Osteopontin counters human immunodeficiency virus type 1-induced impairment of neurite growth through mammalian target of rapamycin and beta-integrin signaling pathways

Abstract

Despite the fact that human immunodeficiency virus type 1 (HIV-1) does not enter or replicate in neurons, its infection of a subset of resident brain glia cells (microglia and astrocytes) induces via disparate mechanisms, dysregulation of glutamate metabolism, neurotoxicity, and inflammation. Antiretroviral therapies suppress viral load, but cellular activation and release of proinflammatory factors, some of which is likely related to viral reservoirs, continue to promote a microenvironment that is injurious to neurons. However, the molecular mechanisms remain to be identified. Osteopontin (OPN) is a proinflammatory cytokine-like, extracellular matrix protein that is elevated within the brain and CSF in several neurodegenerative disorders, including HIV-associated cognitive disorder. However, the impact of elevated OPN on neuronal integrity and function in HIV-infected individuals who exhibit cognitive dysfunction remains unknown. In this study, using a neuronal cell line and primary cultures of cortical rat neurons, we identify the mammalian target of rapamycin pathway involvement in a signaling interaction between OPN-β1-integrins and the HIV-1 envelope glycoprotein, which stimulates neurite growth. These findings link for the first time HIV X4-envelope receptor engagement and osteopontin-mediated signaling through β1-integrin receptors to the mTOR pathway and alterations in the cytoskeleton of cortical neurons.

Keywords: Cytoskeleton; Dendrites; HIV-associated neurocognitive disorder; Integrins; Neurons.

Fig. 1

gp120 from CCR5-using HIV_BaL or HIV_SF162, but not CXCR4-utilizing strain HIV_IIIB, promotes neurite growth in retinoic acid–differentiated SHSY5Y neuronal-like cells. SHSY5Y cells differentiated for 7 days with 10 um retinoic acid were treated with 50–400 pM of recombinant HIV-1 protein from the X4-tropic IIIB or R5-tropic BaL or SF162 strains for 48–72 h before immunofluorescent staining for β-III-tubulin. Data shown are the average of five independent assays per group. GraphPad Prism was used to determine statistical significance by one-way ANOVA and subsequent Tukey’s test compared with control as indicated. Arrows indicate shortened axons, and stars show axons with increased length. Quantification of axonal length. Values represent the mean ± SD of axonal length (μM), calculated from measurement of 20 axons per group. Tukey’s test or Dunn’s test compared with control. a Control. b IIIB Env. c BaL Env. d None–IIIB200, p = .0187; none–IIIB400, p = .0032; IIIB50–100, p = .029; IIIB50–200, p = .0014; IIIB50–400, p = .0002. e None–BaL100, p = .0335; none–BaL 200, p = .0011; none–BaL 400, p = .0002; BaL50–400, p = .0444. f None–SF162 200, p = .0023; none–SF162 400, p < .0001; SF162 50–400, p = .0011; SF162 100–400, p = .0083. g None–IIIB50, p < .0001; IIIB50–BaL50, p = 0.0092; IIIB50–SF162 50, p = 0.0028. h None–BaL100, p = 0.0005; None–SF162 100, p = 0.0497.i None–BaL200, p < .0001; None–SF162 200, p = 0.0036; IIIB200–SF162 200, p = 0.0015. j None–BaL400, p < .0001; none–SF162 400, p < .0001; IIIB400–BaL 200, p = 0.0002; IIIB400–SF162 200, p = 0.0002

Fig. 2

OPN reverses HIV IIIB gp120–induced decreases in neurite length and potentiates the impact of R5-tropic Env in retinoic acid–differentiated SHSY5Y neuronal-like cells. SHSY5Y cells were differentiated, treated, subjected to IF, and analyzed as described in Figure legend 1 in the presence or absence of OPN at 50–200 ng/ml and/or with HIV BaL Env or IIIB gp120 at 400 pM as indicated. Arrows indicate shortened axons, and stars show axons with increased length. GraphPad Prism was used to determine statistical significance by one-way ANOVA and subsequent Tukey’s test or Dunn’s test compared to control (parametric, OPN, Fig. 2g) or Krusalis-Wallis (non-parametric, IIIB, Fig. 2e). Representative images of treated cells stained for β-III-tubulin: a control, b IIIB only, c BaL and OPN 100 ng/ml, d OPN 100 ng/ml. e IIIB: IIIB–OPN100, p = .0049; IIIB–OPN200, p = .0001; IIIBOPN50–OPN100, p = .0452; IIIBOPN50–OPN200, p = .0018. f BaL: none–BaLOPN100, p = .0002; none–BaL OPN200, p < .0001; BaLOPN50–BaLOPN100, p = .0412; BaLOPN50–BaLOPN200, p = .0046; BaL–BaLOPN200, p = .0182. g OPN: none–OPN100, p = .0198; none–OPN200, p = .0377

