Neurovirulent and non-neurovirulent strains of HIV-1 and their Tat proteins induce differential cytokine-chemokine profiles

Materials

PBMCs were purified from leukopaks from New York Blood Center. SHSY-5Y neuroblastoma cells were obtained from ATCC (Catalog number – CRL-2266) and maintained in a 1:1 mixture of Eagle’s Minimum Essential Medium and F12 Medium with 10 % serum. Recombinant purified Tat proteins were from HIV-1_BL43 [9], a subtype C virus and HIV-1_YU2, a subtype B virus and were generated as described previously [9].

Isolation of peripheral blood mononuclear cells

Peripheral blood mononuclear cells (PBMCs) were isolated from HIV-1 negative donor leukopaks (New York Blood Center) using density gradient centrifugation. Leukopaks were diluted in RPMI/FBS and overlaid over a Ficoll-Hypaque (Sigma-Aldrich, St. Louis, MO) gradient. Following centrifugation at 400×g for 30–40 min at room temperature with no brakes, PBMCs were recovered from the Ficoll-Hypaque interface, diluted in PBS, and centrifuged at 200×g for 10 min at room temperature to eliminate platelets. PBMCs were re-suspended in Gibco RPMI1640 (ThermoFisher), supplemented with 10 % FBS (Gibco) and 1 % Penicillin and Streptomycin. Cells were incubated at 37 °C in T75 flasks.

HIV-1 infection of PBMCs

PBMCs were activated with PHA (5 μg/ml) and IL-2 (100 IU/ml) for 48 h prior to infection. HIV-1_ADA and HIV-1_IndieC1 virus stocks were generated by transfecting the HEK 293T cells with infectious molecular clone DNAs. Virus released into the medium was quantitated by the p24 Enzyme-linked immune-sorbent assay (ELISA; Applied Biosystems, Inc. Waltham, MA), as described previously [6]. Approximately 50 × 10⁶ PBMCs were infected with viral stocks at low multiplicity of infection (5 and 10 ng of HIV-1_ADA and HIV-1_IndieC1 respectively) to achieve comparable levels of viral infectivity at the assay end point 4-h or five days following infection as described previously [6].

Immunodepleting Tat and gp120

HIV-1 Tat and gp120 were immuno-depleted from HIV-infected spent media from PBMCs using broadly neutralizing anti-gp120 (VRC 03, NIH AIDS Reagent Repository) and anti-Tat E1.1 [9] antibodies, as described previously [6], 7]. VRC03 was chosen because of its broadly neutralizing capacity and its availability and E1.1 monoclonal antibodies were selected because of their ability to neutralize both subtype B and C Tat proteins [7]. E1.1 mAbs were raised in BALB/c mice using recombinant full-length Tat protein. E1.1 mAb was of IgG1 isotype. Immune characterization showed that E1.1 displayed a Tat binding affinity, with a K_d value of 1.9 nM and in Tat-neutralization properties in preventing the Tat-mediated transactivation in HLM-1 cells (Dr. Ranga, unpublished results).

For immunodepletion, Pansorbin beads (Millipore Sigma, Burlington, MA; Cat no. 507861) were incubated with anti-gp120 and anti-Tat antibodies for an hour on ice. Antibodies were used in excess, at 3-fold (for VRC03) or 100-fold excess (for E1.1), concentrations (VRC03 and E1.1 antibodies were both used at 300 ng respectively) to ensure complete protein depletion from spent media. The beads were centrifuged, re-suspended in 25 μL RPMI, and added to 10 ml of spent media. Following incubation for an hour on ice, the Pansorbin beads are centrifuged at 3,000 rpm for 5 min at 4 °C. The antigen-depletion step was repeated with a fresh round of antibodies to ensure the complete removal of gp120 and Tat proteins from the supernatant. The antibody-treated supernatants were used in the in vitro neurotoxicity assays.

