Supplementing HIV-ART with cannabinoids increases serotonin, BHB, and Ahr signaling while reducing secondary bile acids and acylcholines
Southwest National Primate Research Center, Texas Biomedical Research Institute, San Antonio, TX 78227, USA.
Abstract
Despite effective antiretroviral therapy (ART), people with HIV (PWH) experience persistent inflammation and metabolic dysfunction, increasing their risk for non-AIDS comorbidities. Accordingly, we evaluated the effects of long-term/low-dose Δ9-tetrahydrocannabinol (THC) supplementation in simian immunodeficiency virus (SIV)–infected, ART-treated rhesus macaques (RMs). THC significantly increased plasma/jejunum serotonin and indole-3-propionate, enhancing gut-brain communication through up-regulation of serotonin receptors (HTR4/HTR7) and aryl hydrocarbon receptor (Ahr) signaling via a cannabinoid receptor (CBR)-2–mediated mechanism. Furthermore, THC enriched cholesterol-metabolizing Oscillibacter and reduced plasma cholesterol and toxic secondary bile acids (SBAs), thus improving cholesterol and SBA homeostasis. Furthermore, THC increased β-hydroxybutyrate (BHB) levels via a CBR1-mediated mechanism, suggesting enhanced hepatic fatty acid oxidation for metabolic and cardiovascular health. THC restored ART/SIV-induced elevation of pro-inflammatory and cardiotoxic long-chain acylcholines to preinfection levels. THC-treated RMs maintained viral suppression despite reduced plasma ART levels, suggesting diminished ART-related toxicity. Our findings demonstrate phytocannabinoids to be a safe adjunct therapy alongside ART to mitigate chronic inflammation and metabolic dysfunction in PWH.
Teaser
Cannabinoids added to ART may reduce comorbidities by increasing serotonin, IPA, and BHB while reducing SBAs and acylcholines.
Article notes
Untitled section
Received 2025 Jan 31; Accepted 2025 Aug 1; Collection date 2025 Sep 5.
INTRODUCTION
Advancements in combination antiretroviral therapy (ART) have greatly improved the life expectancy of people with HIV (PWH), transforming HIV from a typically fatal condition into a manageable chronic disease with proper access to treatment (1). As PWH are living longer, age-related comorbidities such as HIV-associated neurocognitive disorders and renal disease (2–4) and those associated with metabolic syndrome like type 2 diabetes, hypertension, dyslipidemia, cardiovascular disease (CVD) (4), nonalcoholic fatty liver disease (NAFLD), and cholestatic liver disease (CLD) (1, 5) have been increasingly observed and reported to occur at least a decade earlier than in HIV-uninfected individuals (3). Both residual inflammation that persists in PWH despite viral suppression by ART and age-related chronic inflammation are increasingly recognized as critical contributors to HIV-associated non-AIDS comorbidities (1, 4). Furthermore, despite their improved safety profile, current generation of ART drugs continue to cause adverse effects such as renal, hepatic, metabolic, and cardiovascular side effects, gastrointestinal (GI) symptoms and dysbiosis, weight gain, dyslipidemia, hypersensitivity, and neuropsychiatric manifestations (6). Thus, developing feasible and effective interventions to reduce residual inflammation in PWH on ART remains an unmet clinical need.
After entry, HIV primarily targets and rapidly infects the gut-associated lymphoid tissue, causing significant depletion of CD4+ T cells within 2 weeks. The CD4+ T cell destruction and ensuing inflammation cause significant structural and functional damage to the intestine that is not reversed by ART. The resultant intestinal epithelial barrier disruption, dysbiosis, and systemic translocation of microbial products have been proposed to fuel chronic immune activation and disease progression. In the central nervous system (CNS), HIV-induced inflammation disrupts the integrity of the blood-brain barrier (BBB), thereby facilitating the entry of not only HIV-infected CD4+ T cells and other inflammatory cells but also bacteria and pro-inflammatory microbial products like lipopolysaccharide (LPS) into the brain, leading to infection, neuroinflammation, and dysregulation of the microbiota-gut-brain axis (MGBA) (7, 8). Other contributors, like ART-associated adverse events, loss of regulatory T cell function, and the activation of opportunistic pathogens like CMV (cytomegalovirus), further perpetuate the persistent inflammatory state that can drive the development of HIV-associated non-AIDS comorbidities (1, 4).
The gut microbiome, with its large collection of microorganisms and their genes, performs a wide range of metabolic functions that affect the homeostasis of multiple organs, including the brain, liver, kidney, and heart. Several beneficial effects of the gut microbiota are mediated by an assembly of metabolites that includes short-chain fatty acids and tryptophan metabolites, the latter serving as one of the most important endogenous ligands of the aryl hydrocarbon receptor (Ahr) (9, 10). Dysbiosis alters the bacterial communities that participate in nutrient breakdown to produce bioactive metabolites that positively regulate signaling pathways affecting the brain, liver, and cardiovascular health. Accordingly, correlated gut microbiome and metabolome alterations, especially gut microbiome–derived secondary bile acids (SBAs), can lead to the development and progression of cholestasis, which can further alter the gut microbiota, worsen systemic inflammation, and exacerbate CLD, leading to cirrhosis and end-stage liver disease, an important comorbidity affecting PWH, in this interactive cycle (5). Although the factors driving chronic inflammation in PWH are multifaceted, developing strategies to target multiple factors rather than approaching them singularly would be advantageous. Accordingly, synchronized modulation of tryptophan metabolism [serotonin and its cognate receptors and indole-3-propionate (IPA)], the microbiome, and the Ahr represents an attractive therapeutic strategy to leverage the intricate interplay between these systems to treat HIV/simian immunodeficiency virus (SIV)–induced intestinal inflammation, metabolic dysfunction, and dysregulation of the MGBA (8, 10). Specifically, we hypothesized that by inhibiting the function of indoleamine-2,3-dioxygenase (IDO1), as we had shown previously (11), it is possible to channel more tryptophan away from the kynurenine pathway and increase its availability to produce serotonin and IPA while concurrently enhancing the growth of beneficial gut microbes involved in these pathways and optimizing Ahr activation.
We previously showed that phytocannabinoids can effectively reduce inflammation in the intestine and brain through reduction of T cell activation and pro-inflammatory gene expression in ART-naïve and ART-experienced SIV-infected rhesus macaques (RMs) (8, 12, 13) and, most recently, via epigenetic modulation of the NLRP6 inflammasome sensor (14). Here, using the same cohort of ART-experienced RMs, we investigated the impact of HIV/SIV infection on the MGBA and microbiota-gut-liver axis (MGLA) under suppressive ART. We showed that persistent intestinal dysbiosis (elevated Gammaproteobacteria and Enterobacteriaceae) is accompanied by reduced expression of the Ahr in the jejunal epithelium. Long-term, low-dose Δ9-tetrahydrocannabinol (THC) administered as an adjunct to ART reduced gut dysbiosis and significantly increased the relative abundance of IPA-producing Clostridia (15), serotonin production–stimulating Lactobacillus plantarum (16), and cholesterol-metabolizing Oscillibacter (17). As a result, THC significantly elevated plasma IPA levels and serotonin levels in both plasma and the intestinal epithelium. Furthermore, SIV/ART-induced dysbiosis disrupted bile acid homeostasis characterized by significantly elevated plasma hydrophobic toxic SBA levels. THC not only maintained SBA homeostasis but also significantly increased plasma levels of the antioxidant, anti-inflammatory, and neuroprotective β-hydroxybutyrate (BHB), which is also a surrogate marker of hepatic fatty acid oxidation (18). In addition, combining ART with low-dose THC decreased plasma cholesterol levels and maintained healthy preinfection plasma levels of acylcholines, an important finding that suggests cardioprotective effects. Overall, our findings provide valuable and actionable insights and mechanisms into the regulatory role of cannabinoids on the MGBA and MGLA that may benefit not only PWH but also people affected by chronic inflammation–driven diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), including those suffering from NAFLD and CLD.
