The Use of CBD and Its Synthetic Analog HU308 in HIV-1-Infected Myeloid Cells

2.2. CBD Decreases Viral RNA and Proinflammatory Cytokine mRNA

As discussed above, EVs released from infected cells contain viral RNA and proteins that promote pathogenesis. In our previous work, we showed that CBD decreases viral RNA and proteins found within EVs [4]. To determine the EV subpopulations impacted by CBD treatment, we analyzed the viral RNA in the 2K, 10K, 100K, 167K(S), 167K(L) subpopulations after treating the cells with CBD. Similar to EV concentration, CBD treatment caused a significant decrease in viral TAR and env RNA within the 100K subpopulation and a slight increase in the 167K(L) subpopulation (Figure 2A,B). As expected, there was an increase in the unprocessed viral spike protein gp160 with a reduction in the cleaved form of gp120 [4], which was found in the 167K(S) and 167K(L) subpopulations released from CBD-treated cells (Supplementary Figure S1).

Although viral products are known to be associated with neuronal damage, the major cause of neurotoxicity observed with HIV-1 infection is through the innate immune system, induced by the release of proinflammatory cytokines such as TNF-α and IL-1β [12]. CBD has been shown to decrease inflammation in a variety of diseases including neurological disorders, such as epilepsy and multiple sclerosis [41,48]. Specifically, CBD has been shown to attenuate the production of proinflammatory factors such as IL-1β and TNF-α to lessen the severity of neuronal inflammation within these disorders which can alleviate seizures and increase mobility, respectively [41,48].

Since HAND is characterized by the neuroinflammation, we assessed the level and impact of CBD on proinflammatory cytokines IL-1β and TNF-α through the levels of mRNA associated with EVs. EV associated mRNA has been shown to be translated by recipient cells, and there is further evidence that mRNA such as IL-1β can impose non-translational changes to the recipient cell as well (such as the upregulation of antiapoptotic gene expression in natural killer cells) [49,50,51,52]. Therefore, we attempted to determine which subpopulation(s) of EVs had proinflammatory cytokine mRNA and whether CBD treatment had any impact on the packaging of these populations. To ensure full length mRNA was analyzed, we synthesized the cDNA with an oligo(dT) primer (3′ extension) and investigated copy numbers with primers that were close to the cap (5′ end).

Results showed that both mRNAs were present in varying amounts in these fractions (Figure 2C,D). However, with CBD treatment, the decrease in proinflammatory cytokine mRNA (IL-1β and TNF-α) was not observed in the 100K subpopulation (compared to decrease viral products in the 100K fraction), and there was a decrease in the TNF-α mRNA levels within the 2K subpopulation (Figure 2C,D). This decrease may be from redistribution of cytokine mRNA between the EV subpopulations which showed altered levels in the 10K, 167K(L), and 167K(S), or the differences in EV subpopulations may be due to the increase in autophagy activation with CBD treatment shown previously [4] and its impact on EV regulation (Supplementary Figure S2). Taken together, CBD treatment shows a decrease in viral RNA transcripts TAR and env within the 100K subpopulation and decrease in proinflammatory cytokine mRNA TNF-α in the 2K subpopulation, likely indicating a package specific mechanism that is different for viral vs. cytokine RNA.

2.3. Analog of CBD (HU308) Similarly Reduces EV Subpopulations

The FDA has currently approved CBD in the medication Epidiolex^® for the treatment of seizures [53]. Epidiolex^® has been shown to be effective. However, CBD, with FDA approval, is isolated from the cannabis plant, and the purification process is expensive. It must go through validation and reproducibility in order to procure the correct dose at large quantities [54]. We therefore looked for a synthetic alternative with a structure similar to CBD’s that also had anti-inflammatory effects (i.e., IL-1β and TNF-α) and settled on HU308 [42]. HU308 is an analog to CBD and is a specific agonist for CB2 receptor [55], as opposed to CBD’s effect on GPR55, 5HT(1A), A2A, TRPV1-2 receptors, among others [56] (Figure 3A,B). CB2 is expressed mainly in the immune system [57], while the CBD receptors are within CNS and throughout the body (i.e., TRPV1 is found on neurons in epithelial cells [58,59]). These receptors are part of the G-coupled protein receptors and are pharmacologic targets of interest as they control inflammation as well as other sensations (i.e., pain) [37,60,61,62,63]. Thus, it is not surprising that HU308 has been shown to decrease peripheral and ocular inflammation [55,61].

