. p < 0.05 was considered significant.

RESULTS

Wnt5a Expression in the Spinal Cord

Our recent work uncovered a critical role of Wnt5a in regulation of cytokine expression in cortical neuron/glia-mixed cultures (41). In this study, we aimed to test the potential role of Wnt5a in regulating gp120-induced cytokine expression in the CNS. To this end, we first determined the expression of Wnt5a protein in the mouse spinal cord by fluorescent immunostaining. As shown in Fig. 1A, Wnt5a protein was detected in the spinal cord, especially the gray matter. Wnt5a was abundant in the spinal dorsal horn (SDH) and ventral horn (SVH) (Fig. 1A). To identify the Wnt5a-expressing cells, we performed double-staining experiments. NeuN, GFAP, and CD11b were used as the cell markers for neurons, astrocytes and microglia, respectively. We found that Wnt5a fluorescence was mainly observed in NeuN-positive cells (i.e. neurons) (Fig. 1B). 93% of Wnt5a-positive cells were neurons and 7% of Wnt5a-expressing cells were non-neuron cells. There was no significant overlap between Wnt5a and GFAP (Fig. 1C) or CD11b (Fig. 1D). These data suggest that Wnt5a is predominantly expressed in neurons. The specificity of the anti-Wnt5a antibody was tested using two approaches. In the first approach, we compare it with the Wnt5a antibody that Ho et al. recently showed to recognize Wnt5a using KO mice (48). The result confirmed that the antibody used in this study recognizes that same Wnt5a band detected by Ho's antibody (Fig. 1E). In the second approach, we showed that the antibody detects the Wnt5a band that could be knocked down by Wnt5a siRNA (Fig. 1F).

FIGURE 1.

Wnt5a protein distribution in the spinal cord. A, fluorescent immunostaining of the spinal cord with anti-Wnt5a antibody. Top panel: overall distribution of Wnt5a in the spinal cord (left) and the spinal dorsal horn (right). Lower panel: left, a scheme of a half of the spinal cord. The red box indicates the area of the higher-power images shown on the right and in B–D. The white and green arrows, respectively, point to cells with or without Wnt5a expression. DAPI staining was performed to visualize nuclei. SDH: spinal cord dorsal horn; SVH: spinal cord ventral horn. B–D, double-staining of Wnt5a with NeuN (neuronal marker) (B), GFAP (astrocytes) (C), or CD11b (microglia) (D) in the SDH. The white and pink arrows, respectively, point to cells with and without NeuN (B). Note the prominent overlap of Wnt5a with NeuN (B) but not GFAP (C) and CD11b (D). E, immunoblotting comparison of the Wnt5a antibody used in this study (Ab1) and a Wnt5a antibody verified previously with knock-out mutants (Fig. S5 in Ref. 48) (Ab2); purpose of this analysis was to confirm that Ab1 recognized Wnt5a as Ab2. Each blot had three lane loaded with total proteins from spinal cord tissues of three mice. F, Wnt5a knockdown in cultured mouse cortical primary neurons revealed by the Wnt5a antibody used in this study. Scale bars: 40 μm.

