Chronic inflammation in HIV-infected individuals has been associated with insulin resistance, cardiovascular diseases, thrombosis, muscle wasting, neurocognitive impairment, renal and liver diseases, obesity, and premature aging (1–10). Constant production of viral proteins (Tat, Nef, and gp120) and elevated pro-inflammatory cytokine levels (TNFα, IL-1β, IL-6, and IL-8) contribute to persistent neuroinflammation and are associated with cognitive impairment in HIV-infected patients (9, 11–17).

HIV-1 Tat is continuously produced in patients even in the presence of viral suppression, thus promoting the persistence of chronic inflammation in both the peripheral and the central nervous system (18). HIV-1 Tat protein is known to induce TNFα production and high levels of this cytokine are present throughout all stages of HIV infection (19, 20). Although essential for the innate immune response to acute inflammation, persistent elevated levels of TNFα contribute to the maintenance of the chronic neuroinflammation observed in HIV-infected patients. These prolonged high levels of TNFα can lead to glutamate-mediated neuronal toxicity (21) as well as stimulate the release of neurotoxic factors, and cell signaling cascades that lead to neuronal dysfunction and death (13, 20, 22, 23). The combined neuronal damage of viral proteins and cytokines is implicated in the pathogenesis of HIV-associated neurocognitive disorders (HAND) (24–28).

Overproduction of pro-inflammatory cytokines has also been implicated in the pathophysiology of insulin resistance and metabolic disorders (29, 30). TNFα can mediate metabolic abnormalities in different tissues by inducing serine phosphorylation of insulin receptor substrate-1 (IRS-1) protein, decreasing the expression of insulin signaling molecules, and down-regulating the expression of the insulin regulated glucose transporter, GLUT 4 (29, 31, 32). Insulin receptor is expressed on the surface of neurons and its activation can lead to increased neurite growth, gene expression, and energy metabolism. IRS-1 is the major substrate of the insulin receptor (33). Abnormalities in this pathway have been linked with neurodegeneration including Alzheimer's disease (34). However, its role in the pathophysiology of HAND has not been explored.

It is known that people living with HIV (PLH) may develop insulin resistance, HAND, or both, as a result of chronic HIV infection, long-term use of combined antiretroviral therapy (cART), aging, genetic predisposition, or a combination of these factors (35–37). Thus, identification of molecular and cellular pathways linking the high incidence of insulin resistance and HAND in PLH is crucial. Defects in the insulin receptor signaling pathway represents a major target for modulating critical energy functions such as glucose and lipid metabolism, neuronal growth and survival, and synaptic plasticity (38–40). Recently, we showed that high plasma and CSF soluble insulin receptor (sIR) levels are associated with the presence and severity of HAND in a cohort of HIV-infected women on cART (41, 42). Moreover, we observed an increased binding of plasma insulin to sIR in patients with worse cognitive performance (42). Hence, there is a need to identify the factors that contribute to the secretion of sIR into the CSF. In this study, we investigated if the CSF from HIV-infected women with HAND influences the secretion of sIR from human neuronal cells. In addition, we evaluated whether this secretion was regulated by HIV-1 Tat protein, TNFα, or other selected cytokines.

The mechanisms involved in the pathophysiology of insulin resistance and cognitive impairment in HIV-infected participants remain unclear. Several mechanisms may explain the generation of the CSF sIR, including secretion from astrocytes, neurons, or PBMCs that cross the blood-brain barrier (BBB), brain tissue injury, or plasma sIR that crosses the BBB (41). We previously showed that higher levels of plasma and CSF sIR full-length were associated with HAND in a population of HIV-infected women (41). In addition, we observed that plasma sIR served as a scavenger receptor through the binding of higher percentages of insulin in patients with the severe forms of HAND (42). These findings suggest that changes in plasma sIR could be associated with an asymptomatic glucose derangement and may contribute to the development of insulin resistance and HAND (41). The observation of a significant IRS-1 tyrosine phosphorylation decrease suggests the possibility of chronic subclinical cellular insulin resistance, which could contribute to the development of HAND (38).

Although the generation of intact and other sIR isoforms was observed in several cultured cell lines (47), their secretion from neuronal cells has never been explored. In our study we exposed human neuronal SH-SY5Y cells to the CSF of HIV-infected women and found that these cells secreted higher sIR levels compared to the CSF of HIV-seronegative controls. When stratified by cognitive status, this increased secretion was associated with the severity of HAND. The mediators that contributed to this secretion were unknown. Thus, identifying the factors that influence the secretion of sIR to the CSF may help in the development of novel strategies for early intervention and, possibly, the prevention of insulin resistance progression and HAND in this population.