Neurotoxicity assay

SH-SY5Y neuroblastoma cells were obtained from ATCC (CRL-2266) and maintained in a 1:1 mixture of Eagle’s Minimal Essential Medium and F12 Medium supplemented with 10 % fetal bovine serum. Approximately 250,000 cells/well were plated in 12-well plates and allowed to stabilize for three days. Three hundred µl (300 µL) of HIV-infected PBMC supernatants (with or without immuno-depletion) were added to 600 µL of SH-SY5Y media and then added to previously plated SH-SY5Y cells for 24 h as described previously [6]. Neuronal apoptosis and viability were assessed, in triplicate wells, using the WST-1 assay (Roche, Basel, Switzerland), which measures mitochondrial activity, providing a robust indication of cell viability and death. The percentage of cell death, compared to control cells treated with uninfected PBMC supernatant, was calculated as previously described [6].

To ensure comparable levels of viral particles in the spent media, we first measured virus replication via p24 ELISA in the HIV-1_ADA and HIV-1_IndieC1 infected PBMC cultures. This calibration allowed for equivalent exposure in the neurotoxicity assays. Viral release was standardized by infecting PBMCs with pre-determined virus inputs (5 and 10 ng of p24 for HIV-1_ADA and HIV-1_IndieC1, respectively) and confirmed by measuring p24 levels in the media at peak virus production points (day 15). These combined approaches ensured that observed neuronal death was attributable to the effects of viral strain differences and associated cytokine-chemokine profiles, rather than variability in infection parameters.

Luminex ELISA and CCL2 ELISA assays

Thirteen different cytokines/chemokines (Interferon-α, CCL-2, CCL-3, CCL-4, IL1-α, IL-1β, IL-4, IL-6, IL-8, IL-10, TNF-α, IP-10 and RANTES) were measured using a high-sensitivity human cytokine array kit. Each sample was assayed in triplicate wells, and the cytokine standards supplied by the manufacturer were included on each plate. Data was acquired using a Luminex-100 system and analyzed using the Bio-Plex Manager software. As CCL2 levels were above the limit of detection of the cytokine array kit, a separate CCL-2 ELISA was performed using the Invitrogen Human CCL2 ELISA kit, as described previously [7], 8].

Tat treatment of PBMCs and real-time PCR

Purification of recombinant HIV-1 Tat proteins from the neurovirulent Subtype-B YU-2 strain and non-neurovirulent Subtype-C BL-43 strain has been reported previously [9]. Neurotoxic properties of these Tat proteins were described previously [10]. PBMCs of healthy subjects were activated with PHA (5 μg/ml) for 24 h, before treating them with 200 ng/ml recombinant Tat protein for 4 h at 37 °C. PBMCs were pelleted, and RNA was extracted using RNAeasy Kit (Qiagen, Germantown, Maryland). cDNA was generated from the RNA using a Taqman reverse transcriptase cDNA kit (Applied Biosystems Inc). cDNA was quantitated, and a qRT-PCR was performed using a Taqman cytokine gene expression SuperArray plate (Applied Biosystems Inc). The delta delta-CT method was used to calculate fold differences between treatments, using GAPDH as a control house-keeping gene. RNA extracted from PBMCs treated with a buffer was used as a control, and the relative fold changes were determined in relation to control CT values.

Statistical analysis

Statistical significance was determined using one-way ANOVA followed by post hoc tests to identify specific group differences, with p-values indicated in the figure legends to denote significance levels.

Contribution of non-viral factors to HIV neuropathogenesis in vitro

To understand the contribution of non-viral immune factors to the pathogenesis underlying HAND, we investigated whether HIV-spent media immunodepleted for viral Tat and gp120 proteins could still retain neurotoxicity. PBMCs of healthy donors were infected with the neurovirulent HIV-1_ADA (5 ng/ml p24) or the non-neurovirulent HIV-1_Indie (10 ng/ml p24) strain to achieve comparable levels of viral infection five days after infection. Following viral infection, p24 levels in the spent media were quantitated using a commercial p24 assay. Spent media of HIV-1_ADA and HIV-1_Indie contained 25 ng/ml and 22 ng/ml of p24, respectively, confirming comparable levels of infection. Spent media were then incubated with VRC03 and E1.1 antibodies to deplete gp120 and Tat antigens from the samples, respectively. Subsequently, undifferentiated SHSY-5Y neuroblastoma cells were treated with immuno-depleted supernatants or an uninfected control medium. Following immuno-depletion of Tat and gp120 (the two primary viral factors responsible for neurotoxicity), the neurovirulent HIV-1_ADA isolate retained about half of its neurotoxicity compared to the non-neurovirulent HIV-1_IndieC1 isolate (Figure 1), that exhibited barely any change in its already low overall neurotoxicity (Figure 1 inset). This result pointed at host factors other than viral proteins, such as pro-inflammatory cytokines and chemokines, as the potential source of cytotoxicity of SHSY-5Y cells.