HIV-1 gp120 (i.t.) Up-regulated Wnt5a Expression in the Mouse SDH

Previous studies revealed that intrathecal injection of HIV gp120 induced the expression of pro-inflammatory cytokines in the dorsal spinal cord (8–11). However, the molecular process through which gp120 elicited the cytokine expression is still unclear. To determine the potential involvement of Wnt5a signaling, we first characterized the effect of i.t. gp120 on the expression of Wnt5a in the SDH of mice by Western blotting analysis. Von Frey tests were performed to monitor the expression of gp120-induced pain. As shown in Fig. 2A, intrathecal administration of gp120 (160 ng)-induced clear mechanical allodynia, indicating the success of i.t. administration. We observed that exposure to i.t. gp120 (30 min) up-regulated the expression of Wnt5a protein in the SDH in a dose-dependent manner (Fig. 2B). On the other hand, the levels of Wnt5a protein were not increased either in the i.t. vehicle or i.t. heat-inactivated gp120 protein (Fig. 2C). Next, we characterized the temporal profile of Wnt5a expression. Wnt5a protein increased to a peak level at 20 min post i.t. gp120 (100 ng), and maintained at significantly high levels by 60 min postinjection (Fig. 2D). Similarly, in the human neuron cultures derived from the differentiation of human neural stem cells (hNSCs), gp120 (10 ng/ml) started to up-regulate Wnt5a at 10 min after being added to the culture medium, and the increase peaked at 15 min (Fig. 2E). These results suggest that gp120 protein induces Wnt5a up-regulation in both rodent and human neurons. However, we do not know if gp120 induces this up-regulation through the same or different mechanisms in these two systems. Furthermore, immunofluorescent staining experiments confirmed gp120-induced Wnt5a increase in the SDH of mice. As shown in Fig. 2F, in comparison with control, i.t. gp120 (100 ng, 30 min) exposure increased Wnt5a staining in the SDH (2.2-fold; p < 0.01). These results collectively show that gp120 rapidly up-regulates Wnt5a in the SDH. A recent study revealed an NMDA receptor-mediated rapid up-regulation of Wnt5a (49); it would be interesting to know if gp120 up-regulates Wnt5a by stimulating NMDARs (50–52).

FIGURE 2.

I.t. gp120 causes Wnt5a up-regulation in the mouse lumbar SDH. All summary graphs of immunoblotting analysis were based on at least three independent experiments (n). A, mechanical allodynia measured by von Frey tests. Compared with vehicle controls (open circles), i.t. gp120 (160 ng; filled circles) increased the response rate to the mechanical stimulation of the hindpaws (n = 7; two way ANOVA). B, dose-dependent effect of i.t. gp120 on the protein level of Wnt5a in the SDH at 30 min after gp120 injection (n = 5; one way ANOVA). C, Wnt5a protein levels at 30 min after i.t. vehicle, heat-inactivated gp120 (100 ng) or gp120 (100 ng) (n = 5; one way ANOVA). D, time course of Wnt5a protein levels after i.t. gp120 (100 ng) (n = 3; one way ANOVA). E, temporal profile of Wnt5a protein levels after gp120 (10 ng/ml) treatment in the human neurons cultured in vitro (n = 3; one way ANOVA). F, fluorescent immunostaining of Wnt5a in the SDH from mice treated with gp120 (100 ng/i.t.) for 30 min. In the quantitative summary graph (12 sections from 3 mice/group), Wnt5a levels were normalized against β-actin and expressed as relative units to vehicle controls. Error bars indicate S.E. *, p < 0.05; **, p < 0.01; ***, p < 0.001 versus controls (0 min or vehicle).