RESULTS
Administration of THC to SIV-infected RMs on ART resulted in a global metabolic profile that exhibited the least separation from preinfection samples
The study design and the timeline are shown in Fig. 1A. VEH/SIV/ART (n = 8) and THC/SIV/ART (n = 8) RMs had undetectable viral loads in both plasma and the jejunum at 5 months postinfection (MPI) (Fig. 1B and Table 1), confirming successful viral suppression by ART. To determine whether low-dose THC administration modulated the plasma metabolome profiles in SIV-infected ART-treated RMs, we first compared the plasma metabolome profiles of THC/SIV/ART and VEH/SIV/ART RMs at 5 MPI versus their respective preinfection samples. It is clear from the principal components analysis (PCA) plot (Fig. 1C) that the global metabolic signature of the VEH/SIV/ART group (green symbols) showed better separation from their preinfection samples (light gray symbols). In contrast, the THC/SIV/ART group (pink symbols) showed the least separation from their preinfection samples (dark gray symbols) (fig. S1A). A higher number of metabolites were markedly increased in the THC/SIV/ART group (n = 106) than the VEH/SIV/ART group (n = 86) compared to their respective preinfection controls. In contrast, a higher number of metabolites decreased in the VEH/SIV/ART group (n = 152) than in the THC/SIV/ART group (n = 126) relative to their preinfection controls (Fig. 1D). These changes were observed across nine metabolic superpathways (amino acids, peptides, carbohydrates, energy, lipids, nucleotides, cofactors and vitamins, xenobiotics, and uncharacterized molecules).
| Animal ID | SIV inoculum | Duration of infection (months) | Plasma viral loads 106/ml | Jejunum viral loads 106/mg RNA | Jejunum histopathology | Opportunistic infections | Age (years) | Weight at necropsy (kg) |
|---|---|---|---|---|---|---|---|---|
| Chronic SIV-Infected, ART and Vehicle Treated (Group 1) | ||||||||
| VEHSIVART1*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.65 | 6.35 |
| VEHSIVART2*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.56 | 7.07 |
| VEHSIVART3*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.61 | 6.60 |
| VEHSIVART4*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.48 | 5.92 |
| VEHSIVART5*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 3.65 | 4.91 |
| VEHSIVART6*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 5.71 | 7.59 |
| VEHSIVART7*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.66 | 6.59 |
| VEHSIVART8*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 3.60 | 5.58 |
| Chronic SIV-Infected, ART andΔ 9 -THC Treated (Group 2) | ||||||||
| THCSIVART1*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.68 | 5.37 |
| THCSIVART2*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.64 | 6.69 |
| THCSIVART3*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.64 | 5.64 |
| THCSIVART4*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 5.30 | 6.52 |
| THCSIVART5*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 3.64 | 5.22 |
| THCSIVART6*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.70 | 7.23 |
| THCSIVART7*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.58 | 6.42 |
| THCSIVART8*†‡ | SIVmac251 | 5 | ND | ND | ND | ND | 4.50 | 5.78 |
| Uninfected Controls (Group 3) | ||||||||
| CONT1§ | NA | NA | NA | NA | NA | NA | 7.86 | 7.16 |
| CONT2§ | NA | NA | NA | NA | NA | NA | 5.78 | 3.85 |
| CONT3§ | NA | NA | NA | NA | NA | NA | 5.60 | 4.14 |
The detection of THC (P = 0.0005) (Fig. 1E), THC carboxylic acid (P = 0.0012) (Fig. 1F), and THC carboxylic acid glucuronide (P < 0.005) (Fig. 1G) only in THC/SIV/ART RMs confirmed the activity of THC and its metabolites.
Tryptophan metabolism in the gut plays a pivotal role in host physiology, particularly in gut-brain axis communication. Because 90% of dietary tryptophan is converted to kynurenine via the kynurenine pathway, 7% to indoles, and 3% to serotonin (Fig. 1H), we hypothesized that inhibition of IDO1 by cannabinoids, as shown by us previously (11), would increase tryptophan availability for conversion to serotonin and IPA. Therefore, we focused on tryptophan metabolites, their receptor-mediated signaling partners, and their interactions with cannabinoids in the context of ART, given their critical implications for gut-brain axis signaling modulation and potential neuroinflammatory outcomes in PWH. Figure 1I shows the complete list of tryptophan metabolites that were significantly increased (red), significantly decreased (dark blue), increased with near statistical significance (pink), and decreased with near statistical significance (light blue) in the plasma of THC/SIV/ART relative to preinfection and VEH/SIV/ART (Fig. 1I).
Cannabinoids beneficially modulated tryptophan metabolism by significantly increasing serotonin and IPA levels
Figure 1I and the metabolic pathway maps (Fig. 2, A and B) show the complete list of tryptophan metabolites detected in the current study and those that showed significant increase (red circles) or decrease (dark blue circles) in VEH/SIV/ART (Fig. 2A) and THC/SIV/ART (Fig. 2B) relative to their preinfection controls. Metabolites in Figs. 1I and 2 (A and B) shown in light blue (decreased) and light pink (increased) showed a tendency to reach statistical significance (0.05 > P < 0.1). Plasma levels of serotonin were significantly increased in THC/SIV/ART RMs compared to their respective preinfection time point (P = 0.002). Because serotonin levels in the preinfection samples from the THC/SIV/ART group were considerably low, we performed a second analysis by combining the preinfection samples from both groups and detected significantly elevated serotonin levels only in the THC/SIV/ART group at 5 MPI (Fig. 2, B to D). Similarly, plasma levels of IPA (P = 0.0418) (Fig. 2, B and E) were also significantly increased in THC/SIV/ART RMs at 5 MPI compared to preinfection controls, whereas IPA levels, although elevated in VEH/SIV/ART RMs, did not reach statistical significance (P = 0.058) (Fig. 2, A and E) compared to their respective preinfection samples and combined preinfection samples (Fig. 2F). In addition, plasma kynurenine levels were also significantly high in THC/SIV/ART RMs (Figs. 1I and 2B, and fig. S1B). Nevertheless, kynurenine/tryptophan ratios, a better indicator of IDO1 functional activity, showed no change in both groups compared to their respective preinfection samples (fig. S1C). To further explore the role of serotonin, we investigated its production in the intestine by performing immunofluorescence staining of jejunum tissue from both THC/SIV/ART and VEH/SIV/ART groups at 5 MPI and their preinfection time points. We did not detect a significant change in the number of serotonin-producing enterochromaffin cells (ECCs) in the VEH/SIV/ART (Fig. 2H) group compared to preinfection samples (Fig. 2G). However, serotonin production (as revealed by the area of serotonin positivity) was significantly increased in ECCs of the THC/SIV/ART group (Fig. 2, J to L) at 5 MPI compared to preinfection (Fig. 2I) samples and the VEH/SIV/ART (Fig. 2H) group. Note that the number of serotonin-positive ECCs was significantly higher in the THC/SIV/ART group compared to the combined preinfection (Fig. 2M) samples from both groups and when analyzed separately (Fig. 2N), suggesting a significant THC-mediated enhancement of serotonin synthesis in the gut.
Phytocannabinoids increase physical signaling along the MGBA by up-regulating serotonin receptor expression
Serotonin plays a significant role in maintaining intestinal homeostasis, with five of the seven serotonin receptor subtypes expressed in the GI tract (19), each with distinct regulatory functions. Mining our recently published colon epithelium (CE) microarray data from a non-ART animal cohort detected significant up-regulation of the serotonin receptors, HTR4 and HTR7, in THC/SIV RMs relative to controls (13). Most notably, serotonin acting via HTR4 and HTR7 reduced inflammation (20) in mouse models of colitis, underscoring their potential as therapeutic targets (19). Compared to HTR4 (2-fold) (fig. S1D), THC increased HTR7 by 10-fold in the CE of THC/SIV RMs (fig. S1E).
Accordingly, we performed immunofluorescence staining to evaluate HTR4 and HTR7 protein expression in the jejunum and detected a significant increase in HTR4 expression at 5 MPI (Fig. 3, B, D, and I) compared to their preinfection (Fig. 3, A, C, and I) time points. Notably, THC/SIV/ART (Fig. 3, D and I) RMs showed a significantly greater increase in HTR4 protein expression compared to both their preinfection levels and VEH/SIV/ART RMs. This finding is important as HTR4 agonists have been proposed for treating inflammatory bowel disease (IBD), and the THC-mediated increase in HTR4 suggests a strong synergistic effect when combined with ART.