Here, we aimed to investigate both CBD and HU308 treatment on HIV-1-infected cells, with HU308 showing no cell death at concentrations ranging from 0.1 to 100 µM during 3-day treatment of uninfected monocytes and T-cells (U937 and Jurkat, respectively) as well HIV-1-infected monocytes and T-cells (U1 and J1.1, respectively) (Supplementary Figure S3). Similar to CBD treatment, HU308 decreased EVs from HIV-1-infected cells released from the 100K but also within 167K(S) subpopulation (Figure 3C). HU308 treatment, however, also shows a slight decrease in EV size across subpopulations 10K, 100K, and 167K(L) while increasing 167K(S), indicating differences between CBD and HU308 treatments (Figure 3D,E). Collectively, these data indicate that HU308 may be a synthesized alternative to CBD in regulating EV release.

2.4. HU308 Reduces Both Viral and Proinflammatory RNAs

We next asked whether HU308 could regulate HIV-1 RNA levels as well as cytokine mRNA associated with the EVs released from HIV-1-infected cells. To determine the impacts of HU308 on HIV-1 infection and inflammation, a similar protocol was used as above. Similar to CBD, HU308 treatment reduced viral TAR and env RNA from the 100K EV subpopulation, as well as a slight decrease of TAR in the 167K(S) and env in the 167K(L) subpopulations (Figure 4A,B). Furthermore, we observed a decrease in TNF-α and IL-1β mRNA in the 2K EV subpopulation. This was seen with an increase in the 10K subpopulation (Figure 4B,C). Taken together, these data indicate that HU308 reduce viral RNA within the classical exosome subpopulation and proinflammatory cytokine mRNA (TNF-α and IL-1β) within the 2K subpopulation, similar to CBD, indicating a similar package specific response to the drug treatment.

2.5. Decrease in Viral RNA in Primary Macrophages Treated with HU308 or CBD

During the initial HIV-1 infection, there is an increase in mobility of activated peripheral blood monocytes to the brain, and it is the microglial (the brain’s resident macrophage population) in which the viral reservoir replicates and contributes to the neuroinflammation observed in HAND [2,64]. We next utilized primary macrophages for infection with dual tropic virus (89.6) and treatment with HU308. Previously, we had shown that CBD decreases viral RNA in primary cells. However, in this study, we aimed to explore the impact of HU308 as well as the combination of these cannabinoids with cART [4].

Here, peripheral blood mononuclear cells (PBMCs) were differentiated into macrophages, infected with HIV-1 dual tropic strain and then treated with CBD, HU308, cART, or combination every 3 days for 10 days. The cells were then collected on day 10, and the RNA was isolated and assessed for presence of viral TAR, TAR-gag, and env. HIV-1 non-coding TAR represents short transcripts; TAR-gag represents transcription into gag open reading frame; and env is representative of full-length transcription observed in infections. When cells were infected with HIV-1 89.6, there was an increase in all three transcripts after 10 days. Treatment with optimal concentration of cART (10 µM each) showed a significant decrease in all three RNA transcripts (Figure 5; lane 3). We then treated cells with lower levels of cART (1 µM of each anti-viral compound) and observed no drop in HIV-1 transcription. However, when a combination of low cART and CBD was added to cells, there was a significant decrease in all transcripts (lane 5) compared to CBD alone (lane 7). Finally, when low cART and HU308 were added to infected cells, there were no significant changes in RNA levels compared to HU308 alone (lane 6 vs. 8). Taken together, these data indicate that CBD along with low cART have a more pronounced effect on HIV-1 RNA levels in primary macrophage infection compared to HU308 treatment.