Wnt5a Antagonist Suppressed i.t. gp120-induced Cytokine Expression

The up-regulation of Wnt5a by i.t. gp120 in the SDH shown above and the reported activity of Wnt5a in cytokine expression (41) suggest an interesting possibility that Wnt5a signaling regulates gp120-induced expression of cytokines. We performed experiments to test this idea directly, by determining the significance of Wnt5a in gp120-induced cytokine expression in the SDH. To this end, we first characterized the temporal effect of i.t. gp120 on the expression of IL-1β, IL-6, and TNF-α by Western blotting. Consistent with previous reports (9, 10), i.t. gp120 caused increase of IL-1β (Fig. 3A), IL-6 (Fig. 3B), and TNF-α (Fig. 3C). Marked cytokine increase occurred within 30 min after gp120 injection (100 ng/i.t.). This increase of cytokines was persistent for at least 24 h. To determine the potential role of Wnt5a in gp120-induced increase of IL-1β, IL-6, and TNF-α, we inhibited the Wnt5a signaling with Box5, which is a modified Wnt5a-derived hexapeptide that can specifically antagonize Wnt5a activity (53). In these experiments, Box5 or vehicle was intrathecally administrated 30 min prior to gp120 injection. Immunoblotting results showed that, although i.t. gp120 (100 ng, 12 h) significantly increased the expression of IL-1β (1.7-fold; p < 0.05) (Fig. 3D), IL-6 (1.7-fold; p < 0.05) (Fig. 3E), and TNF-α (2.0-fold; p < 0.05) (Fig. 3F) in the SDH, Box5 attenuated the increase of IL-1β, IL-6, and TNF-α, by 71%, 86%, or 100%, respectively (all with p < 0.05). Box5 alone did not affect basal cytokine levels. To confirm the observed effect of Box5-mediated Wnt5a blockage on gp120-induced cytokine increase, secreted Frizzled Related Protein 3 (sFRP3), a Wnt antagonist, was administered. Similar to Box5, sFRP3 (2 μg/i.t.) also significantly blocked the gp120-induced up-regulation of IL-1β (Fig. 3G), IL-6 (Fig. 3H), and TNF-α (Fig. 3I). These results suggest that Wnt5a is crucial for gp120 to induce cytokine expression. However, because Box5 and sFRP3 appear not to be able to completely block gp120-induced IL-1β increase, other pathways may contribute to gp120-induced up-regulation of this cytokine.

FIGURE 3.

Wnt5a is critical for gp120 to induce IL-1β, IL-6, and TNF-α in the SDH. A–C, time courses of IL-1β (A), IL-6 (B), and TNF-α (C) protein levels after i.t. gp120 (100 ng) administration. D–F, inhibition of Wnt5a by Box5 impaired gp120-induced increase of IL-1β (D), IL-6 (E), and TNF-α (F) protein in the SDH at 12 h after gp120 injection. G–I, sFRP3 attenuated gp120-induced increase of IL-1β (G), IL-6 (H), and TNF-α (I) protein levels in the SDH at 12 h after gp120 injection. Box5 (10 μg), sFRP3 (2 μg) or vehicle was intrathecally administered 30 min prior to 100 ng of gp120 or vehicle injection. In summary graphs, the levels of target proteins were normalized against β-actin and expressed as relative units to the control (0 h or vehicle). Data presented in graphs are means ± S.E. from at least 3 mice per group from three experiments. *, p < 0.05; **, p < 0.01 versus control (one way ANOVA).

Wnt5a Stimulated Cytokine Expression in the SDH

Having shown that Wnt5a is critical for gp120 to induce cytokines (Fig. 3), we next investigated if exogenous Wnt5a is sufficient to stimulate cytokine expression in the SDH. To this end, we tested the effect of i.t. Wnt5a on IL-1β, IL-6, and TNF-α. Western blotting analyses showed that i.t. Wnt5a (20 ng) significantly increased the protein levels of IL-1β (Fig. 4A), IL-6 (Fig. 4B), and TNF-α (Fig. 4C). Wnt5a-induced IL-1β and IL-6 expressions were time-dependent. A significant increase of IL-1β was observed at 30 min after Wnt5a administration (i.t.) (Fig. 4A). IL-1β continued to increase afterward until the end of observation at 24 h (Fig. 4A). Although a detectable but statistically insignificant increase of IL-6 appeared 30 min after Wnt5a injection, this increase did not reach a statistical significance until 6 h after Wnt5a administration (Fig. 4B). The Wnt5a-induced TNF-α increase peaked at 1 h after Wnt5a injection (Fig. 4C). These results indicate that Wnt5a by itself is sufficient to elicit cytokine expression in the spinal cord. The different temporal profiles of IL-1β, IL-6, and TNF-α expression indicate that Wnt5a may use distinct mechanisms to regulate the expression of different cytokines.

FIGURE 4.