Similarly, HTR7 protein expression was significantly elevated in the jejunum of VEH/SIV/ART RMs at 5 MPI (Fig. 3, F and J) compared to their preinfection time point. The combination of ART and THC in THC/SIV/ART animals further increased HTR7 protein expression compared to both VEH/SIV/ART RMs and preinfection samples (Fig. 3, F, G, H, and J), as confirmed by quantification of fluorescence intensity. HTR7, like HTR4, is known for its anti-inflammatory effects in the intestine, further reinforcing the translational significance of these findings (20, 21).
To determine whether THC directly modulated HTR4 and HTR7 protein expression, we cultured and treated human small intestinal epithelial (SIE) cells with increasing concentrations of THC (0.5, 1.0, 1.5, 2.0, and 3.0 μM) for 24 hours and performed immunofluorescence (Fig. 3K). Quantification of immunofluorescence signal intensity demonstrated a concentration-dependent increase in HTR4 expression, with the highest signal intensity observed at 3 μM THC concentration (Fig. 3, K and L). Similarly, HTR7 protein expression also showed a similar trend with the highest protein up-regulation detected at 3 μM (Fig. 3, K and M), confirming the dose-dependent effect of THC on the protein expression of both serotonin receptors. Nevertheless, unlike HTR4, even 1 μM THC induced a substantial increase in HTR7 protein expression (Fig. 3, K and M), which aligns well with the in vivo fold change (FC) detected in CE (fig. S1D). To investigate the receptor-mediated mechanisms by which THC increased HTR4 and HTR7 expression, we pretreated SIE cells with cannabinoid receptor (CBR) antagonists, AM251 (CBR1 antagonist) and AM630 (CBR2 antagonist), followed by THC treatment and immunofluorescence quantification. Our findings revealed a significant decrease in both HTR4 and HTR7 expression in the SIE cells treated with AM630 (CBR2 antagonist) (Fig. 3, N to P), indicating that the THC-induced up-regulation of both serotonin receptors is mediated via CBR2.
To confirm the completeness of the serotonin signaling circuitry along the gut-brain axis, we next evaluated HTR4 and HTR7 protein expression in the myenteric plexus, a network of neuronal cell bodies (ganglia) located between the smooth muscular layers of the GI tract that is innervated by the vagus nerve. Immunofluorescence staining revealed strong expressions of HTR4 and HTR7 in the myenteric plexus (green in Fig. 3, Q and R). The presence of neuronal cell bodies in the myenteric plexus was confirmed by staining with the neuronal marker PGP 9.5 or UCHL1 (ubiquitin C-terminal hydrolase L1) (fig. S1F). These findings indicate a physical mechanism by which cannabinoids can enhance signaling along the gut-brain axis, wherein gut ECC-derived serotonin can directly bind HTR4 and HTR7 on the myenteric plexus to relay signals to the brain via the vagus nerve (Fig. 3S) (22).
Phytocannabinoids enhance chemical signaling along the MGBA by increasing IPA-Ahr interactions
In Fig. 2 (A and B), we demonstrated that, although ART treatment alone did not significantly increase neuroprotective plasma IPA levels, supplementing ART with THC significantly elevated IPA levels during SIV infection. Intestinal microbiota-derived IPA can traverse the intestinal epithelium into the lamina propria and, via the circulation, reach the brain, where it can exert neuroprotective effects (10). This pathway underscores the chemical connection between the gut and brain, highlighting the role of IPA in modulating the gut-brain axis.
To further investigate the mechanisms underlying this observation, we examined Ahr protein expression, which is known to regulate inflammation and a receptor for IPA (23), in the jejunum epithelium of VEH/SIV/ART and THC/SIV/ART RMs. Ahr protein expression was significantly higher in the jejunum epithelium of THC/SIV/ART (Fig. 4, D and E) compared to VEH/SIV/ART RMs (Fig. 4, B and E) and preinfection samples (Fig. 4, A, C, and E). To determine whether THC directly induced Ahr protein expression, we treated SIE cells with increasing concentrations of THC (0.5, 1.0, 1.5, 2.0, and 3.0 μM) and quantified its protein expression 24 hours later. THC induced Ahr protein expression in a dose-dependent manner, with the highest expression seen at 3.0 μM THC (Fig. 4, F and G).
Because IPA crosses the BBB, we determined whether receptors capable of binding IPA and mediating signal transduction were expressed in the brain. Given Ahr’s binding affinity for IPA and its role in regulating neuroinflammation (24), we next examined its expression in the brain, focusing on basal ganglia sections from both treatment groups and a separate cohort of uninfected controls. Ahr protein expression (red) localized predominantly to NeuN (green)–positive neurons and was significantly higher in the THC/SIV/ART (Fig. 4, H, bottom, and I) compared to both VEH/SIV/ART (Fig. 4, H, middle, and I) and preinfection controls (Fig. 4, H, top, and I), suggesting that THC-mediated up-regulation of Ahr in the brain may contribute to its neuroprotective effects.
To further explore the effects of THC on neuronal cells, we treated HCN-2 neuronal cells with increasing concentrations of THC (0, 0.5, 1.0, 1.5, 2.0, and 3.0 μM). We observed a dose-dependent increase in Ahr expression, with the strongest Ahr protein induction detected with 2.0 μM THC (Fig. 4, J and K). Our data suggest that increased plasma IPA levels, facilitated by THC supplementation, have the potential to not only cross the BBB but also activate neuronal Ahr receptors to exert neuroprotective effects. This represents a critical chemical mechanism by which cannabinoids enhance signaling along the gut-brain axis (Fig. 4L). The combined effects of elevated IPA and enhanced Ahr expression in the brain could be more effective in mitigating neuroinflammation (8) and promoting neuronal health, particularly in PWH on suppressive ART (24).
Cannabinoid supplementation during ART shifted the gut microbiome, decreasing Gammaproteobacteria while increasing Clostridia and L. plantarum
Given that IPA and serotonin are also produced or stimulated by gut bacteria such as Clostridia and Lactobacillus species, respectively, we next hypothesized that cannabinoids may reduce inflammation-driven gut dysbiosis (12) and better preserve and support the growth of commensal anaerobic bacteria capable of producing beneficial metabolites. Furthermore, the significantly high plasma levels of 3-hydroxyhippurate (fig. S1G) in THC/SIV/ART RMs provided indirect evidence of Clostridial enrichment as elevated hippurate levels have been previously associated with increased abundance of OTUs within the bacterial order Clostridiales (25). Accordingly, we performed shotgun metagenomic sequencing of jejunum contents to obtain deeper insights into the impact of long-term cannabinoid administration on the jejunum microbiome.
At the level of bacterial phylum, Actinobacteria (blue arrow in Fig. 5, A and G), known to exert pivotal roles in intestinal homeostasis (26), was significantly enriched (P = 0.03) in jejunum contents of THC/SIV/ART relative to preinfection controls and VEH/SIV/ART RMs (P = 0.01). Although statistically nonsignificant, levels of Gammaproteobacteria (red arrow in Fig. 5, B and H), a class of dysbiotic bacteria known to translocate into the systemic circulation and induce systemic immune activation in PWH (27) and SIV-infected RMs (28), were comparable between THC/SIV/ART and preinfection controls (Fig. 5H) but markedly enriched and elevated in VEH/SIV/ART RMs. Bacteria belonging to the class Clostridia known to produce the entero-, neuro-, and cardioprotective IPA (15) were substantially enriched at 5 MPI exclusively in THC-supplemented SIV-infected RMs (green arrow in Fig. 5, B and I).
When looking at the bacterial order, the relative abundance of pathogenic Bacteroidales (blue arrow in Fig. 5, C and J) and Enterobacterales (red arrow in Fig. 5, C and K) were significantly decreased (P = 0.026 and P < 0.02, respectively) in THC/SIV/ART compared to preinfection controls.