Figure 1:

Contribution of cytokines and chemokines to neurotoxicity: 300 µl of the spent media of HIV-1_ADA- or HIV-1_Indie-infected PBMCs were incubated with pooled VRC03 and E1.1 (anti-gp120 and anti-Tat, respectively) antibodies for 24 h to deplete both gp120 and Tat, via two rounds of immunodepletion. Neuronal SH-SY5Y cells were incubated with the spent media, with or without (inset) immuno-depletion. WST-1 cell viability was performed using the WST-1 assay (Inset-adapted from reference # [6]). This experiment was performed 3 times.

Differential induction of cytokines and chemokine by HIV-1_ADA and HIV-1_IndieC1

In order to examine the types of cytokines that were induced by the infection of subtypes B and C HIV-1, we next infected PBMCs with HIV-1_ADA or HIV-1_IndieC1. Upon infection of PBMCs, HIV-1_ADA induced significantly higher levels of pro-inflammatory cytokines IL-1α, IL-1β, IL-6 and TNF-α as measured via Luminex assay (Figure 2). These factors have been implicated [6], [11], [12], [13], [14], [15] in causing neuronal damage in various neuroimmune disorders. In contrast, HIV-1_IndieC1 induced a higher expression level of IL-8, which is an exception in not being a neurotoxic cytokine. Further, HIV-1_ADA also induced higher levels of CC family of chemokines, CCL2, CCL3, and CCL4, and also IP-10 (CXCL-10), a CXC family chemokine. Although the CC-chemokines are conditionally neuroprotective [16], 17], at higher levels, they can promote increased immune cell trafficking to the brain, leading to augmented neuronal damage. Furthermore, HIV-1_ADA infection of PBMCs also resulted in a more significant down-regulation of anti-inflammatory cytokines IL-4 and IL-10, compared to HIV-1_IndieC1-infection or control uninfected sample. Additionally, HIV-1_IndieC1 also suppressed the induction of IL-10 to a greater extent than that of HIV-1_ADA. These results suggest that the cytokines and chemokines are vital in causing neurotoxicity, which is subjective to subtype-specific differences.

Figure 2:

Differential induction of chemokines and cytokines upon HIV-1 infection of PBMCs. A panel of thirteen cytokines/chemokines (Interferon-α, CCL-2, CCL-3, CCL-4, IL-1α, IL-1β, IL-4, IL-6, IL-8, IL-10, TNF-α, IP-10, and RANTES) was measured in the spent media from HIV-1_ADA- or HIV-1_Indie-infected PBMCs, using a high-sensitivity human cytokine array kit. Each sample was assayed in triplicate wells, and data was acquired using a Luminex-100 system and analyzed using the Bio-Plex Manager software. This experiment was performed 3 times, each time with triplicates.

Tat plays a key role in the differential induction of chemokines and cytokines

It is well-known that HIV-1 Tat protein transactivates numerous cellular genes in addition to HIV-1 LTR. In fact, Tat from subtypes B and C are known to differentially transactivate cytokine genes in primary monocytes [18]. In order to test whether the differential neurotoxicities of spent media from the PBMCs infected with HIV-1_ADA and HIV-1_IndieC1 are likely a reflection of the types of cytokines induced by Tat protein in PBMCs, PBMCs were treated with recombinant Tat proteins from HIV-1_YU-2 and HIV-1_BL-43 strains representing subtypes B and C, respectively. After a 4-h treatment, RNA was extracted from PBMCs, and gene expression changes were compared using quantitative real-time PCR on a macroarray of selected cytokine/chemokine genes. The neurovirulent HIV-1_YU-2 Tat protein induced 5- to 30-fold increases in the level of pro-inflammatory cytokine transcripts compared to the control or the non-neurovirulent HIV-1_BL-43 Tat (Figure 3). Similar elevated patterns for subtype B Tat-treated PBMCs were observed with neurotoxic host factors, such as TNF-α and IP-10.

