Background
The advent of antiretroviral therapy (ART) has increased the lifespans of people living with HIV (PLWH) [
8]. However, a significantly high percentage of PLWH remain comorbid with drug abuse, (cocaine and marijuana [
9‐
11]) leading to rapid disease progression [
12‐
17], immune activation, or protection from immune activation in cases of dual drug use (marijuana and cocaine) [
9]. Although HIV-induced neuroinflammation may drive HIV-associated neurocognitive disorder (HAND), the mechanisms by which cannabinoids, such as THC, a component of marijuana modulates HIV-induced neuroinflammation are not completely understood.
Investigating the longitudinal effects of HIV and THC on the brain is difficult to conduct in humans, because humans are polydrug users, cigarette smokers, and alcohol consumers. Humans also have variability in patterns and lengths of drug, alcohol, and cigarette usage. These extrinsic factors make studies with human subjects’ complex, and data associated with such studies correlational. Animal models, such as the SIV-infected rhesus macaque (SIV/RM) model [
18,
19] of HIV provides a valuable animal model/approach and more controlled environment to study HIV-induced neuroinflammation, the response to long-term treatment with THC, and the effects of HIV/SIV alone or SIV and THC on EV cargo composition and function.
Previously, we showed that the anti-inflammatory effects of long-term low dose THC was associated with its ability to stimulate the release of bioactive blood-derived extracellular (BEVs) [
18] that induced divergent actin cytoskeletal and signaling cues in SIV-infected RMs [
18]. Whether or not THC is capable of reversing neuroinflammatory effects of HIV/SIV is unknown. In addition to the effects of HIV proteins and RNA on persistent inflammation/immune activation, EVs play multifaceted roles in the function/dysfunction of HIV target cells, such as monocytes, microglial, astroglial and non-target cells, including neurons. Broadly defined, EVs released by many cell types [
20‐
32] encompassing exosomes, microvesicles, and apoptotic bodies carry bioinformation (proteins, DNA, diverse RNA profiles) and regulate intercellular/organ communications [
24,
29,
31]. EVs carry markers of the producer cells and if the producer cells are healthy or diseased, EVs will carry markers corresponding to their state [
33,
34]. Thus, EVs are important in research [
26,
33,
35‐
37] and drug discovery/therapeutics [
38].
The cargo of EVs, including RNA mediate dynamic intercellular crosstalk. EV-associated RNA (exRNA) consists of diverse RNA biotypes that are incorporated into or associated with various circulating carriers, including lipoproteins [
39,
40], EVs [
41,
42] and extracellular or membraneless condensates (MCs) [
42]. EV-associated exRNA include several classes of long and small RNAs but not limited to miRNAs [
32]. Through association with EVs, miRNAs are protected from degradation by RNAses [
43].
EVs traverse the blood brain barrier (BBB) and have been shown to regulate the availability of neuroprotective factors [
4]. Emerging evidence indicate that brain-derived EVs are linked to neurogenesis, neural development, synaptic communication, nerve regeneration, and neuroinflammation [
44‐
47]. EVs are used to establish molecular signatures associated with drug abuse in HIV-infected individuals [
4,
18,
48,
49], and as such, may serve as intercellular conveyors of bioactive molecules within the CNS. Indeed, brain-derived EVs have been purified from cultured neurons, oligodendrocytes, microglia, astrocytes, and cerebrospinal fluid (CSF) [
50‐
52]. However, there are no available studies describing the properties and functions of EVs from the basal ganglia (BG), a series of interconnected subcortical nuclei and a major target/reservoir of HIV in the CNS [
5,
6]. BG dysfunction is a hallmark of HIV infection and cognitive impairment in PLWH with neuronal death in the BG [
53]. Moreover, HIV targets the BG leading to loss of dopaminergic neurons [
54]. A prior study showed an increase in miR-29b in the BG of morphine-dependent SIV-infected RMs compared with controls [
4].
In this study, we demonstrate that BG contains bioactive EVs. A significant number of miRNAs are significantly decreased in the EVs isolated from BG of SIV-infected RMs. We further demonstrate that low-dose chronic THC treatment counteracts the suppressive effects of SIV on BG-derived EV miRNA repertoire, and conversely restores the levels of all of the suppressed miRNAs. Furthermore, BG-EVs internalized by astrocytes alter astrocyte activation and gene expression profiles in an EV and CX3CR1 dependent manner.