Exogenous Wnt5a evokes the expression of IL-1β, IL-6, and TNF-α in the SDH. The time courses of IL-1β (A), IL-6 (B), and TNF-α (C) protein dynamics after i.t. Wnt5a (20 ng). In summary graphs, the levels of target proteins were normalized against β-actin and presented as relative units to 0 h. Data are expressed as means ± S.E. from at least three independent experiments with at least three animals. *, p < 0.05; **, p < 0.01 versus 0 h (one way ANOVA).

Wnt5a Activated αCaMKII and JNKs in the SDH

The above results suggested a role of Wnt5a in the control of cytokine expression in the SDH. Next, we wanted to understand the mechanism by which Wnt5a regulates cytokine expression. Previous studies in other experimental systems revealed that Wnt5a can activate the non-canonical Wnt signaling cascades such as Wnt/CaMKII (54), Wnt/PKC (55), and Wnt/JNK pathways (56). To elucidate the specific pathways that are activated by Wnt5a in the SDH, we measured the protein levels of phosphorylated forms of αCaMKII (pT286-αCaMKII), PKC [pT514-PKC (pan)), and JNK (pT183/Y185-JNK). Western blotting analyses showed that i.t. Wnt5a (20 ng) significantly increased phosphorylated αCaMKII (Fig. 5A) and JNK (Fig. 5B), but did not affect the phosphorylation of PKC (Fig. 5C). The pT286-αCaMKII significantly increased 1 h after Wnt5a administration, with a 2.2-fold of peak increase (p < 0.05) at 6 h, while pT183/Y185-JNK increase was significant 30 min after Wnt5a injection, with a 1.6-fold of peak increase (p < 0.05) for p-JNK1 and a 1.8-fold of peak increase (p < 0.05) for p-JNK2/3 at 12 h. These results indicate that Wnt5a specifically activates the Wnt/CaMKII and/or Wnt/JNK pathways in the SDH.

FIGURE 5.

Wnt5a increases phosphorylated forms of αCaMKII and JNK in the SDH. The time courses of pT286-αCaMKII (A), pT183/Y185-JNK (B), and p-PKC (C) proteins after i.t. Wnt5a (20 ng). In summary graphs, the levels of target proteins were normalized against β-actin and expressed as relative units to 0 h. Data are expressed as means ± S.E. from at least three independent experiments with at least three mice. *, p < 0.05; #, p < 0.05; ##, p < 0.01 versus 0 h (one way ANOVA).

gp120 Activated Wnt5a/CaMKII and Wnt5a/JNK Signaling Pathways

The data described above identified the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways in the SDH. Because gp120 up-regulated Wnt5a (Fig. 2), we next wanted to determine if gp120 also activated these pathways. Western blotting results showed that i.t. gp120 (100 ng) markedly increased pT286-αCaMKII (Fig. 6A) and pT183/Y185-JNK (Fig. 6B) in the SDH. pT286-αCaMKII increased significantly 1 h after gp120 administration, with a 3.5-fold of peak increase at 6 h (p < 0.01); the total αCaMKII did not show significant change during the course of experimentation (Fig. 6A). Similarly, the pT183/Y185-JNK (p-JNK1/2/3) also significantly increased from 0.5 to 24 h, but the total JNK did not. The p-JNK peaked at 1 h with a 1.8-fold increase (p < 0.05) for p-JNK1 and 2.7-fold (p < 0.01) for p-JNK2/3 (Fig. 6B). These results demonstrated that gp120 activated αCaMKII and JNK in the SDH.

