TNF-alpha-Induced Neuroregeneration through an NF-kappaB-dependent Pathway: A New Mechanism Involving EphB2 in the Context of HIV-1 Neuroinflammation

Abstract

The use of highly active antiretroviral therapy (HAART) has significantly decreased the mortality rate of HIV-1 patients, however the increased survival has led to the development of complications associated with the persistence of the viral infection. Nearly half of HIV-1-infected individuals develop HIV-associated neurocognitive disorders (HAND) as the effects of the chronic infection leads to neuronal injury and synaptic loss in the central nervous system (CNS). The neurotoxicity of HIV-1 has largely been attributed to the inflammation caused by viral replication and the altered signaling of astrocytes, microglia, and macrophages. Although HAART has improved the control of viral replication, the effects from inflammation remain a concern, particularly those of the pro-inflammatory cytokine, tumor necrosis factor alpha (TNF-α). TNF-α has been a therapeutic target for other diseases associated with chronic inflammation, such as rheumatoid arthritis, but emerging evidence has suggested that TNF-α signaling can have a dual role, especially in the CNS, proving the complexity in the modulation of the TNF-α pathway. Although the detrimental effects of TNF-α have been well-characterized, we lack a complete understanding of the beneficial role of TNF-α. TNF-α signaling has largely been considered to be neurotoxic but has been able to regulate neurite outgrowth in the context of neural development. Since TNF-α is upregulated in various neurodegenerative conditions, we considered potential outcomes of TNF-α on neurite outgrowth following injury. Initially, most would assume that TNF-α would prevent neurite outgrowth as apoptosis is a common outcome of TNF-α-induced signaling. If TNF-α signaling strictly prevents neurite outgrowth, anti-TNFα therapies could be considered to reverse this effect. However, upon induced injury, we observed an increase in neurite regrowth following induced injury in human primary fetal neurons, demonstrating a strong need for a deeper understanding of this dual role of TNF-α. Anti-TNF-α therapies have been considered for HIV-1-infected patients to reduce the chronic inflammation, however inhibiting TNF-α signaling could have side-effects that could prevent neuronal recovery from HIV-1 effects. Targeting pathways downstream of TNF-α signaling would be more advantageous to mediate the beneficial role of TNF-α in the CNS. We investigated the transcriptional effects of TNF-α treatment on neurons to uncover a potential pathway to promote neurite outgrowth. One pathway we have discovered to be beneficial in primary human fetal neurons is TNF-α-induced Ephrin B2 upregulation. Ephrin B2 (EphB2) receptors are important mediators of neuronal development and synaptic plasticity, however little has been established in regards to their role in HIV and inflammation, particularly in the CNS. EphB2 can mediate axonal development by providing retractive cues to assist the axon to reach the target, but EphB2 can also promote dendritic branching to improve learning and memory, which would be particularly beneficial for HAND patients that experience cognitive deficits. We observed a correlation between the upregulation of EphB2 in response to TNF-α and neurite outgrowth, which provides a potential pathway to repair damaged neurons and re-establish lost neuronal connections. Dendritic pruning and neuronal loss has been observed in HAND patients, so this ability to promote repair could prevent, improve, or recover the cognitive deficits experienced by HIV-patients with HAND. TNF-α, although primarily known to induce neurotoxicity, strongly activates the nuclear factor-kappaB (NF-κB) pathway, which can have a very wide range of transcriptional effects. Therefore, our hypothesis is that the TNF-α-induced neurite regrowth occurs through an upregulation EphB2 in an NF-κB-dependent pathway. TNF-α has been well established to induce NF-κB signaling, mostly by promoting the translocation of the NF-κB p65 DNA binding factor to the nucleus for transcriptional regulatory effects. NF-κB can regulate neuronal growth and process development of both dendrites and axons, which would correlate to the neurite regrowth observed following TNF-α upon induced injury. The regulation of EphB2 by NF-κB has not been extensively studied, but EphB2 can be negatively