Methods
Macaques and viruses
A total of nine age and weight-matched Mamu-A0*1
−/B08
−/B17
− specific-pathogen-free (free of SIV, D retrovirus, STLV and Herpes B) male Indian rhesus macaques were randomly assigned to three experimental groups. One group (
n = 3; Group 1) received twice daily injections of vehicle (VEH/SIV) (1:1:18 of emulphor:alcohol:saline) and second (
n = 3; Group 2) received twice-daily injections of Δ
9-THC (THC/SIV) beginning 4 weeks prior to SIV infection until 6 month post-SIV infection [
55]. Group 3 (
n = 3) macaques served as uninfected controls (Table
1). THC (NIDA/NIH) was prepared as an emulsion using alcohol, emulphor, and saline (1:1:18) as vehicle before use. Chronic administration of VEH (Group 1) or Δ
9-THC (Group 2) was initiated 4 weeks before SIV infection at 0.18 mg/kg as used in previous studies [
7,
55‐
57]. This dose of Δ
9-THC was found to eliminate responding in a complex operant behavioral task in almost all animals [
57]. Groups 1 and 2 macaques were infected intravenously with 100 TCID
50 dose of the CCR5 tropic SIVmac251. Beginning the day of SIV infection, the THC dose was increased for each subject to 0.32 mg/kg, over a period of approximately 2 weeks when responding was no longer affected by 0.18 mg/kg on a daily basis (i.e., tolerance developed), and maintained for the duration of the study. The optimization of the THC dosing in RMs accounts for the development of tolerance during the initial period of administration. Because in previously published studies [
7,
57] this dose of THC showed protection, the same dose was used in this study. At necropsy, BG tissues were collected in RNAlater (Thermo Fisher Scientific) and Z-fix for total RNA extraction and embedding in paraffin blocks. SIV levels in plasma and BG were quantified using the TaqMan One-Step Real-time RT-qPCR assay that targeted the LTR gene [
55,
56].
Table 1
Animal IDs, SIV inoculum, duration of infection, viral loads, and brain histopathology in vehicle or delta-9-tetrahydrocannabinol (Δ9-THC) treated chronic SIV-infected and uninfected rhesus macaques
Chronic SIV-Infected and Vehicle treated (Group 1) |
IV95 | SIVmac251 | 180 | 0.02 | 2.0 | ND | ND |
JD66 | SIVmac251 | 180 | 0.04 | 0.2 | ND | ND |
JR36 | SIVmac251 | 180 | 0.5 | 0.2 | ND | ND |
Chronic SIV-Infected and Δ9-THC treated (Group 2) |
JI45 | SIVmac251 | 180 | 3 | 0.01 | ND | ND |
JC85 | SIVmac251 | 180 | 0.02 | 0.09 | ND | ND |
IV90 | SIVmac251 | 180 | 0.02 | 0.06 | ND | ND |
Uninfected Controls (Group 3) |
IR97 | NA | NA | NA | NA | NA | NA |
IT18 | NA | NA | NA | NA | NA | NA |
GT18 | NA | NA | NA | NA | NA | NA |
BG-EV purification and characterization
The schematic and workflow for isolation of basal ganglia EVs is shown in Additional file
1: Fig. S1. Briefly, small chunks of RNALater-stored BG tissues, ranging from 35 to 118 mg, were finely chopped and digested with collagenase III. Samples were clarified and supernatant was purified on a 20 × 0.5 cm Sephadex G-50 size exclusion column, using a particle purification liquid chromatography (PPLC) system as previously described [
42]. Fifty fractions of 200 µL were collected, and the 3D UV–Vis (230–650 nm) fractionation profiles were recorded. A no-tissue collagenase control was used as background. After background subtraction and PPLC analysis for particle size and concentration, EV-containing fractions were pooled and stored in small aliquots at − 80 °C. For further characterization, EVs were diluted in 0.1X PBS (1/1000). Zeta potential (ζ-potential) measurements were acquired using nanoparticle tracking analysis (ZetaView) as described previously [
58].
Energy dispersive X-ray transmission electron microscope with immunogold-labelling (TEM–EDX-IL)
Equal volumes of BG-EVs from each group were pooled (n = 4). 10 µL were spotted onto TEM grids. Specimens were incubated with anti-CD9 at 4 °C overnight. Following washing, samples were incubated with 10 nm gold-conjugated anti-mouse IgG for 1 h, washed, and then followed by a post-stain with uranyl acetate (1%). Specimens were characterized using TEM.