FIGURE 6.

gp120 regulates the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways in the SDH. A and B, time courses of pT286-αCaMKII (A) and pT183/Y185-JNK (B) proteins after i.t. gp120 (100 ng). C and D, inhibition of Wnt5a signaling by Box5 blocked i.t. gp120-induced pT286-αCaMKII (C) and pT183/Y185-JNK (D) increase. E and F, effect of sFRP3 on gp120-induced pT286-αCaMKII (E) and pT183/Y185-JNK (F) increase. Box5 (10 μg), sFRP3 (2 μg) or vehicle was intrathecally administered 30 min prior to gp120 (100 ng) or vehicle. The SDH was collected at 12 h after gp120 injection. In summary graphs, the levels of target proteins were normalized with the β-actin loading control and expressed as relative units to the control (0 h or vehicle). Data presented in graphs are means ± S.E. from at least 3 mice. *, p < 0.05; **, p < 0.01; #, p < 0.05 (one way ANOVA).

Next, we further investigated if the gp120-induced activation of αCaMKII and JNK via Wnt5a signaling. To this end, we determined if the Wnt5a antagonists, Box5 and sFRP3, were able to block the gp120-induced phosphorylation of αCaMKII and JNKs. In these experiments, Box5 (10 μg) was intrathecally administered 30 min prior to i.t. gp120, and the lumbar dorsal spinal cord was dissected at 6 h after gp120 injection, when significant increase of gp120-induced phosphorylation of αCaMKII and JNK was observed (Fig. 6, A and B). Immunoblotting results showed that Box5 abolished the gp120-induced increase of pT286-αCaMKII (Fig. 6C) and pT183/Y185-JNK (Fig. 6D). Box5 alone did not affect the basal level of both pT286-αCaMKII and pT183/Y185-JNK. In consistence with the Box5 results, i.t. sFRP3 (2 μg) also significantly attenuated the gp120-induced increase of pT286-αCaMKII (Fig. 6E) and pT183/Y185-JNK (Fig. 6F). Taken together, these results suggest that gp120 activates the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways in the SDH.

Wnt5a/CaMKII and Wnt5a/JNK Pathways Differentially Regulated Cytokine Expression

Because Wnt5a not only was critical for gp120 to induce cytokine expression (Fig. 3) but also was sufficient to stimulate the expression of cytokines (Fig. 4), the findings that Wnt5a stimulated CaMKII and JNK activation (Fig. 5) and that gp120 activated the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways (Fig. 6) inspired us to further determine if Wnt5a controls cytokine expression through the CaMKII and the JNK signaling. To test this hypothesis, we used two widely used specific inhibitors KN-93 (57) and SP600125 (58) to block αCaMKII and JNKs, respectively. We determined the effects of these inhibitors on i.t. Wnt5a-induced expression of pro-inflammatory cytokines in the SDH. The PKC inhibitor stearoylcarnitine was also used as the control. KN-93 (25 μg), SP600125 (10 μg), and stearoylcarnitine (20 μg) were intrathecally administered 30 min before Wnt5a injection (i.t., 20 ng), and the lumbar dorsal spinal cord was dissected for immunoblotting at 12 h after Wnt5a injection. Under our experimental conditions, KN-93 inhibited Wnt5a-induced phosphorylation of αCaMKII (pT286) by 95% (Fig. 7A), while SP600125 suppressed Wnt5a-indced phosphorylation of JNKs (pT183/Y185) by 100% (Fig. 7B). Meanwhile, KN-93 suppressed Wnt5a-induced IL-1β increase by 79% (p < 0.01) (Fig. 7C). SP600125 and stearoylcarnitine did not have an inhibitory effect on the IL-1β increase (Fig. 7C). Either KN-93 or SP600125 significantly attenuated Wnt5a-induced increase of IL-6 (p < 0.01) (Fig. 7D). SP600125 showed a significant inhibitory effect on Wnt5a-induced up-regulation of TNF-α (p < 0.01), while KN-93 and stearoylcarnitine had no effect (Fig. 7E).

FIGURE 7.