regulated by an NF-κB family member, c-Rel. We analyzed the EphB2 promoter and identified three NF-κB p65 binding sites upstream from the transcriptional start site, which provided insight to our hypothesis. We established that p65 directly binds to and can regulate EphB2 promoter activity in response to TNF-α. Since the dual role of TNF-α can be dependent on the receptor through which the signaling proceeds, either TNF-α receptor 1 (TNFR1) or TNF-α receptor 1 (TNFR2), we investigated if this upregulation of EphB2 is receptor dependent and determined EphB2 is induced primarily through activation of TNFR2. Neurons express both receptors, however, the effects of TNF-α to promote neuroprotection and repair primarily occur through the TNF-α/TNFR2 regulatory axis. Although we have been established the mechanism of TNF-α-induced EphB2 and there is a strong correlation with neurite outgrowth following induced injury, we considered the possibilities to modulate EphB2 in the absence of TNF-α to demonstrate the direct effects of EphB2 expression. Several approaches could be used to mediate EphB2 activation or inhibition in vitro. RNA interfering techniques, such as small interfering RNA (siRNA), are useful, but we were interested in a complete knockout strategy. Since our approach was to assess the effects of EphB2 knockout only on neurite outgrowth following induced injury, a knockout animal model would not be appropriate, as a lack of EphB2 would affect the development of the neurons, unless an inducible knockout model was established. This is a lengthy and elaborate process and, more importantly, would only be available in a non-human model. Other techniques, such as transcription activator like effector nucleases (TALENs), can generate knockout systems that are targeted to specific regions of a gene, but specific binding proteins must be created to recruit the endonucleases to the target. Clustered regularly-interspaced short palindromic repeats/CRISPR associated protein 9 (CRISPR/Cas9) has emerged as a specific and relatively easy technique to knockout genes of interest and uses short RNA sequences to guide Cas9 endonucleases to target regions to create double stranded breaks in the DNA to silence the gene. Once concern with Cas9 is specificity to target only the desired region of the gene, as off-target effects can occur and may result in unwanted gene silencing. A Cas9 mutant, Cas9 nickase (Cas9n), has been created to have more specificity by requiring two guide RNAs to recruit two Cas9 nickases to generate a double stranded break as they function as nickases to only create a nick in one DNA strand. We developed this strategy to remove exon 1 of the EphB2 gene by using two pairs of Cas9 nickases, with four guide RNAs, to eliminate any chance for off-target effects but retaining the desired outcome of and EphB2 knockout. We validated the system by demonstrating that a knockout of EphB2 increases adhesion and prevents migration in human embryonic kidney 293T (HEK293T) cells. Although this cell model is not a neuronal cell model, the migration assay demonstrates the functional loss of EphB2. We also created an inducible Ephb2 system to overexpress EphB2. Together these provide essential tools to verify the direct involvement of EphB2 in neurite outgrowth. Taken together, our studies characterize a novel mechanism for neurite outgrowth following injury in neurons: TNF-α/TNFR2-induced EphB2 signaling in an NF-κB p65-dependent manner. In addition to the established mechanism, we developed a technique to assess the effects of EphB2 knockout and overexpression in the context of neurite outgrowth: EphB2-targeted-Cas9n and EphB2 inducible construct. This mechanism yields insight into a potential downstream pathway to be utilized to repair damaged regions in the brain and reverse cognitive deficits in neurodegenerative conditions, especially in a chronic inflammatory environment, such as HIV-1 infection. The strategies created provide a valuable toolset to demonstrate the direct effects of modulating EphB2 signaling, not only in neurons for effects on neuronal health and synaptic plasticity, but also in other disease models, such as glioblastoma, in which EphB2 was demonstrated to promote invasion and migration of tumor cells. These observations and the usefulness of the modulatory strategies likely extend to multiple neurodegenerative diseases that demonstrate cognitive deficits that correlate to neuroinflammation.