2.6. Effect of HU308 and CBD in a 3D Mini-Brain Infection Model

HAND is a comorbidity due to the low-level viral transcription and inflammation within the CNS. We therefore utilized a 3D infection model to score for the effects of HU308 and CBD. The 3D neurospheres used here have previously been shown to be composed of neuronal progenitor cells, neurons, astrocytes, and microglia-like cells and were shown to be a suitable model for HIV-1 infection and cART treatment [65]. In this study, neurospheres were pretreated with cART, CBD, or HU308 for 24 h prior to infection and then treated every 48 h for 7 days (Figure 6A). At 168 h post-infection, neurospheres were pelleted, and cells were collected for viral and cytokine RNA analysis. There was a decrease in intracellular HIV-1 RNA levels post cART (Figure 6B; lane 3) and a drop in viral RNA with CBD or HU308 alone (lane 4 and 5).

The neuronal death and injury associated with HAND is mainly caused by viral products and proinflammatory cytokines [2]. To better understand the potential decrease in proinflammatory cytokine RNA levels, infected neurospheres were treated with CBD and HU308 and assayed for viral RNA but also for proinflammatory cytokine mRNA in these cultures. We observed the presence of both cytokine RNAs in all infected neurospheres. As expected, with cART treatment, there was a decrease in IL-1β mRNA levels in these cultures (Figure 6C; lane 3). This trend was also seen post HU308 treatment (lane 5). Interestingly, there was an increase in IL-1β levels post-CBD treatments (lane 4). We also assayed for presence of TNF-α and again observed a similar trend in which there was a decrease in levels with cART or HU308 treatment and an increase with CBD (Figure 6D). Collectively, these data implicate that in a multi-cellular infection model (3D), CBD and HU308 can decrease viral RNA levels, and that HU308 can reduce IL-1β and TNF-α proinflammatory cytokine mRNA levels.

4.2. Primary Cells

A set of healthy, primary PBMCs was purchased (Precision For Medicine, Frederick, MD, USA) and cultured in vitro. The cells were placed in fresh complete RPMI overnight, and Phorbol 12-Myristate 13-Acetate (PMA; CAT: 16561-29-8, Cayman Chemical) and phytohaemagglutinin (PHA) were added the next morning. PMA/PHA was added every other day, and media were added throughout 10 days. On day 10, T-cells cells in suspension were removed and in vitro differentiated. Primary macrophages were allowed to grow with fresh media and PMA treatments for 8 more days. Macrophages were infected with HIV-1 89.6 strain (MOI:10). The next day, the infected cells were treated with cART at high levels (10 µM; Lamivudine, Tenofovir, Emtricitabine, Indinavir), low level cART (1 µM), and with or without CBD (5 µM) or HU308 (5 µM). This treatment was added every 3 days for a total of 3 treatments, and on the 2nd treatment, fresh media were added. Macrophages were collected on day 10 and processed for RNA analysis.

4.3. Serum EV-Depleted Medium

FBS was depleted of EVs as previously described [81], where FBS was at ultracentrifugation 100,000× g for 90 min in a Ti70 rotor (Beckman Coulter, Indianapolis, IN, USA). The EV-depleted supernatant was then carefully collected and stored at −20 °C for future use.

4.4. ZetaView Nanoparticle Tracking Analysis (NTA)

ZetaView Z-NTA (Particle Metrix, Inning, Ammersee, Germany) with software (ZetaView 8.04.02, Particle Metrix) was used to preform NTA. Machine calibration and user settings were previously described and performed in accordance to the manufacturer’s protocols [11,81]. The ZetaView software was used to analyze the mean size (diameter; nm) and the sample concentration (EV number) after the measurement from the 11 measured positions and removal of outliers. The measurement data provided from ZetaView were then inserted into Microsoft^® Excel^® to calculate averages and standard deviation from the technical triplicates.