Fig. 3

Osteopontin (OPN), in a dose-dependent manner, blocks HIV_IIIB Env inhibition of neurite growth and significantly increases the expression of dendrites in cultured primary rat cortical neurons. a, b, c, d E18 primary rat neurons isolated from the prefrontal cortex were differentiated for 4–5 days before treatment with increasing concentrations of OPN (6.25 pM, 12.5 pM, 25 pM, 50 pM, 100 pM, or 200 pM) in the presence or absence of 400 pM HIV IIIB envelope as indicated and stained and quantified for either β-III-tubulin or MAP2 expression. a Representative images of stained neurons. b None–IIIB, p < .0001; none–IIIB50, p < .0001; none–IIIB100, p < .0001; none–IIIB200, p < .0001; IIIB only–IIIB6.25, p < .0001; IIIB only–IIIB12.5, p < .0001; IIIB only–IIIB25, p < .0001; IIIB only–IIIB50, p < .0001; IIIB only–IIIB100, p < .0001; IIIB only–IIIB200, p < .0001; IIIB6.25–IIIB50, p < .0001; IIIB6.25–IIIB100, p < .0001; IIIB6.25–IIIB200, p < .0001; IIIB12.5–IIIB50, p = .0029; IIIB12.5–IIIB100, p < .0001; IIIB12.5–IIIB200, p < .0001; IIIB25–IIIB50, p < .0001; IIIB25–IIIB100, p < .0001; IIIB25–IIIB200, p < .0001; IIIB50–IIIB100, p < .0001; IIIB50–IIIB200, p = .0002. c None–OPN6.25, p = .0053; none–OPN25, p < .0001; none–OPN50, p < .0001; none–OPN100, p < .0001; none–OPN200, p < .0001; OPN6.25–OPN50, p < .0001; OPN6.25–OPN100, p < .0001; OPN6.25–OPN200, p < .0001; OPN12.5–OPN25, p < .0001; OPN12.5–OPN50, p < .0001; OPN12.5–OPN100, p < .0001; OPN12.5–OPN200, p < .0001; OPN25–OPN50, p = .0101; OPN25–OPN100, p < .0001; OPN25–OPN200, p < .0001; OPN50–OPN200, p = 0.106. d None–IIIB6.25, p = .0087; none–OPN100, p = .0119; IIIB–IIIB25, p = .0402; IIIB–OPN100, p < .0001; IIIB6.25–IIIB25, p = .0006; IIIB6.25–IIIB50, p = .0038; IIIB6.25–IIIB100, p = .0017; IIIB6.25–OPN100, p < .0001. e Blocking β1-integrin receptor interferes with OPN modulation of β3-tubulin levels in cultured primary rat cortical neurons. Differentiated rat cortical neurons were treated with inhibitory antibodies against β1- or β3-integrin in the absence or presence of 100 ng/ml OPN and the level of β-III-tubulin expression quantified at 48 h post-treatment. OPN-anti-β1, p = .0407

Fig. 4

OPN-induced increases in cortical neurite expression in the presence or absence of HIV IIIB Env are blocked by the mTOR inhibitor rapamycin. Differentiated rat cortical neurons in poly-D-lysine-coated 24-well plates were treated in triplicate with increasing concentrations of OPN (6.25 pM, 12.5 pM, 25 pM, 50 pM, 100 pM, or 200 pM) in the presence or absence of 400 pM HIV IIIB envelope and rapamycin (20 nM) as indicated. To obtain sufficient protein for western analyses, the triplicate wells were pooled and 1–2 μg of protein was loaded onto 4–12% Bis-Tris gradient gels for SDS-PAGE and western blot analyses for quantification of β-III-tubulin and β-actin expression. a IIIB25–100, n = 3, p = .0193; b OPN6–25 w/wo rapamycin, n = 3, p = .0004. OPN25–100 w/wo rapamycin, p = .0199