BG-EV RNA isolation
2 mL of PPLC-purified BG-EVs (equivalent of 9.6 × 1011 to 5.7 × 1012 particles or 254–984 µg of EV proteins) were concentrated under reduced pressure at low temperature, and the total RNA was isolated using miRNeasy serum/plasma kit, per manufacturer’s protocol. RNA was eluted and the eluate was measured using a NanoDrop 1000.
Small RNA-Seq
Libraries were prepared using 25 ng of RNA and 20 cycles of PCR following the manufacturer’s recommendations. The libraries were pooled to equal nanomolarity concentrations and then purified and size selected using Pippin Prep (Sage Biosciences, Beverly, MA, USA). The library pool was profiled using a TapeStation (Agilent Technologies, USA) and Qubit (ThermoFisher) before sequencing on the NextSeq 550 (Illumina, San Diego, CA, USA). Sequencing was performed with single 75 bp reads.
sRNA-Seq data were processed for filtering, trimming, and QC analyses before generating count matrices. After adapters are trimmed, reads are filtered based on length (5 bp, 15 bp). Filtering reads shorter than 5 bp determines RNA degradation. 15 bp is the minimum length for meaningful alignments. Count matrices were obtained using trimmed reads (minimum length 15) by alignment to the
Macaca mulatta genome and the
Macaca mulatta data set miRBase (miRNA database). Raw miR Counts are provided in Additional file
2: Table S1.
Mouse model
All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC; covered by Animal welfare assurance No A3011-0), SUNY Stony Brook, School of Medicine and conducted in accordance with National Institutes of Health “Guide for the Care and Use of Laboratory Animals” guidelines. Experiments were performed using adult (1–3 days) male mice [C57BL/6 J (wt) and CX3CR1-GFP (Jackson Labs, 005582 model B6.129P-Cx3cr1tm1Litt/J)9]. The mouse lines were backcrossed onto a C57BL/6 J background, bred in-house, and genotyped by PCR. The brains from these mouse strains were used for preparation of primary astrocyte cultures.
Generation of primary cortical astrocytes
Primary astrocytes were isolated from P1–P3 mouse pups [
59,
60]. Briefly, T-75 flasks were coated with Poly-
d-Lysine for 1 h at 37 °C. Flasks were washed, filled with DMEM 10% FBS and placed at 37 °C, 5% CO
2. Brains were dissected, the cerebellum removed, and the Cortices placed on ice in 1X HBSS to slow metabolic function. Four cortices were then placed in one mL of papain solution for 15 min at 37 °C for chemical dissociation, with a brief mechanical dissociation through a p1000 pipette tip. After additional incubation at 37 °C for 15 min, p1000 pipette dissociation, a p200 pipette was used to further digest the tissue into a homogenous single-cell suspension. Papain activity was neutralized using 20 mL of DMEM 10% FBS media and the solution spun at 3000 RPM for 10 min. The pellet was resuspended in 1 mL DMEM 10% FBS, placed into T-75 flask containing 37 °C media. On days 3, 6, and 9 of culture, a portion of 10 mL of media were removed and replaced with equivalent volume of fresh media. On day 10 of culture, microglia were detached and removed. On day 12 of culture, OPCs were detached (in a 16–18 h shake at 300 RPM) and removed. After removal of the OPCs, the remaining cells are astrocytes and were trypsinized and plated in a 10 cm dish.
BG-EV internalization
PBS control or BG-EVs were stained with SYTO™ RNASelect™ Green Fluorescent cell stain and purified using Exosome Spin Columns. Labelled BG-EVs were added to cells for kinetic imaging using Lionheart FX automated scope.