4.5. EV Isolation via Differential Ultracentrifugation

HIV-1-infected monocytes (U1 cells) were grown in T75 flasks and expanded up to 25 mL of culture volume. The cells were treated with PBS or CBD (5 µM) daily for three days. The cells were then separated from the supernatant via low-speed centrifugation (300–400× g for 10 min). Supernatants were then transferred into ultracentrifugation tubes, and the tubes were placed in Ti70 (Beckman, Brea, CA, USA) rotor and spun at 2000× g (2K) at 4 °C for 45 min. Next, the supernatant was transferred to new tubes and further centrifuged at 10,000× g (10K) for 45 min, while the pellets were resuspended with PBS (150 µL per tube) and transferred to fresh 1.5 mL tubes. After the 10K spin, the supernatant was again separated into new tubes and spun at 100,000× g (100K) for 90 min, and the pellets were resuspended and transferred as described above. This process of supernatant removal and pellet resuspension was repeated, and the supernatant was spun down at 167,000× g for 4 h (h; 167K(S)), the pellet was collected, and the final spin was done at 167,000× g for 16 h (167K(L)). The final supernatant was discarded, and the 167K(L) pellet was resuspended in PBS. Isolated EVs were stored at −20 °C for future assays.

4.6. RNA Isolation, cDNA Synthesis, and Quantitative Real-Time PCR (RT-qPCR)

Total RNA isolation of cells and EVs were prepared as follows. Cells were harvested, washed once in 1× PBS free of calcium and magnesium, and resuspended in 50 µL of 1× PBS free of calcium and magnesium. EVs were isolated via differential ultracentrifugation as described above. Trizol Reagent (Invitrogen, Carlsbad, CA, USA) was used to isolate total RNA from cell pellets and isolated EVs according to the manufacturer’s protocol. GoScript Reverse Transcription Systems (Promega, Madison, WI, USA) was used to generate cDNA with Envelope Reverse: (5′-TGG GAT AAG GGT CTG AAA CG-3′), TAR Reverse: (5′-CAA CAG ACG GGC ACA CAC TAC-3′), and oligo(dT) reverse primer (Promega) with a GoScript kit (Promega). These were compared to quantitative standards that were created from serial dilutions of a CEM T-cell line that contained a single copy of HIV-1 LAV provirus per cell (8E5cells).

For viral RNA, cDNA samples (2 µL) or standards (2 µL) were plated in a master mix (18 µL) composed of IQ Supermix (Bio-Rad, Hercules, CA, USA), TAR Forward Primer (5′-GGT CTC TCT GGT TAG ACC AGA TCT G-3′), TAR Reverse Primer (5′-CAA CAG ACG GGC ACA CAC TAC-3′), or (viral env forward primer (5′-GGC AAG TCT GTG GAA TTG G-3′) and env reverse (5′-TGG GAT AAG GGT CTG AAA CG-3′) were used to measure full length viral genome) and TAR Probe (5′56-FAM-AG CCT CAA TAA AGC TTG CCT TGA GTG CTT C-36-TAMSp-3′). The BioRad CFX96 Real Time System was used for RT-qPCR with the following conditions: a single 2 min cycle at 95 °C followed by 41 cycles with 15 s at 95 °C and 40 s at 60 °C.

For all other RNA, cDNA samples or standards (2 µL) were plated with a master mix (18 µL) of SYBR Green (Bio-Rad) that included corresponding primers below and set to the following cycles.

TNF-α: forward primer (5′-CCC GAG TGA CAA GCC TGT AG-3′), reverse primer (5′-GAT GGC AGA GAG GAG GTT GAC-3′). The conditions were a single 50 °C 2 min, a 95 °C cycle for 2 min, then 41 cycles of 95 °C for 15 s, 57.3 °C for 30 s, and 72 °C for 40 s.