At the family level, the relative abundance of Enterobacteriaceae (blue arrow in Fig. 5, D and L) (P = 0.02, respectively) was significantly reduced in the jejunum contents of THC/SIV/ART RMs relative to preinfection levels. In contrast, the relative abundance of Clostridiaceae (P = 0.05) (green arrow in Fig. 5, D and M) showed significantly increased abundance in THC/SIV/ART RMs compared to preinfection controls. At the level of bacterial genera, Prevotella (blue arrow in Fig. 5, E and N), which has been previously linked to intestinal epithelial barrier disruption (29), showed significantly reduced (P < 0.001) abundance in THC/SIV/ART RMs compared to controls. At the species level, serotonin-producing L. plantarum (blue arrow in Fig. 5, F and O) was significantly enriched (P = 0.04) in the jejunum contents of THC/SIV/ART RMs relative to preinfection controls. Intrigued by the recent findings on the identification and characterization of cholesterol-metabolizing bacteria (Oscillibacter) in humans (17), we next looked to see whether these bacteria showed differential abundance in our treatment groups. Unexpectedly, Oscillibacter was present at very low to undetectable levels in jejunum contents (fig. S1H). However, after mining our recently published colonic microbiome data (8), we detected significantly high relative abundance of Oscillibacter in ART-naïve THC/SIV RMs that was significantly higher than ART-naive VEH/SIV and control RMs (Fig. 5P).
Overall, THC enriched the relative abundance of the commensal Actinobacteria, Clostridia, and Lactobacillus while reducing those of the pathogenic Enterobacteriaceae (Gammaproteobacteria) and Prevotella, confirming that THC maintains the gut microbiome homeostasis during HIV/SIV infection. THC increased the relative abundance of the serotonin production–stimulating L. plantarum and cholesterol-metabolizing Oscillibacter.
Metabolic alterations related to CLD are more significant in SIV-infected RMs on ART
PWH have been reported to have elevated cholesterol and triglyceride levels that have been linked to HIV infection and ART medications (30). Although ART has slowed HIV disease progression, prolonged use of certain widely used ART drugs is closely related to alterations in lipid metabolism resulting in increased cholesterol and triglyceride (hypertriglyceridemia) levels, which increases the risk of CVD in PWH (30, 31). Furthermore, because Oscillibacter showed high relative abundance in THC/SIV RMs, we hypothesized that plasma cholesterol levels would be lower in this group. Although unchanged in VEH/SIV/ART RMs (Fig. 6, A and C), plasma cholesterol levels were significantly decreased in THC/SIV/ART RMs (P = 0.004) (Fig. 6, B and C) compared to preinfection controls (Fig. 6, B and C). Although the mechanisms by which THC lowered plasma cholesterol may be multifactorial, we detected a statistically nonsignificant (P = 0.061) increase in beta-sitosterol, one of the several phytosterols (plant sterols) that was elevated only in THC/SIV/ART RMs compared to preinfection levels (Fig. 6B and fig. S1I). Phytosterols have received extensive attention owing to their excellent cholesterol-lowering activity (32) and their contribution to CVD prevention. Therefore, our data suggest that low-dose THC may decrease total cholesterol levels by increasing not only beta-sitosterol levels but also the relative abundance of the cholesterol-metabolizing Oscillibacter.
In the liver, cholesterol is converted into primary bile acids (PBAs), chenodeoxycholic acid (CDCA) and cholic acid (CA) (Fig. 6D). These PBAs are transported to the intestines, where the gut microbiota converts them into SBAs, namely, lithocholic acid (LCA), deoxycholic acid (DCA), and ursodeoxycholic acid (UDCA) (Fig. 6, D and E) (33). These SBAs, although cytotoxic at high concentrations due to their hydrophobic nature, serve as key regulators of PBA synthesis in the liver through a feedback mechanism (Fig. 6D). A major portion of the SBAs is reabsorbed in the ileum and transported back to the liver through the portal circulation (Fig. 6D), whereas the rest undergoes sulfation or glucuronidation to become hydrophilic, less toxic forms that are excreted through the urine or feces. In the liver, SBAs may remain in their unconjugated form or become conjugated with taurine or glycine, forming taurolithocholic acid or glycolithocholic acid, all of which are cytotoxic at elevated levels (Fig. 6E) (5, 33). Hepatic bile acid synthesis is tightly regulated through a negative feedback mechanism to maintain a constant bile acid pool. Disruption of the bile acid–gut–microbiota cross-talk can lead to significant alterations in bile acid homeostasis, affecting the liver and gut health (33–35). This intricate interplay is depicted in Fig. 6D.
The pathway figure (Fig. 6E) highlights the involvement of the gut-microbiota in the conversion of PBAs into SBAs, detoxification by sulfation, and clearly shows elevated levels of LCA, taurolithocholate, and glycolithocholate only in VEH/SIV/ART (red solid circles in Fig. 6E) but not in THC/SIV/ART RMs (fig. S1J) relative to controls. Elevated hepatic levels of highly hydrophobic SBAs LCA and DCA can directly activate the hepatic expression of pro-inflammatory cytokines, chemokines, and cell adhesion molecules (35). Hepatic bile acid synthesis is tightly regulated by bile acids through negative feedback mechanisms, which maintains a constant bile acid pool size. The disturbance in the bile acid–intestinal–microbiota cross-talk leads to the development and progression of cholestasis, which can further alter gut microbiota and exacerbate cholestasis in this interactive cycle. In this context and most notably, VEH/SIV/ART RMs showed significantly elevated (Fig. 6F, red circles) levels of 7 microbial-derived SBAs of the 12 total SBAs detected, including cholestatic/pro-inflammatory LCA or lithocholate (P = 0.002) (Fig. 6H), taurolithocholate (P = 0.0032) (Fig. 6I), glycolithocholate (P = 0.03) (Fig. 6J), lithocholic acid sulfate (P = 0.006) (Fig. 6K), taurolithocholate-3-sulfate (P = 0.02) (Fig. 6L), lithocholate sulfate (P = 0.0005) (Fig. 6M), and lithocholic acid glucuronide (P = 0.03) (Fig. 6N) compared to their preinfection time point. Conversely, THC/SIV/ART RMs showed significantly elevated levels of only two SBAs—lithocholate sulfate (P = 0.01) (Fig. 6, G and M) and lithocholic acid glucuronide (P = 0.039) (Fig. 6, G and N)—compared to preinfection controls. These findings demonstrate the ability of phytocannabinoids (THC) to maintain normal plasma SBA levels in SIV-infected RMs despite long-term daily ART treatment. The highly hydrophobic and potent cholestasis-inducing SBA salts lithocholate, taurolithocholate, and glycolithocholate were maintained at healthy preinfection levels in THC/SIV/ART RMs, whereas they were significantly elevated in the plasma of VEH/SIV/ART RMs. Only lithocholate sulfate and lithocholic acid glucuronide, which are water-soluble urinary excretory forms of toxic lithocholate, were significantly elevated in THC/SIV/ART RMs, and the FCs were higher than in VEH/SIV/ART RMs versus controls. Overall, the supplementation of ART with phytocannabinoids facilitates the conversion of lithocholate to its less toxic water-soluble forms for easier excretion via urine and bile, thereby reducing the levels of toxic SBAs in the circulation and potentially liver. Moreover, as bile acids are synthesized from cholesterol in the liver, decreased cholesterol levels in THC/SIV/ART RMs (Fig. 6C) may partly explain the decreased SBA levels.
Supplementing ART with cannabinoids increased BHB and maintained homeostatic acylcholine levels
BHB, a ketone body and a by-product of hepatic fatty acid oxidation, plays a critical role as an anti-inflammatory and neuro-, hepato-, and cardioprotective metabolite (36). To assess whether THC enhanced fatty acid oxidation during HIV/SIV infection under ART, we compared plasma BHB levels between THC/SIV/ART and VEH/SIV/ART groups. THC supplementation significantly increased plasma BHB levels compared to the VEH/SIV/ART group (Fig. 7A).