Wnt5a/CaMKII and Wnt5a/JNK pathways differentially regulate the expression of IL-1β, IL-6, and TNF-α in the SDH. A and B, effects of KN-93 (KN) and SP600125 (SP) on i.t. Wnt5a-evoked pT286-αCaMKII (A) and pT183/Y185-JNK (B) in the SDH. C–E, effects of KN-93, SP600125, and stearoylcarnitine (ST) on Wnt5a-induced IL-1β (C), IL-6 (D), and TNF-α (E) in the SDH. F–H, effects of KN-93, SP600125, and stearoylcarnitine (ST) on gp120-induced IL-1β (F), IL-6 (G), and TNF-α (H) in the SDH. KN-93 (25 μg), SP600125 (10 μg), and ST (20 μg) were intrathecally administered 30 min prior to i.t. Wnt5a (20 ng) or gp120 (100 ng). At 12 h after Wnt5a, gp120 or vehicle injection, the SDH was collected. In summary graphs, the levels of target proteins were normalized with the β-actin loading control and expressed as relative units to the vehicle. Data in the graphs are presented as means ± S.E. *, p < 0.05; **, p < 0.01; ***, p < 0.001 (one way ANOVA).

The above results indicate that the CaMKII mediates the activity of Wnt5a in regulation of the expression of IL-1β and IL-6, while JNKs mediates the expression of TNF-α and IL-6 in the SDH. Next, we wanted to know if CaMKII and JNKs were critical for gp120 to up-regulate cytokine expression. As shown in Fig. 7F, the gp120-induced IL-1β up-regulation was markedly inhibited by 65% (p < 0.01) when αCaMKII was blocked by KN-93, while SP600125 and stearoylcarnitine did not. In addition, both KN-93 and SP600125 attenuated gp120-induced increase of IL-6, although the inhibitory effect of KN-93 did not reach statistical significance (Fig. 7G). Furthermore, only SP600125, neither KN-93 nor stearoylcarnitine, caused significant inhibition of gp120-induced TNF-α increase (p < 0.01) (Fig. 7H). The above results showed that gp120-induced cytokine increase requires αCaMKII or JNKs. Because Wnt5a antagonists (Box5 and sFRPs) blocked the gp120-induced cytokine expression (IL-1β, IL-6, and TNF-α) (Fig. 3) and the activation of αCaMKII and JNKs (Fig. 6), gp120 likely regulates the cytokines via Wnt5a signaling.

DISCUSSION

In this work, we report for the first time a critical role of Wnt signaling in regulation of the expression of pro-inflammatory cytokines in the CNS. Our results reveal that HIV gp120 activates the Wnt5a/CaMKII and Wnt5a/JNK signaling pathways in the SDH, and that these pathways differentially mediate gp120-induced expression of the cytokines. In particular, the Wnt5a/CaMKII pathway is critical for IL-1β expression whereas the Wnt5a/JNK pathway for TNF-α. Meanwhile, both pathways are involved in the expression of IL-6. These observations collectively suggest the following molecular mechanism of gp120-elicited neuroinflammatory response in the SDH (Fig. 7F): 1) gp120 up-regulates Wnt5a; 2) Wnt5a then activates CaMKII and JNKs; 3) the Wnt5a-activated CaMKII and JNK signaling not only stimulate the expression of IL-1β and TNF-α, respectively, but also co-regulate the expression of IL-6. This working model illustrates a Wnt5a signaling cascade that coordinates the neuroinflammaory responses to HIV gp120.

gp120 is a major viral coat protein that causes HIV-1-associated CNS neuroinflammation and neurotoxicity (59, 60). However, the molecular mechanism by which gp120 elicits neuroinflammation in vivo is still incompletely understood. We found that gp120 induced a rapid increase of Wnt5a protein within 5 min after i.t. gp120 (Fig. 2C). Emerging evidence reveals a role of Wnt5a in regulation of peripheral inflammatory response (32–34, 36, 61). More recently, we showed that Wnt5a is critical for Aβ-induced cytokine expression in cortical cultures (41). These prior findings suggest an interesting possibility that the gp120-induced up-regulation of Wnt5a is an early molecular event leading to gp120-elicited neuroinflammation. In support of this idea, we found that gp120-evoked IL-1β, IL-6, and TNF-α up-regulation were inhibited by the Wnt5a-specific antagonist (Fig. 3). Furthermore, i.t. Wnt5a was sufficient to stimulate the expression of these cytokines (Fig. 4). These results together suggested that Wnt5a signaling mediates the activity of gp120 in the induction of the cytokine expression.