IL-1β: forward primer (5′-AGA TGA TAA GCC CAC TCT ACA G-3′), reverse primer (5′-ACA TTC AGC ACA GGA CTC TC-3′). Which were later verified with primers closer to the 5′ end: IL-1β forward (5′-TCA GCC AAT CTT CAT TGC TC-3′ and reverse (5′-AAC AAG TCA TCC TTG CC-3′). The BioRad CFX96 Real Time System conditions were set at 50 °C for 2 min, 95 °C for 2 min, and a 41 cycle of 95 °C for 15 s, 54 °C for 40 s, and 72 °C for 40 s.

IL-10: forward (5′-GCA AAA CCA AAC CAC AAG ACA GA-3′) and reverse (5′-TCT CGA AGC ATG TTA GGC AGG-3′). The conditions were as follows, the initial 50 °C for 2 min, then a single 95 °C cycle for 2 min, 41 cycles of 95 °C for 15 s, 59.0 °C for 30 s, and 72 °C for 40 s.

Cellular GAPDH: forward primer (5′-GAA GGT GAA GGT CGG AGT CAA C-3′), reverse primer (5′-CAG AGT TAA AAG CAG CCC TGG T-3′), and conditions were set at 50 °C for 2 min, 95 °C for 2 min, 41 cycles of 95 °C for 15 s and 57.5 °C for 30 s.

Each reaction was performed in triplicate, and quantification was determined through the comparison of cycle threshold (Ct) values to the 8E5 standard curve in the BioRad CFX Manager Software. The analysis of generated raw data was performed in Microsoft^® Excel^®.

4.7. Preparation of Whole Cell Extracts

HIV-1-infected monocytes (U1) cell pellets were collected, washed with 1× PBS, and resuspended in 50 µL of lysis buffer [50 mM Tris-HCl (pH 7.5), 120 mM NaCl, 5 mM EDTA, 0.5% Nonidet P-40, 50 mM NaF, 0.2 mM Na₃VO₄, 1 mM DTT, and 1 complete protease inhibitor cocktail tablet/50 mL (Roche Applied Science, Penzberg, Germany)]. The resuspended cells were then incubated for 20 min on ice with vortexing every 4 min. After the 20 min incubation, the cell debris were centrifuged at 10,000× g for 10 min. A Bradford protein assay was then performed in accordance of the manufacturer’s protocol (Bio-Rad) to determine protein concentrations of each cell lysate.

4.8. Western Blot

Laemmli buffer was added to isolated EV pellets or to sample cell lysates (15 µg) and was then heated for 3 min at 95 °C. Next, 10 µL of cell lysate sample and EV sample were then loaded onto a 4–20% Tris/glycine gel (Invitrogen). The gels were run at 100 V and then transferred onto Immobilon PVDF membranes (MilliporeSigma, Burlington, MA, USA) at 50 mA overnight. The membranes were blocked for 2 h at 4 °C with 5% milk in PBS with 0.1% Tween-20 (PBS-T). The gels were then incubated in PBS-T with the appropriate primary antibody overnight at 4 °C: α-gp120 (CAT: 522; NIH AIDS Reagent Program), α-GAPDH (CAT: 47724; Santa Cruz Biotechnology, Paso Robles, CA, USA), α-CD63 (CAT: EXOAB-CD63A-1; Systems Biosciences, Palo Alto, CA, USA), α-catalase (CAT: 724810; Novus Biologicals, Centennial, CO, USA), α-Phosphor-Cofilin (Ser 3) (CAT: 3311; Cell Signaling Technology), α-Rab7 (B-3) (CAT: 376362; Santa Cruz Biotechnology), α-HSP90 α/β (F-8) (CAT: 13119; Santa Cruz Biotechnology), α-Hsc70 HRP (B-6) (CAT: 7298; Santa Cruz Biotechnology), α-coronin-1A (CAT: NB110-58867; Novus Biologicals). Membranes were washed with PBS-T and then incubated at 4 °C for 2 h with the indicated HRP-conjugated secondary antibody. Clarity Western ECL Substrate (Bio-Rad) was used with the Molecular Imager ChemiDoc Touch System (Bio-Rad) to view HRP luminescence.