To further investigate this mechanism, we performed in vitro assays using the hepatocyte cell line, HEPG2, treated with oleic acid as a substrate to assess fatty acid oxidation. Cells were subsequently treated with THC, the peroxisome proliferator–activated receptor alpha (PPARα) agonist fenofibrate (a positive control for fatty acid oxidation), or CBR antagonists AM251 (CB1R) and AM630 (CB2R). After 24 hours, BHB levels were quantified using a commercially available ELISA (enzyme-linked immunosorbent assay). As evident in Fig. 7 (B and C), THC significantly increased BHB production to levels comparable to fenofibrate, indicating enhanced fatty acid oxidation. In addition, cannabidiol (CBD) and THC:CBD combinations also significantly increased BHB levels compared to dimethyl sulfoxide (DMSO)–treated cells (Fig. 7C). Notably, blocking CB2 receptors with AM630 significantly increased BHB levels compared to both 1 and 3 μM THC concentrations (Fig. 7C), suggesting that THC’s enhancement of fatty acid oxidation is stronger when signaling through CB1R (Fig. 7C). These findings reveal that THC supplementation enhances fatty acid oxidation and increases BHB levels, suggesting a potential role in reducing hepatic lipid accumulation, dyslipidemia, and associated CVD risks in PWH under ART. By promoting hepatic metabolic balance and enhancing neuro- and cardioprotective metabolites like BHB, THC may serve as a valuable adjunct therapy in managing metabolic complications associated with HIV and long-term ART.
To determine whether long-term low-dose THC exerted cardioprotective effects during HIV/SIV infection under ART treatment, we analyzed the levels of metabolites related to CVD in THC/SIV/ART and VEH/SIV/ART RMs compared to preinfection controls. A recently discovered group of long-chain unsaturated fatty acid acylcholines (37) has been reported to inhibit the nonneuronal acetylcholine signaling pathway, particularly in macrophages, by affecting the cholinergic receptor sensitivity of acetylcholine with varying degrees of antagonism, thereby promoting pro-inflammatory cytokine production by macrophages and other innate immune cells (Fig. 7D) (38). In a previous clinical study, untargeted UPLC-MS profiling identified three acylcholines, namely, oleoylcholine, linoleoylcholine, and arachidonoylcholine, to be elevated in blood vessels surgically resected from patients with CVD (abdominal aortic aneurysm, stenotic carotid plaques, femoral stenotic plaques, and intimal thickening) (depicted in Fig. 7E) (37, 39, 40). In another study, acylcholines were increased in blood samples obtained from patients with a high risk of pulmonary embolism (PE) compared to samples from patients with an intermediate risk of PE (41). In our study, seven of the eight acylcholines detected were significantly increased only in the plasma of VEH/SIV/ART RMs (Fig. 7E) compared to preinfection controls, whereas THC/SIV/ART RMs (Fig. 7F) showed levels that mirrored preinfection levels. The significantly increased acylcholines in VEH/SIV/ART included those associated with CVD and inhibition of acetylcholine signaling, namely, linoleoylcholine (P = 0.0004) (Fig. 7G), oleoylcholine (P = 0.0008) (Fig. 7H), arachidonoylcholine (P = 0.0006) (Fig. 7I), and palmitoylcholine (P = 0.0004) (Fig. 7J). Other acylcholines that significantly decreased in VEH/SIV/ART RMs included stearoylcholine (P = 0.0003) (Fig. 7K), dihomo-linolenoylcholine (P = 0.0009) (Fig. 7L), and docosahexaenoylcholine (P = 0.0008) (Fig. 7M). These pivotal findings strongly suggest that long-term low-dose THC administered as an adjunct to ART may help maintain healthy plasma acylcholine levels and thereby reduce systemic inflammation (8, 11, 12), which may potentially help lower CVD risk in PWH.
Plasma ART concentrations are reduced in THC-treated SIV-infected RMs despite undetectable viral loads
As previously mentioned, plasma viral loads were at undetectable levels in both the THC/SIV/ART and VEH/SIV/ART groups at 5 MPI. Metabolic pathway figures show all ART drugs detected [significantly increased (red)] in VEH/SIV/ART (Fig. 8A) and THC/SIV/ART (Fig. 8B) relative to preinfection samples. Unexpectedly, plasma levels of ART drugs 9-[9(R)-2-(phosphonomethoxy)propyl]adenine (PMPA) also known as tenofovir (P = 0.1356, FC = 38) (Fig. 8, B and C), and emtricitabine (P = 0.2321, FC = 28) (Fig. 8, B and D) were modestly increased in THC/SIV/ART RMs at 5 MPI compared to preinfection controls. Conversely, VEH/SIV/ART RMs showed significantly elevated levels and extremely high FCs of tenofovir (P = 0.03, FC = 231.55) (Fig. 8, A and C) and emtricitabine (P = 0.005, FC = 204.89) (Fig. 8, A and D) compared to preinfection controls. Nevertheless, at 5 MPI, plasma levels of both Tenofovir (P = 0.061) and Emtricitabine (P = 0.045) were substantially reduced in THC/SIV/ART compared to VEH/SIV/ART RMs (Fig. 8, C and D). Dolutegravir was detected at significantly elevated levels in both groups relative to their preinfection time points (Fig. 8E); however, THC/SIV/ART (FC = 2.69) (Fig. 8, B and E) showed a lower FC compared to VEH/SIV/ART (FC = 3.32) (Fig. 8, A and E).
DISCUSSION
Although the advent of modern combined ART has helped PWH achieve a near-normal life span similar to uninfected individuals, residual inflammation and metabolomic alterations have increased their risk for developing non-AIDS–associated comorbidities. The irreversible disruption of GI homeostasis during acute HIV/SIV infection and the consequent dysregulation of the microbiome and its metabolites can influence chronic inflammation. The kynurenine pathway is an important signal transduction pathway that metabolizes tryptophan using the enzyme IDO1 to produce nicotinamide adenine dinucleotide. Because the induction and maintenance of IDO1 protein expression represents an important triggering event in the maintenance of the chronic inflammatory state in progressive HIV/SIV infection (42), targeting the kynurenine-IDO1 pathway represents a viable therapeutic option to reduce chronic inflammation and age-related comorbidities in PWH. In this context, we previously showed that long-term low-dose THC successfully inhibited IDO1 activity via a CB2 receptor mechanism, resulting in reduced plasma kynurenine, kynurenate, and quinolinate levels in SIV-infected RMs (11), including reduced kynurenine/tryptophan ratios (8), but concurrently increased the levels of another beneficial tryptophan metabolite, IPA. The increase in IPA was associated with a significant increase in the relative abundance of three Clostridial species that encode the phenyllactate dehydratase gene cluster that catalyzes the conversion of tryptophan to IPA in the colon. In the present study, we show that, under suppressive ART, cannabinoid-mediated inhibition of IDO1 resulted in not only significantly high levels of serotonin and IPA in plasma but also expression of their respective signaling receptors in the intestine and brain. In addition, low-dose THC stimulated fatty acid oxidation and increased BHB levels while reducing the production and levels of toxic SBAs and pro-inflammatory acylcholines.