Fig. 5

In HIV envelope–treated rat cortical neurons, low, but not higher, levels of OPN activate mTORC1. Differentiated rat cortical neurons were prepared and treated as described in figure legend 4 and subjected to western analyses for mTOR S2448 and mTOR total protein expression. The differences were not significant

Fig. 6

Activation of p70 S6 kinase, a downstream target of mTORC1 signaling by low doses of OPN in the presence or absence of HIV Env in cotreated cortical neurons. Differentiated rat cortical neurons were prepared and treated as described in figure legend 4 and subjected to western analyses for p70 S6 kinase (Ser371), pS6 ribosomal protein (Ser235/236), and β-actin expression. a IIIB6.25–25, with/without rapamycin n = 3, p = .0138; b OPN6–25 w/wo rapamycin, n = 3, p = .0355

Fig. 7

Activation of the downstream mTORC2 substrate, stress-glucocorticoid kinase 1 (SGK1) in HIV Env-OPN-cotreated cortical neurons. Differentiated rat cortical neurons were prepared and treated as described in figure legend 4 and subjected to western analyses for stress- and glucocorticoid-regulated kinase-1 (SGK1) Ser422, SGK1 total protein, and β-actin expression. a IIIB25–100, with/without rapamycin n = 3, p < .0001; and rapamycin-IIIB6.25–25 and rapamycin-IIIB50–100, n = 3, p < .0001. b OPN6–25 w/wo rapamycin, n = 3, p = .0202; OPN6–25 and OPN25–100, p = .0192

Fig. 8

AKT phosphorylation is activated via an mTORC2-mTORC1 pathway interaction, in a PDK-independent manner. Differentiated rat cortical neurons were prepared and treated as described in figure legend 4 and subjected to western analyses for the expression of protein kinase B (Akt Ser473), total Akt protein, phosphoinositide-dependent kinase-1 (PDK1 Ser421) and total PDK protein levels. a IIIB6.25–25, with/without rapamycin n = 3, p < .0001; IIIB25–100, with/without rapamycin n = 3, p < .0001. b IIIB25–100, with/without rapamycin n = 3, p = .0207

Fig. 9

Osteopontin (OPN)-induced β1-integrin activation and HIV-1 envelope signaling converge on mTORC1/2 pathways in a mechanism that promotes neurite outgrowth. HIV X4-tropic envelope protein is a well-known neurotoxin. A novel phenotype of increased neurite growth was observed in primary rat cortical neurons exposed to increasing doses of osteopontin in the presence or absence of HIV X4-tropic envelope. Osteopontin promoted neurite growth as assessed by increases in b3-tubulin and MAP2 staining. Increases in osteopontin-mediated b3-tubulin expression were inhibited by rapamycin, a well-known inhibitor of mTORC1, thus implicating this pathway in the mechanism of action. The signaling cascade can be initiated through β1-integrin receptors, as functional antibody blocking studies abrogate the effect of osteopontin. A modest level of activation of downstream mTORC1 substrate p70 S6 kinase was found suggesting activation of mTORC1. However, in the presence of HIV IIIB envelope and osteopontin, SGK1 and Akt, both established targets of mTORC2 signaling, are activated. Interestingly, the activation of SGK1 in neurons cotreated with HIV IIIB envelope and higher levels (50–100 pM) of osteopontin are insensitive to rapamycin inhibition confirming the activation of mTORC2. In contrast, low levels of osteopontin (6–12.5 pM) induce SGK activation in a manner that depends on mTORC1 signaling. Collectively, our data suggest that there is a feedback loop operating between mTORC1 and mTORC2, the outcome of which is increases in neurite growth. The mTORC1 and mTORC2 pathways as well as signaling via β1-integrin receptors have been implicated in the regulation of actin cytoskeletal dynamics, learning and memory, axon guidance, and synaptic plasticity. Therapeutic strategies which can modulate osteopontin expression in the central nervous system in the context of neuronal dysfunction may be advantageous in addressing deficits accompanying neurodegenerative processes. Image rights: © 2018 Johns Hopkins University, by Lydia Gregg