Immunofluoresence of activation markers
10,000 astrocytes were treated with PBS (vehicle) or 100 µg/mL of pooled BG-EVs (n = 4, 25 µg/sample) for 24 h. Cells were imaged immediately after treatment and at 24 h post-treatment. Subsequently, cells were fixed, permeabilized, incubated with anti-GFAP, and detected with appropriate fluorescently conjugated secondary antibody. Cells were finally stained with DAPI and imaged again using a confocal microscope (Leica Sp8-x) or an automated scope (Lionheart FX). Briefly, nine fields of views per well and three wells per condition were recorded. Images were then pre-processed, deconvoluted, and stitched, and cells were identified using DAPI channel. A secondary mask was then applied in which Texas red MFI (representing GFAP) was calculated. For the WT: PBS: 1062 cells, Control EVs: 972 cells, VEH/SIV EVs: 817 cells, THC/SIV EVs: 1039 cells. For CX3CR1+/GFP: PBS: 591 cells, Control EVs: 714 cells, VEH/SIV EVs: 1001 cells, THC/SIV EVs: 828 cells.
Real-time quantitative PCR (RT-qPCR)
250,000 astrocytes plated overnight were treated with PBS or with 100 µg/mL of pooled BG-EVs (n = 4, 25 µg/sample) for 24 h. RNA was extracted from cells and used for cDNA synthesis. RT-qPCR was performed using a 7500 FAST machine and Power Track SYBR Green master mix.
For validation of the let-7 family of miRNA, we used the Thermo Fisher Scientific mml-let-7a-5p (Assay ID: 000377) and mml-let-7c-5p (Assay ID: 000379) TaqMan PCR specific assays, following manufacturer’s instructions.
Cell viability analysis
Percentage of viable cells was inferred by quantitation of cellular ATP, measured using CellTiter-Glo Luminescent viability assay [
61,
62].
Statistical analysis
Differential expression data were generated using Graphpad Prism. The significance cutoff was set to fold change (FC) > 1.5 or < − 1.5 and a p value < 0.05. Ordinary one-way ANOVA multiple comparison test or two-way ANOVA test with (Dunnett’s or Tukey’s corrections were used to assess statistical differences. When stated, unpaired T test with Welch’s correction was also used. Details of specific statistics are presented in each figure legend, where the p values are listed.
Discussion
HIV invades the brain within 2 weeks after infection [
74,
75], infecting resident CNS cells, including microglia and astrocytes [
76,
77]. Glial cells maintain brain homeostasis [
78] via scavenging for excess toxic neurotransmitters, maintaining BBB integrity, regulation of immune activation/inflammation, and release of neurotrophic factors. Moreover, activation of glial cells contributes to neuropathology induced by mitochondrial toxins [
79]. HIV and its proteins (Tat and gp120) have been implicated in mediating astrocyte toxicity [
80,
81]. Although astrocytes make up about 40% of the total CNS cell population, their exact function in HIV-induced neuroinflammation remains unclear.
Our data using the SIV-infected RM model exposed to chronic THC treatment, along with WT and CX3CR1
+/GFP astrocytes, revealed that BG contains EVs with previously unrecognized functions. There was no significant difference in BG-EV size distributions, concentrations, protein content, and zeta potential (Fig.
1). Similarly, we detected no difference in the structure of EVs (Fig.
2) from BG samples of uninfected control, VEH/SIV, and THC/SIV RMs. However, significant differences were observed in the BG-EV miRNA profile (Fig.
3). The cargo composition and functions of the BG-EVs were modulated by HIV/SIV infection. BG-EVs from these animals exhibited a proinflammatory profile and induced an activated/proinflammatory state in primary CNS astrocytes. Strikingly, long-term low dose THC treatment of SIV-infected RMs selectively counteracted the generalized proinflammatory nature of SIV BG-EVs. Specifically, both mml-let-7a-5p and mml-let-7c-5p were significantly lower in BG-EVs from VEH/SIV RMs than the levels in uninfected controls and THC/SIV-BG-EVs (Fig.
3F). Interestingly, the let-7 family members which have identical seed sequence [
82] are abundantly expressed in the brain, glial progenitor cells, astrocytes [
83] and they exhibit high cross-species sequence conservation [
84‐
86]. The let-7 family of miRNAs is involved in regulating CNS inflammation and neurological outcomes. Studies have linked let-7a-5p and let-7c-5p overexpression to the suppression of TNFα expression [
87,
88], while let-7c-5p improved neurological outcomes in a murine model of traumatic brain injury by suppressing neuroinflammation [
89]. With respect to HIV, Swaminathan et al., showed significant down-regulation of Let-7 family of miRNAs in patients with chronic HIV infection compared to healthy controls [
90], while Zhang et al. showed that HIV infection resulted in suppression of the let-7i/IL-2 axis leading to cell death [
91]. Aside from HIV, let-7d-5p, let-7a, let-7c, and miR-122-5p decreased over time in agreement with the progression of liver fibrosis in hepatitis C-infected people [
92]. In the context of EVs, let-7a has been shown to regulate EV secretion and mitochondrial oxidative phosphorylation [
93]. Furthermore, EV-associated let-7a-5p and let-7c-5p levels were significantly reduced in liver cirrhosis patients and let-7a-5p levels significantly correlated with hepatic fibrosis markers and could predict hepatic cirrhosis more accurately than other markers of hepatic fibrosis [
92].