Given the importance of tryptophan metabolism to maintaining immune homeostasis, we focused on tryptophan metabolites and detected significantly high plasma serotonin levels in SIV-infected RMs that received supplemental long-term low-dose THC, thereby confirming our hypothesis. Supplemental THC significantly increased not only serotonin production by ECCs but also the expression of its signaling receptors HTR4 and HTR7 in jejunum epithelial and lamina propria cells via a predominantly CB2R-mediated mechanism. Although peripheral serotonin does not cross the BBB, the strong expression of HTR4 and HTR7 on myenteric plexus neurons identified a signaling circuitry for gut-derived serotonin to signal to the brain and regulate brain function and activity (22). L. plantarum supplementation has been shown to increase the number of serotonin-producing ECCs (43) and enhance serotonin production by ~34%. THC modulated the gut microbiome, particularly L. plantarum species, suggesting its potential to augment serotonin production via exerting probiotic-like effects. In this context, the HTR4 agonist (Tegaserod/Zelnorm) inhibited dextran sulfate sodium–induced colitis (44) and recently was Food and Drug Administration (FDA) approved as a prescription drug for relieving constipation, abdominal pain, and bloating in female patients with (<65 years) IBS free of cardiovascular risk factors (45–49). Patients with IBD have low serum serotonin levels, and its supplementation reduced intestinal inflammation (50). Similarly, PWH on ART also have low plasma serotonin (51) levels. Moreover, serotonin levels decreased during acute COVID-19 and showed drastic reduction in severe cases of post–acute sequelae of COVID-19 (Long Covid) (52). Recent studies showed that the intestinal epithelium–derived serotonin alone is sufficient to modulate mood, anxiety, and depression symptoms via afferent vagal pathways, suggesting a gut-targeted therapeutic approach to boost serotonin production (53). Put together, our findings go one step further to show that low-dose phytocannabinoids not only increased serotonin levels through IDO1 inhibition and modulating L. plantarum levels but also induced HTR4 and HTR7 protein expression in the intestine and, in doing so, could enhance physical signaling along the MGBA through vagal afferent pathways. As a result, low-dose cannabinoids may potentially help PWH, IBD, and potentially patients with Long Covid in strengthening the intestinal epithelial barrier and alleviating neuropsychiatric symptoms in a much safer way than selective serotonin reuptake inhibitors that are often associated with serious adverse effects (54).
Under physiological conditions, ~7% of the absorbed tryptophan is metabolized by gut bacteria to indole metabolites. In ART-naïve SIV-infected RMs, we showed that low-dose THC increased the relative abundance of bacterial species encoding the phenyllactate dehydratase cluster (Clostridium botulinum, Clostridium paraputrificum, and Clostridium cadaveris) (15) in the colon that can convert tryptophan to IPA. Because of its strong antioxidant, intestinal barrier–strengthening, and cardio- and neuroprotective properties, stimulating IPA production via microbiome modulation using low-dose cannabinoids offers a feasible intervention to reduce intestinal and neuroinflammation in PWH, who otherwise have low plasma IPA (55) levels despite undetectable plasma viremia. Owing to its BBB permeability, increasing IPA levels represents an important mechanism by which cannabinoids could increase chemical signaling along the MGBA in HIV/SIV infection. Apart from tryptophan metabolites, low-dose THC also increased plasma levels of BHB, a ketone body with potent anti-inflammatory and neuro-, hepato-, and cardioprotective properties (56, 57). Because BHB is a by-product of fatty acid oxidation, this finding suggests that low-dose THC may stimulate mitochondrial beta-oxidation of long-chain fatty acids in the liver. Consistent with this assumption, an in vitro BHB assay further confirmed the ability of cannabinoids to stimulate fatty acid oxidation mainly through a CB1R-mediated mechanism. This is an important translational finding with high clinical relevance as dietary modulation, including caloric restriction (intermittent fasting), is being used to increase fatty acid oxidation and BHB levels to improve metabolic complications caused by insulin resistance and reduce aging, including cellular senescence and chronic inflammation. Overall, our findings point to an important, previously unknown physiological effect of phytocannabinoids that can be clinically used to increase fatty acid oxidation and potentially help reduce the risk of metabolic syndrome, dyslipidemia, and NAFLD in PWH.
Among many other non-AIDS–associated comorbidities, PWH have a high risk of developing NAFLD (58) and CLD (59), which, if left untreated, leads to liver fibrosis, biliary cirrhosis, and ultimately end-stage liver disease (34). Hepatotoxicity, or drug-induced liver injury, is one of the most common adverse reactions induced by many widely used ART drugs, including dolutegravir (60–66). Considering that a significant portion of PWH is coinfected with hepatitis B/C virus, which has a cumulative impact on the liver, preexisting liver disease may exacerbate the development of liver toxicity in PWH (67). In cholestasis, the accumulation of toxic hydrophobic SBAs in the hepatobiliary system damages bile duct epithelial cells and hepatocytes, causing liver injury and inflammation (33). In the intestine, gut bacteria metabolize PBAs CA and CDCA to SBAs DCA and LCA, respectively (33). UDCA, the only drug approved to effectively reduce the retention of toxic bile acids in liver cells and alleviate liver damage (68, 69), has limited efficacy in CLD (70), with some patients experiencing adverse GI effects (71). Therefore, there is an urgent need for in-depth investigations to understand the pathogenesis of HIV/ART-induced cholestasis and develop more effective interventions to maintain the gut microbiota–bile acid–liver cross-talk at homeostatic levels in PWH. The important finding of significantly reduced plasma SBA levels in THC/SIV/ART RMs has high translational relevance and may be explained by the reduced gut dysbiosis, reduced plasma cholesterol levels, and faster ART drug clearance from the blood. Recently, Li et al. (17) reported the presence of Oscillibacter species to be associated with decreased cholesterol levels in the blood and stool. In this context, the ability of THC to significantly increase the relative abundance of Oscillibacter in the colonic contents may partly explain the reduced plasma cholesterol levels as these bacteria encode the intestinal stool metabolism A (IsmA) gene to metabolize cholesterol into coprostanol.
We recently described changes in the colonic microbiome in response to SIV infection and THC administration (8) and successfully identified the high relative abundance of IPA-producing Clostridial species along with numerous Lactobacilli and Bifidobacteria in the colonic contents of THC-treated SIV-infected RMs. Therefore, in the current study, we were intrigued to find out whether we could detect the same bacterial communities in the jejunum and whether differences existed between the colon and jejunum. Although we detected several bacterial phyla, classes, families, and genera that were common to the colon and jejunum, our findings demonstrated that IPA-producing Clostridial species and cholesterol-metabolizing Oscillibacter species were enriched in the colon but not in the jejunum and that phytocannabinoids represented a clinically viable intervention to increase the relative abundance of Oscillibacter to decrease cholesterol levels in PWH. These findings clearly address an important knowledge gap regarding microbial differences between the small and large intestines. Our findings emphasize the importance of focusing on the colon, preferably the descending colon, for future microbiome studies as it contains several 100-fold more microbes than any other intestinal segment and is a major source of microbial metabolites that can affect host physiology. Therefore, examining the colon microbiota instead of fecal microbiota is of significant interest to human health. Furthermore, THC’s ability to stimulate fatty acid oxidation and increase BHB levels suggests its clinical utility in reducing hepatic lipid accumulation, dyslipidemia, and associated atherosclerotic CVD risk in PWH under ART. By promoting hepatic metabolic balance and increasing the production of neuro- and cardioprotective metabolites like BHB, cannabinoids may serve as a valuable adjunct therapy in managing metabolic complications associated with HIV and long-term ART. Last, reduced plasma ART drug concentrations despite undetectable SIV RNA in plasma and the jejunum suggest faster ART metabolism in THC/SIV RMs, which may reduce circulating levels of ART and protect the liver from drug-mediated toxicity.
Because PWH on ART showed depletion of Clostridia (72) and enrichment of Enterobacteriaceae (73), our findings suggest that, by reducing intestinal inflammation, THC may maintain a state of physiologic hypoxia in the intestinal lumen, thereby facilitating the growth of anaerobic commensals like Clostridia and Lactobacilli while reducing the levels of inflammation-promoting Gammaproteobacteria and Enterobacteriaceae. Owing to their anti-inflammatory and anti-allergic properties, Clostridium species are now being considered as next-generation probiotics for promoting human and animal health (74). Yet, another clinically applicable finding that emerged from this study is the significant elevation of acylcholine levels in SIV-infected RMs on long-term suppressive ART and its prevention by phytocannabinoid supplementation. Because Ach activates the cholinergic anti-inflammatory signaling pathway, particularly in macrophages, elevated long-chain acylcholines can outcompete Ach by binding Ach receptors, thereby promoting macrophage activation and potentially driving atherosclerotic CVD by increasing pro-inflammatory cytokine production within plaques. However, it remains to be determined whether chronic immune activation, ART toxicity in the liver, and microbiome dysbiosis are the factors driving high acylcholine levels. In contrast, the maintenance of normal acylcholine levels in THC/SIV/ART RMs is a very important and beneficial finding, which could be attributed to THC’s anti-inflammatory, hepatoprotective, and antidysbiotic properties. These data, together with recent clinical trial findings (75) showing the effectiveness of oral low-dose THC:CBD combination in reducing pro-inflammatory cytokine production and T cell senescence in PWH on ART (75), are a clear testament to the high translational significance of our nonhuman primate (NHP) findings.