4.9. Cell Viability Assay

Cells (uninfected monocytes (U937), HIV-1-infected monocytes (U1), uninfected T-cells (Jurkat), HIV-1-infected T-cells (J1.1)) cultured in RPMI with 10% exosome-free FBS were plated at 5 × 10⁴ cells per well in 96-well plates. Cells were treated with HU308 at varying concentrations (0.1, 1, 5, 10, 100 µM) every 24 h for 3 days. A Cell Titer-Glo assay was used to measure cellular ATP activity according to the manufacturer’s protocol (Reaction Biology, Malvern, PA, USA). Briefly, 100 µL of Cell Titer-Glo was added to each well and then mixed for 2 min (in dark conditions) and left in room temperature undisturbed for 10 min. The readings were taken with GloMax Explorer (Promega) and analyzed in Microsoft^® Excel^®.

4.10. Neurosphere 3D Model

Using STEMdiff™ Neural Progenitor Medium (STEMCELL Technologies™; Vancouver, BC, Canada), Neural Progenitor Cells (NPCs; CAT: ACS-5003; ATCC. Manassas, VA, USA) were grown and used to generate 3D neurospheres (1 × 10⁵ cells per well) as described before [65]. Briefly, 3D neurospheres were differentiated with DMEM:F12 (CAT: 30-2006; ATCC) with Neural Progenitor Cell Dopaminergic Differentiation Kit (CAT: ACS-3004; ATCC) and incubated for 14 days. Neurospheres were infected with 10 µL of Infectin™ (Virongy, LLC; Manassas, VA, USA) and HIV-1 89.6 dual-tropic strain (MOI:10) and were incubated for 48 h in a total volume of 200 µL. Neurospheres were treated with 1× PBS ± 10 µM cART cocktail (Lamivudine, Tenofovir, Emtricitabine, Indinavir) ± CBD (10 µM), ±HU308 (10 µM) every other day for 7 days. At day 7, RNA was isolated from neurospheres for downstream RT-qPCR analysis.

4.11. Humanized Mice

Since there is considerable interest in the use of selective CB₂ receptor agonists (i.e., HU308, which are devoid of psychoactive properties of CB₁ agonists), we utilized humanized mice to test its effect on HIV-1 replication. All mice used in this study were maintained within the National Center for Biodefense and Infectious Disease’s breeding colony (George Mason University, Manassas, VA, USA). All experiments were carried out in bio-safety level 3 (BSL-3) facilities and in accordance with the Guide for the Care and Use of Laboratory Animals (Committee on Care and Use of Laboratory Animals of The Institute of Laboratory Animal Resources, National Research Council, National Institutes of Health Publication 86-23, revised 1996). NOD.Cg-Rag1^tm1Mom Il2rg^tm1Wjl/SzJ mice were obtained from The Jackson Laboratory (007799, Bar Harbor, ME, USA) and were humanized [66]. Briefly, five groups of three neonatal animals were sub-lethally irradiated. The next day, mice were intraperitoneally injected (100 µL) with ~5 × 10⁵ human cord blood-derived CD34⁺ hematopoietic stem cells (Lonza, Walkersville, MD, USA). At ~3 months post-engraftment, animals were subcutaneously infected with the dual-tropic HIV-1 89.6 (100 µL, 20 ng of p24/µL). Mice receiving cART (emtricitabine 210 mg/kg, tenofovir 20 mg/kg, ritonavir 60 mg/kg; maraviroc 60 mg/kg), HU308 (1 mg/kg) [67,68], or a combination of HU308 and cART were injected intraperitoneally three times a week prior to blood isolation, and PCR. HU308 was obtained from and dissolved in anhydrous ethanol at 20 mg/mL prior to preparation in 1:1:18 solution of HU308: Kolliphor EL: and saline stored at 4 °C.

We thank all members of the Kashanchi lab for their assistance, especially Gwen Cox, Vivian Gao, and Emre Nalbantoglu.