It is evident that EVs can be used as biomarkers for specific conditions, as suggested for neuron-derived EVs from PLWH bearing biomarkers of cognitive impairment [
94]. EVs are also prospective carriers of drugs and other exogenous compounds with the potential to regulate neuropathogenesis [
95]. Using the SIV-infected RM model, we detected several miRNAs that regulate inflammation/immune regulation, TLR signaling, Neurotrophin TRK receptor signaling, cell death/response to stress. Neurotrophins, including brain-derived neurotrophic factors (BDNF) are a family of closely related proteins identified to control many aspects of survival, development, and functions of neurons. Continued presence of neurotrophins is essential as it controls synaptic function and plasticity, sustain neuronal survival, morphology, and differentiation, in addition to other roles outside the nervous system [
96]. Furthermore, HIV suppresses BDNF expression and reduces BDNF activity, resulting in neurodegeneration in infected individuals [
97,
98]. Remarkably, long term low dose THC administration led to significant upregulation of all the SIV-downregulated miRNAs. THC also upregulated 26 miRNAs and some neuromodulatory miRNAs, including the let-7 family members [
64] that regulate biological processes, such as apoptosis [
99,
100], immune system modulation [
101,
102], TLR7 activation [
103], axon guidance [
104], and BBB permeability [
105]. In addition, let-7c has been shown to promote polarization of macrophages from M1 to M2 phenotype [
102]. These neuromodulatory miRNAs were either downregulated or were unchanged in the VEH/SIV group. The various pathways altered by SIV or THC have been implicated in neuroinflammation. These observations indicate that the changes SIV and/or THC imprinted on the brain, manifest in BG-EVs, which may then serve as a conduit for dissemination of miRNAs to CNS cells. It is also possible that BG-EVs may spread their miRNA cargos to distant sites in the periphery via cell-to-cell transfer.
Aside from serving as biomarkers, EVs mediate intercellular communication, both within and across species. The cross-species efficacy of EVs and their cargo has been established by our group and others. For example, human semen-derived EVs delivered human Apobec3g and Apobec3f gene products to mice in vivo [
106]. Moreover, EVs derived from human BMD2a cells were incorporated within mice brains and the EVs mediated permeability of mouse brain blood vessels [
107]. If EVs mediate intercellular communication via their cargo, and EVs from one species can function in another species, cross-species transfer of miRNA may be likely, especially since miRNAs are conserved throughout bilaterian evolution [
108]. In our study, mouse astrocytes tolerated up to 100 μg of rhesus macaque BG-EVs. However, increasing BG-EV concentration beyond 100 µg showed that the tolerance of astrocytes was dependent on the origin (uninfected control, VEH/SIV, THC/SIV) and concentration (50, 100, 200 µg) of the EVs. The viability of astrocytes remained unchanged and comparable to PBS treated cells and tolerated up to 200 µg EV concentration. However, viability of astrocytes significantly declined in the presence of 200 µg of VEH/SIV EVs. THC/SIV EVs prevented VEH/SIV EV-mediated decline in astrocyte viability.
In the steady state, astrocytes express GFAP. However, the expression of GFAP is increased during activation [
109]. In our study, VEH/SIV EVs significantly increased the level of astrocyte GFAP (Fig.