Despite these findings, our study has limitations, and several questions remain unanswered. Although data on the relative abundance of Gammaproteobacteria, Clostridiaceae, Enterobacteriaceae, and Lactobacilli have been reported in humans (72, 73), the reader should still exercise caution while interpreting the findings, given that diet and housing conditions can shape the composition and diversity of the gut microbiome in NHPs. Although NHPs in this study were fed a nutritionally and fiber-balanced primate diet, the different levels of fiber and fat in the diets of PWH can affect the gut microbiome differently as a high-fiber diet promotes a healthy and diverse microbiome, whereas a saturated fat–rich diet disrupts the microbial balance and promotes dysbiosis. Similarly, pair housing promotes affiliative interactions and allows for a more natural social experience to reduce stress, anxiety, and abnormal behaviors, all of which can affect the gut microbiome composition. For these reasons, NHPs in all our past and current studies were pair-housed to ensure their psychological well-being and overall welfare. It remains to be determined whether phytocannabinoids increase serotonin and IPA levels in the brain and the cerebrospinal fluid. BHB is known to exert antiseizure effects in various epileptic models (76). CBD was recently FDA approved for the treatment of childhood epilepsy (76), and given that both THC and CBD increased BHB levels, future studies are needed to determine whether the neuroprotective effects of phytocannabinoids like CBD are mediated through increased production of BHB in the brain. Unlike the intramuscular route used in the present study, because oral cannabinoids are first transported to the liver and metabolized, it is important to determine how oral administration, which generally requires slightly higher doses, affects PBA and SBA levels. Given the pro-inflammatory properties and the limited information known about the biological activity of acylcholines, future research is needed in this area owing to the high prevalence of CVD and metabolic syndrome in PWH (30, 31). Although THC reduced plasma ART drug levels, it is unclear how this affects the intracellular concentrations of the active drug metabolites and whether these alterations influence the size of the replication-competent latent HIV/SIV reservoir in CD4 T cells and macrophages. Last, as NHPs are “an acutely scarce resource,” the use of only male NHPs, as done in the current study, is specifically acceptable under NOT-OD-15-102 (Consideration of Sex as a Biological Variable in NIH-funded Research). Nevertheless, future studies should look at the effect of cannabinoids on systemic inflammation and other outcomes and parameters in both male and female ART-suppressed SIV-infected Indian RMs.
Taken as a whole, our findings uncover numerous hitherto unknown mechanisms of cannabinoid action and provide multiple lines of evidence for its utility as an effective and relatively safe adjunct therapy to ART. Among these, an important uncharacterized mechanism is its ability to increase serotonin production and its signaling both directly and indirectly through gut microbiome modulation (Fig. 9A). A second mechanism is through inducing the expression of Ahr, a receptor for the tryptophan metabolite, IPA, and other indoles (24), in the intestine and brain (Fig. 9A). Ahr activation enhances communication along the gut-brain axis, attenuating systemic and neuroinflammatory processes while promoting neuronal survival, highlighting the need for selective Ahr modulators to optimize therapeutic outcomes (77–79). A third and very important mechanism is through stimulating the oxidation of long-chain fatty acids and increasing the production of its by-product, BHB, which can not only exert protective effects in the intestine and liver but also cross the BBB and exert neuroprotective effects (Fig. 9B). Furthermore, by reducing the levels of toxic SBAs (via reducing dysbiosis) and long-chain acylcholines, low-dose cannabinoids may protect against NAFLD, CLD, and CVD (Fig. 9B). Last, although faster clearance of ART drugs from the circulation may appear to be disadvantageous, the complete suppression of viral loads in the blood and intestine suggests that this effect of cannabinoids may be beneficial in protecting PWH against ART toxicity and even promote adherence. The reduced SBA and acylcholine levels are likely the strongest indicators of reduced ART toxicity, particularly in the liver. On the basis of the findings from the recent human clinical trial on the positive effects of oral low-dose cannabinoids (THC:CBD combination) on chronic immune activation in PWH (75), our extensive findings, in addition to uncovering clinically applicable effects of low-dose cannabinoids, suggest the need to continue researching the effects of cannabinoids (THC) as PWH and those suffering from chronic inflammatory and metabolic diseases may benefit from its inclusion as an adjunct to standard ART and other medications. Because the CBD and THC:CBD (1:1) combination also increased BHB levels, combining THC with CBD should help reduce any residual psychotropic effects of THC and concurrently improve each phytocannabinoid’s ability to positively modulate the endocannabinoid system (80). Overall, our findings represent a substantial advance as they open up fresh areas of investigation (next-generation safe synthetic cannabinoids or CBR agonists with targeted therapeutic benefits), introduce uncharted concepts or mechanisms of cannabinoid action that were previously unknown, unexplored, or not widely accepted, and have broad applicability to not only HIV but also other diseases such as IBD, IBS, chronic liver diseases, CVD, and diseases affecting the brain such as AD and PD.
MATERIALS AND METHODS
Animal care, ethics, and experimental and clinical procedures
All experiments using RMs were approved by the Tulane Institutional Animal Care and Use Committee (protocols 3581 and 3781). The Tulane National Primate Research Center (TNPRC) is an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International–accredited facility (AAALAC #000594). The NIH Office of Laboratory Animal Welfare assurance number for the TNPRC is A3071-01. All clinical procedures, including administration of anesthesia and analgesics, were carried out under the direction of a laboratory animal veterinarian. Animals were anesthetized with ketamine hydrochloride for blood collection procedures. Intestinal pinch biopsies were performed by laboratory animal veterinarians. Animals were preanesthetized with ketamine hydrochloride, acepromazine, and glycopyrrolate, intubated, and maintained on a mixture of isoflurane and oxygen. All possible measures were taken to minimize the discomfort of all the animals used in this study. Tulane University complies with the NIH policy on animal welfare, the Animal Welfare Act, and all other applicable federal, state, and local laws.
Animal model and experimental design
Sixteen Mamu A01/B08/B17 negative male Indian RMs were randomly divided into two groups and pair-housed until the end of the study. Group 1 (VEH/SIV/ART; n = 8) and group 2 (THC/SIV/ART; n = 8) were infected intravenously with 100 TCID50 (median tissue culture infectious dose) of SIVmac251. Beginning 2 weeks post–SIV infection, both groups received ART (tenofovir or PMPA, 20 mg/kg; FTC or emtricitabine, 30 mg/kg; and dolutegravir, 2.5 mg/kg) daily by a subcutaneous route. Group 2 received twice-daily injections of Δ9-THC (THC) as previously described (8, 14) until 6 months post–SIV infection. We initiated THC at 0.18 mg/kg and subsequently increased the dose for each subject to 0.32 mg/kg over a period of ~2 weeks when responding was no longer affected by 0.18 mg/kg daily (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 our previously published studies (8, 14), this dose of THC showed protection, we used the same dose in this study. As mentioned previously (13, 14), four macaques in group 1 and four macaques in group 2 received two injections of anti-α4β7 integrin (VEHSIVART3, VEHSIVART8, THCSIVART6, and THCSIVART8) or control immunoglobulin G (IgG) (VEHSIVART6, VEHSIVART7, THCSIVART5, and THCSIVART7) (50 mg/kg of anti-α4β7 or control IgG) beginning 4 MPI at 3-week intervals as part of another study before the jejunum resections were collected at 5 MPI. Note that the anti-α4β7 antibody functions by blocking α4β7 positive T cells from trafficking to the intestine and did not influence viral rebound after six to eight treatments at 3-week intervals. These animals were also used in our previously published studies (13, 14), and the data showed that both anti-α4β7 and control IgG did not affect the MMP25-AS1 mRNA (CT values) and MMP25 and NLRP6 protein expression (fluorescence intensity) in the jejunum. Similarly, we did not see any impact of anti-α4β7 and control IgG on HTR4, HTR7, and Ahr protein expression. Nevertheless, we have and continue to report this information in all manuscripts that use these animals to maintain transparency about the experimental manipulations.