We further elucidated the downstream mechanism through which Wnt5a controls the expression of the cytokines in the SDH. CaMKII and JNK are the key components of the non-canonical Wnt/CaMKII (54) and Wnt/JNK pathways (56), respectively. Our results indicated that Wnt5a activated CaMKII and JNKs (Fig. 5, A and B). Importantly, specific CaMKII or JNK inhibitors blocked the activity of Wnt5a in stimulating the expression of IL-1β, IL-6, and TNF-α (Fig. 7, C–E). These results indicate that the activation of CaMKII and JNK is necessary for Wnt5a to induce cytokine expression in the SDH. Our findings provide the initial evidence that the non-canonical Wnt5a signaling pathways play a crucial role in modulation of neuroinflmmatiion in the CNS, by controlling the expression of pro-inflammatory cytokines.

It is interesting to note that following gp120-injection (i.t.) Wnt5a was only transiently up-regulated, but the increase of phosphorylated CaMKII and JNKs were much longer-lasting. How did gp120-induced transient Wnt5a elevation have such a lasting effect? We do not know the answer at present. One way by which transient Wnt5a elevation could cause a lasting effect may involve autophosphorylation mechanism suggested for CaMKII persistent activation (Ref. 67). Alternatively, because CaMKII and JNKs are the downstream target of cytokines (e.g. IL-1β) (Ref. 68), another possibility is that the cytokines induced by the transient Wnt5a elevation can in turn stimulate CaMKII and JNKs, and thus the subsequent phosphorylation of these kinases could be sustained in a Wnt5a-independent manner by cytokines.

Our data suggest that Wnt5a/CaMKII and Wnt5a/JNK pathways differentially regulate the expression of pro-inflammatory cytokines. Whereas the Wnt5a/CaMKII pathway is required for the up-regulation of IL-1β (Fig. 7C), the Wnt5a/JNK critical for TNF-α (Fig. 7E). Interestingly, IL-6 is co-regulated by both pathways (Fig. 7D). Because previous studies reported that IL-6 was regulated by IL-1β and TNF-α (62–66), we cannot conclude if the Wnt5a-induced IL-6 up-regulation is a primary effect of the activation of the Wnt5a signaling or a secondary effect from the increase of IL-1β and TNF-α. The finding that the different pathways of the Wnt5a signaling cascade differentially regulate the expression of IL-1β, IL-6, and TNF-α illustrates a molecular mechanism for coordinating the action of the spectrum of pro-inflammatory cytokines in the CNS. Such coordination is presumably important for the expression of neuroinflammation.

It is known that in macrophages the activation of the Wnt5a/CaMKII pathway causes to the nuclear translocation of inflammatory transcription factors NF-κB and NF-AT (30). Pereira et al. reported that the Wnt5a/CaMKII pathway in macrophage stimulates cytokine expression (33). Recently Ouchi et al. (34) reported that Wnt5a/JNK signaling pathway mediated the expression of TNF-α and IL-6 in macrophages and promoted the inflammatory process in type II diabetes. Our data extend the previous findings by uncovering the critical roles of Wnt/CaMKII and Wnt/JNK in regulation of cytokine expression in the CNS.

Based on the observations described above, we propose that HIV gp120 in the CNS activates the Wnt/CaMKII and Wnt/JNK pathways to stimulate the expression of proinflammatory cytokines. This provides a molecular mechanism for gp120-induced neuroinflammation in vivo. This new mechanistic understanding may help design approaches to control HIV-1-associated chronic neuroinflammation in the CNS.