5). The conversion of GFAP
low astrocytes into GFAP
high astrocytes which occurred via interaction with EVs is indicative of their potential to activate these cells. Activated astrocytes, and other brain resident cells are key contributors to HIV-induced neuroinflammation. These cells release neurotoxic factors and inflammatory mediators such as TNFα [
110] that may lead to deleterious consequences, including neurotoxicity. TNFα is mitogenic to astrocytes and increased levels of TNFα are associated with elevated GFAP expression [
111]. Astrocytes promote chronic inflammation and progressive neurodegeneration via overexpression of TNFα [
112]. Increased production of TNFα is linked to HIV-induced immunological abnormalities [
113‐
115] and astrocyte apoptosis, a major feature of cellular injury in HAND [
116]. In the CNS, CD40 is expressed by astrocytes and its interaction with CD40L on other resident CNS cells such as T, monocytic, natural killer, and mast cells mediates intracellular signaling events that promote the production of proinflammatory factors and neurotoxins [
117]. The reprogramming of astrocytes from CD40/TNFα
low to CD40/TNFα
high by EVs is suggestive of the proinflammatory nature of VEH/SIV BG-EVs. Similarly, the conversion of CD40/TNFα
high to CD40/TNFα
intermediate by THC/SIV EVs is suggestive of the potential anti-inflammatory nature of THC/SIV EVs. These observations are significant, because the hallmarks of HAND include widespread microglial activation, accompanied by reactive astrogliosis and their secretory products, including cytokines and chemokines [
118,
119].
In addition to TNFα, other inflammatory mediators were induced by BG-EVs. MMP2 and MMP9 mRNA were variably altered in astrocytes by all EVs in a CX3CR1-dependent manner. In some cases, THC/SIV EVs counteracted SIV EV-induced effects on MMP2 and MMP9 mRNA. These observations are intriguing, because under pathological conditions, dysregulated expression of MMPs induces inflammation and promotes progression of neurodegenerative diseases [
120]. MMP2 and MMP9 are ECM-degrading enzymes involved in inflammation and tissue remodeling. Through their induction of soluble TNFα and proteolytic activity on the ECM, MMPs may promote brain injury. It is likely that astrocyte activation by EVs may trigger the expression of inflammatory mediators (TNFα, MMP2, and MMP9) as observed in the present study. Furthermore, MMPs are thought to be involved in the pathogenesis of HAND and other neurodegenerative disorders via degradation of ECM and compromising the BBB [
121‐
123]. In addition, MMP2 and MMP9 are present in the CSF, plasma, and brain tissue of HIV patients [
122]. The suppressive effect of THC/SIV EVs on inflammatory mediators show the potential of THC/SIV EVs to ameliorate the effect of pathological VEH/SIV EVs.
With regard to the role of CX3CR1 in HIV infection, the ligand of CX3CR1, CX3CL1, also known as Fractalkine (FKN) is increased in the CSF of HIV-infected individuals who exhibit neurocognitive impairment [
124,
125]. However, exogenous FKN has been shown to protect cultured neurons from neurotoxicity induced by Tat or Tat + morphine-induced dendritic losses [
126,
127]. It is worth mentioning that CX3CR1 isoforms produced by alternative splicing may function as fusion co-receptors for HIV envelope protein [
128], although the significance of CX3CR1 among other HIV co-receptors for HIV entry is still not clear. Nonetheless, HIV-infected individuals homozygous for CX3CR1-I249 M280 (that affects two amino acids—isoleucine-249 and methionine-280) exhibit a more rapid progression to AIDS [
129], perhaps due to reduced FKN binding. Our findings, together with the literature evidence, suggest a possible involvement of CX3CR1 mediated response to BG-EV alteration of astrocyte gene expression and function.
The ability of low dose THC, which is also prescribed [FDA-approved synthetic THC (Marinol)] as an appetite stimulant in PLWH [
130‐
134] to reprogram BG-EVs and affect their functions is significant. Chronic cannabis use may slow disease progression, prolong survival, reduce viral load, and attenuate infection-induced inflammation/immune activation in SIV-infected RMs [
7,
55‐
57,
135‐
138] and ART-treated PLWH [
139,
140]. The effect of THC is systemic―affecting many organs. As a result, THC and other cannabinoids are recommended for the treatment of digestive disorders [
141‐
146] and FDA approved their use for clinical management of wasting and appetite stimulation, in PLWH [
130,
131,
133,
134,
147].
It is also evident in the comparative analysis of plasma and BG viral loads (Table
1), that BG viral loads were generally lower in the THC/SIV group compared to the VEH/SIV group, although not statistically significant. The lack of significance in viral load between the two groups may be due to the limited number (
n = 3/group) of study subjects used. Thus, studies with increased sample size are warranted to assess whether THC can reduce CNS viral load.