Figure 3:

Differential secretion of cytokines and chemokines by PBMC treated with Tat: PBMCs of healthy donors were treated with 200 ng/ml of recombinant Tat proteins of HIV-1_YU-2 (clade B) or HIV-1_BL-43 (clade C) viruses. Transcription of selected cytokine and chemokine genes was evaluated using real-time PCR. This experiment was performed two times with triplicates.

The Tat-induced patterns of the anti-inflammatory cytokines, IL-4 and IL-10, differed from those of the pro-inflammatory cytokines. Both Tat proteins exerted a suppressive effect on the expression of the two anti-inflammatory cytokines, compared to a ‘no Tat’ control. Importantly, HIV-1_BL-43 Tat exerted a weaker inhibitory effect on the secretion of these two anti-inflammatory cytokines than the HIV-1_YU-2 Tat, thus inducing still significantly higher levels of these host factors. Thus, compared to subtype B, subtype C Tat induced relatively low levels of pro-inflammatory cytokines and relatively higher levels of anti-inflammatory cytokines, resulting in an overall lower potential for neurotoxicity

A potential role of non-viral factors in neuroinflammation

The spent media of virus-infected PBMC retained neurotoxic properties even after depleting Tat and gp120, suggesting an important role for immune dysregulation in the causation of adverse CNS outcomes. Using the highly sensitive Luminex assay, several pro- and anti-inflammatory cytokines and chemokines could be measured in such media. Importantly, the neurovirulent subtype B strain up-regulated the cytokines higher than the non-neurovirulent subtype C strain. Consistent with these results, differential expression of the same cytokines was also shown in PBMCs treated with recombinant Tat proteins.

The analysis of chemokines and cytokines in Figures 2 and 3 highlights distinct inflammatory profiles associated with subtype B (HIV-1_ADA/HIV-1_YU2) and subtype C (HIV-1_IndieC1/HIV-1_BL43). In Figure 2, the depleted media from PBMCs infected with subtype B HIV-1_ADA shows significantly higher levels of pro-inflammatory chemokines (CCL2, CCL3, CCL4, and CXCL10/IP-10) and cytokines (IL-1α, IL-1β, IL-6, TNF-α) compared to subtype C HIV-1IndieC1, which induces a less pronounced inflammatory response. Subtype B also suppresses anti-inflammatory cytokines (IL-4 and IL-10) to a greater extent, shifting the balance toward a more pro-inflammatory state. The results in Figure 3, where recombinant Tat proteins were used, reveal subtype-specific effects of Tat proteins with HIV-1B Tat inducing significantly higher levels of pro-inflammatory markers, including CCL2, CCL3, CCL4, IL-1β, IL-6, and TNF-α, compared to subtype C Tat. Both Tat proteins suppress IL-4 and IL-10, but subtype C Tat exhibits a weaker suppressive effect, potentially altering the intensity of inflammation. The comparison underscores that subtype B HIV-1 strains (ADA/YU2) drive a more aggressive pro-inflammatory and neurotoxic response than subtype C strains (IndieC1/BL43), reflecting their differential contributions to neuropathogenesis.