Increasing evidence suggests that inflammation is actively involved in the neuropathogenesis of HAND, and several cytokines have been identified as potential contributors. The importance of these cytokines, especially TNFα, has been emphasized in the propagation and maintenance of a CNS inflammatory response. We were able to quantify six cytokines in the CSF of our cohort of HIV-infected women and observed that the pro-inflammatory IL-2, IFNγ, TNFα, and the anti-inflammatory IL-10 were associated with the presence and severity of HAND. These results are consistent with other studies that show an association between cognitive function and increased levels of IFNγ and TNFα in the CSF of several ethnic populations of people living with HIV (PLH) (13, 48–50). Experimental evidence also shows that higher levels of IL-6 in CSF of HIV-infected patients with HAND confirm its specific role as mediator of neuroinflammation during chronic HIV inflammation (51). In addition, several studies identified possible pathways involved in the generation of IL-6 from brain microvascular endothelial cells, as well as astrocytes (52, 53). Interestingly, when the neuronal cells were exposed to these elevated cytokines, only TNFα was able to induce higher secretion of sIR from neuronal cells. This finding was supported by the fact that R7050, a TNFα receptor blocker, when added to the culture medium for 24 h, antagonized the effects of TNFα in sIR secretion. This is in accordance with previous reports showing that TNFα can induce the secretion of other soluble receptors in the culture medium (54).

Cognitive impairment in HIV-infected patients may be caused directly by the action of the HIV virus or its well-known auxiliary proteins (55). HIV-1 Tat, Rev, Nef, Vif, Vpr, and Vpu can promote neurotoxicity through several mechanisms which result in neuronal dysfunction and cell loss (16, 56). Of these viral proteins, HIV-1 Tat (one of the first HIV proteins to be expressed after infection occurs), has been shown to have a deleterious role in the development and progression of HAND. Indeed, HIV-1 Tat protein can be released from infected cells even in patients who are adequately treated with antiretroviral drugs to affect HIV non-permissive cells such as neurons (18, 57). In our study, we found that HIV-1 Tat increased the secretion of sIR from neuronal cells and R7050 was able to antagonize this effect. The mechanisms employed by HIV-1 Tat on sIR secretion can be explained by its interaction with the toll-like receptor 4 (TLR4). Studies have revealed that HIV-1 Tat interacts physically (with high affinity) with TLR4-MD2 to promote TNFα expression (11). Other studies also show that HIV-1 Tat can induce the secretion of TNFα and IL1β from U87 glioma cells (58). Thus, it is then conceivable that the autocrine effects of TNFα induced by HIV-1 Tat can be conceptualized as a part of the physiological process used by the toxin to regulate neuronal function during HIV infection. Previously we published that sIR acted as a scavenger receptor for insulin promoting intracellular insulin resistance (42). In this study, we observed that the release of sIR is induced by Tat and TNFα suggesting an effect on the insulin signaling pathway and contributing to the neuropathogenesis of HAND. These findings suggest that the replacement of insulin alone such as intranasal insulin may not be sufficient for treatment of these patients.

Our study has several limitations. First, the sample size is small, particularly for the sub-group analyses that we performed. We believe that our sub-group comparisons were sufficiently sized to generate meaningful conclusions; however, our findings should be interpreted with caution. The need to replicate these findings is important. Second, for the purpose of this study we focused on the role of HIV-1 Tat and inflammatory cytokines in the release of sIR from neuronal cells. The presence of glutamate in the CSF and the possible influence of TNF and Tat combination on the secretion of sIR was not measured. Third, we did not adjust for multiple comparisons in our analysis of sIR secretion from neuronal cells. We believe that adjustment was not necessary for two reasons: (1) we conducted planned comparisons that each asked the same basic question, and (2) there was consistency in the results with all comparisons pointing to the same conclusion. Fourth, we only studied a cohort of women with HIV infection since our cohort was designed to study the effects of HIV in women. Further studies to address sex differences are warranted. Fifth, our study design did not consider differentiating the chronic vs. acute effects of HIV-1 Tat. Future studies comparing this effect should be considered.

Soluble receptors, including insulin, may be secreted from cells in several forms such as intact or with an associated cell membrane. Recent studies have observed that the latter form may be secreted as a vesicle as an exosome (59, 60). In this study we proposed to determine the association of the sIR as measured previously in Gerena et al. (41, 42) and inflammation. We understand that further studies including the determination of exosomes should be performed.

Our study provides evidence that the activation of the TNFα receptor plays an important role in the secretion of sIR from neuronal cells. Both TNFα and HIV-1 Tat regulate the secretion of sIR from neuronal cells. Further studies will investigate whether the effects of HIV-1 Tat on sIR secretion are mediated by neuronal TNFα being released to the extracellular medium.

YG, RM-D, JV, AN, and VW were responsible for the conception and design of the study. YG, RM-D, AD-N, JM-Á, JS-P, LH, AN, and VW contributed to the data acquisition. RS, YG, RM-D, AN, and VW were involved in the analysis and interpretation of data. YG, RM-D, and VW drafted the first version of the article. All authors revised it critically for important intellectual content. All authors gave final approval of the version to be submitted.