Both purified Tat proteins and HIV-infection display subtype-specific differences

Our observations with HIV-infected PBMCs correlated well with the findings using recombinant purified Tat proteins added to PBMCs. Subtype C Tat protein induced low levels of pro-inflammatory cytokines than subtype B Tat protein and higher levels of anti-inflammatory cytokines. Furthermore, previously published studies [17], [18], [19] on IL-10 levels induced by the subtype B and C Tat proteins in monocytes have yielded inconsistent results suggesting the involvement of additional viral or host factors affecting IL-10 regulation. It is well known that Tat protein is secreted and is taken up by bystander cells [2]. In the uninfected bystander cells, HIV-1 Tat can transcriptionally transactivate proinflammatory cytokine genes. We previously showed that the differential uptake of Tat B and C proteins is likely the basis of differential transactivation of such inflammatory host genes and the resulting differential neuroinflammation [20]. These observations indicate that differences between neurovirulent and non-neurovirulent HIV-1 strains are reflected in their Tat protein, inducing the differential regulation of pro- and anti-inflammatory chemokines/cytokines in PBMCs.

Among the factors up-regulated, both IL-1α and IL-1β have been implicated in causing and enhancing neuronal damage [21]. These variant forms of Interleukin-1 act through the IL-1 receptor. Several models of brain injury demonstrated that IL-1 receptor agonist (IL-1Rα) reduces neuronal damage caused by IL-1α and IL-1β [11]. IL-6 has been implicated in enhancing NMDA-mediated neurotoxicity, leading to increased membrane depolarization [22]. In the CNS, TNF-α orchestrates a diverse array of functions, including but not limited to immune surveillance, cellular homeostasis, and protection against specific neurological insults. However, in the context of HIV-1 infection, TNF-α enhances HIV-1 Tat-induced neurotoxicity by oxidative stress leading to neuronal apoptosis [23]. TNF-α can also inhibit glutamate uptake by astrocytes, leading to an extracellular buildup of glutamate, causing neuronal excitotoxicity [24]. Furthermore, inhibiting TNF-α activity abrogates HIV-1-induced neurotoxicity in vitro [25]. IL-1β and TNF-α dysregulate glutamate production in neurons through the induction of glutaminase [15]. IP-10, while serving as a T cell chemoattractant, can cause neurotoxicity in primary neurons due to elevated intracellular calcium levels [12], leading to caspase activation and neuronal death [26].

We found that infection of PBMC by the neurovirulent HIV-1_ADA and treatment of PBMC by recombinant Tat of HIV-1_YU-2 strain, both up-regulated chemokines CCL-2, CCL-3, and CCL-4 to significant levels. CCL-2 (MCP-1), although typically neuroprotective [16], can cause monocyte migration to the CNS [27], thus exacerbating neuronal damage. Likewise, CCL-3 (MIP1α) and CCL-4 (MIP1β) also can enhance the migration of immune cells of diverse lineages to the CNS [28]. All the processes can contribute to increased chronic inflammation and progressive neuronal damage, finally manifesting as HAND. Neuronal injury is further augmented by the suppression of anti-inflammatory chemokines, IL-4 and IL-10, as shown in vitro in neuronal cultures and in vivo in patients. Despite viral suppression with cART, indirect neuronal damage is persistent and is mediated by upregulated inflammatory cytokines and chemokines, particularly in individuals with chronic HIV infection. Concomitant up-regulation of several pro-inflammatory cytokines, such as TNF-α and IL-1α, and down-regulation of anti-inflammatory cytokines, such as IL-4 and IL-10, have been documented in HIV-1 patients suffering from HAND [13].

Our findings suggest that in addition to using cART to control viral replication, adjunctive therapies incorporating neuroprotective agents, such as minocycline [29], 30] and memantine [31] might be beneficial in suppressing CNS inflammation. Alternative approaches to reducing CNS inflammation by targeting leukocyte trafficking across the blood-brain barrier with Natalizumab, an anti-alpha-4-integrin monoclonal antibody [32], 33] can potentially improve CNS outcomes in people with HIV. Additionally, novel therapies targeting HIV-1 Tat function [34] could reduce overall neuronal damage, both from the direct neurotoxicity of Tat and the up-regulation of pro-inflammatory cytokines and chemokines induced by Tat. Although substantial in vitro and in vivo evidence demonstrates attenuation of neuronal damage using adjunctive approaches, therapeutic interventions targeting CNS dysfunction in HIV-infected individuals remain clinically ineffective, necessitating further research to develop interventions that can improve patient outcomes and quality of life.