At 5 MPI, jejunum resections were collected in RNAlater (Thermo Fisher Scientific) and Z-fix for total RNA extraction and embedding in paraffin blocks, respectively. Plasma and jejunum samples collected at preinfection served as controls (n = 16). Basal ganglia (brain) samples from both groups were collected at necropsy. In addition, brain samples were collected from three uninfected control macaques (CONT1, CONT2, and CONT3) (group 3) for immunofluorescence studies (Fig. 4H). The global shortage of RMs due to their extensive use for COVID-19 research made it harder and more challenging to find many uninfected male Indian RMs. Therefore, we included one uninfected female RM (CONT2) for immunofluorescence studies that was immediately available at our facility. We undertook a similar approach in our previous published study (8). Table 1 contains all the relevant information about the RMs used in the current study.
HIEC-6 and HCN-2 cell culture experiments
HIEC-6 cells [American Type Culture Collection (ATCC), Manassas, VA, USA; catalog no. CRL-3266] were first cultured in Opti-MEM (Thermo Fisher Scientific, Waltham, MA, USA) containing 20 mM Hepes, sodium bicarbonate (2.4 g/liter), 10 mM GlutaMAX, and epidermal growth factor (10 ng/ml) and supplemented with 4% fetal bovine serum (FBS) in 75 cm2 flasks (Thermo Fisher Scientific) at 37°C in a humidified atmosphere with 5% CO2. HCN-2 cells (ATCC, Manassas, VA, USA; catalog no. CRL-3592) were first cultured in Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific) containing sodium bicarbonate (1500 mg/liter), glucose (4500 mg/liter), 4 mM l-glutamine, and 1 mM sodium pyruvate and supplemented with 10% FBS in 8-well chamber slides (Cellvis) at 37°C in a humidified atmosphere with 5% CO2. To determine THC dose-response on HTR4, HTR7, and Ahr expression in HIEC and HCN-2 cells, the cells were plated in 8-well chambers (Cellvis) and treated with 0.5, 1.0, 1.5, 2.0, and 3.0 μM THC. After 24-hour incubation, the cells were fixed with 2% paraformaldehyde and immunostained with HTR4, HTR7, and Ahr antibodies. In addition, another set of 8-well chambers was plated with HIEC-6 cells and treated with THC (3 μM) only, or AM251, and AM630 first, followed by THC (3 μM) after 1 hour. After 24-hour incubation, cells were fixed and stained with HTR4 and HTR7 antibodies. The HTR4, HTR7, and Ahr protein expression images were obtained using a Zeiss 700 confocal microscope. Signal intensity was quantified using HALO software (Indica Labs).
Immunofluorescence staining
The immunofluorescence staining for serotonin, HTR4, HTR7, Ahr, and NeuN on the jejunum, colon, and basal ganglia and fixed in vitro cultured cells was performed as described previously (14). Information about antibodies used for immunofluorescence studies is provided in table S1 (Supplementary Materials).
Quantitative image analysis
Briefly, two slides containing colon, jejunum, or basal ganglia tissue sections from each animal were stained with antibodies specific for serotonin, HTR4, HTR7, and Ahr. No differences in staining intensity were detected between slides for each macaque. A total of five to eight bright-field sections for each RM (from VEH/SIV/ART, THC/SIV/ART, and uninfected controls) were scanned using a Zeiss LSM700 confocal microscope (Carl Zeiss Microscopy, LLC) at 20X objective. Digital images were imported into the HALO software (Indica Labs) for image quantitation analysis. Because colonic epithelial cells were used for mRNA profiling (fig. S1, D and E) (14), we first demarcated the epithelial regions in the jejunum before fluorescence signal quantitation. Similarly, for basal ganglia (brain), we first identified and demarcated mature neuronal cells expressing the NeuN marker. Next, we used the area quantification module available on HALO v3.2 (Indica Labs) to quantify serotonin, HTR4, and HTR7 (green signal/Alexa Fluor 488) and Ahr (red signal/Alexa Fluor 568) fluorescence. Similarly, for in vitro cell culture experiments (HIEC-6 or HCN-2 cells stained with HTR4, HTR7, and Ahr), five bright-field images from each treated well in 8-well chambers were captured. In this new computational method, the artificial intelligence–driven software identifies all cells that express serotonin, HTR4, and HTR7 in green, Ahr in red, and nuclei in blue [4′,6-diamidino-2-phenylindole (DAPI)] and categorizes the cells on the basis of predefined fluorescence intensity levels. Specifically, the HALO software normalizes the threshold across all images and enables quantitation of the number of cells and relative intensity of fluorescence per cell of single-channel fluorescence (green or red) corresponding to the expression of serotonin, HTR4, HTR7, and Ahr that was intensely expressed in the cells. The output values (total area and average positive intensity) were used to calculate the total fluorescent intensity/tissue area. The data were graphed using Prism v9 software (GraphPad software).
BHB colorimetric assay
For the BHB assay, HEPG2 cells (ATCC, Manassas, VA, USA; catalog no. HB-8065) were first cultured in EMEM (ATCC, Manassas, VA, USA) containing sodium bicarbonate (1500 mg/liter), 2 mM l-glutamine, and 1 mM sodium pyruvate and supplemented with 10% FBS in 8-well chamber slides (Cellvis, Mountain View, CA, USA) at 37°C in a humidified atmosphere with 5% CO2. Cells were plated in 6-well plates and allowed to reach 90% confluence. Once confluent, the cells were treated with oleic acid (10 mM) in all wells as the fatty acid substrate. Subsequently, three wells each were treated with DMSO as a control, THC (3 μM), fenofibrate (PPARα agonist, 10 μM), and GW6471 (PPARα antagonist, 10 μM). In addition, three wells each were treated with AM251 (5 μM) and AM630 (5 μM) followed by THC (3 μM) after 1-hour incubation at 37°C. The cells were then incubated for 24 hours posttreatment, harvested, and lysed using the lysis buffer provided in the Abcam BHB assay kit (catalog no. ab83390). The lysates were centrifuged, supernatants collected, and the BHB assay was performed following the Abcam BHB colorimetric assay protocol (catalog no. ab83390). The assay’s absorbance was measured at OD450nm (optical density at 450 nm).
Statistical analysis
For metabolite data analysis (by Metabolon Inc.), after log transformation and imputation of missing values, if any, with the minimum observed value for each compound, Welch’s two-sample t test was used as a significance test to identify biochemicals that differed significantly (P < 0.05) between experimental groups as previously described (8). Imputation for replacing missing values in the metabolite data analysis was performed by replacing the missing values with their observed minimum (performed after batch normalization). Imputing with the minimum was chosen on the basis of Metabolon’s internal simulation studies comparing this to other methods regarding the type I error and power for the two-sample t test.
For metagenomic sequencing, raw sequencing reads were processed to obtain valid reads for further analysis as previously described (8). First, sequencing adapters were removed from sequencing reads using cutadapt v1.9. Second, low-quality reads were trimmed by fqtrim v0.94 using a sliding-window algorithm. Third, reads were aligned to the host genome using bowtie2 to remove host contamination. Once quality-filtered reads were obtained, they were de novo assembled to construct the metagenome for each sample by IDBA-UD. All coding regions (CDSs) of metagenomic contigs were predicted by MetaGeneMark v3.26. CDS sequences of all samples were clustered by CD-HIT v4.6.1 to obtain unigenes. Unigene abundance for a certain sample was estimated by transcripts per million on the basis of the number of aligned reads by bowtie2 v2.2.0. The lowest common ancestor taxonomy of unigenes was obtained by aligning them against the NCBI NR database by DIAMOND v 0.7.12. To determine the abundance profile of unigenes, differential analysis was carried out at each taxonomic level by Fisher’s exact test. Unigenes with P value < 0.05 and |log2 FC| > 0.585 were defined as differentially expressed. One-way analysis of variance (ANOVA) (GraphPad, Prism v9) was used for immunofluorescence image quantification analysis of serotonin (in the jejunum), HTR4, HTR7 (in the jejunum and HIEC-6 cells), and Ahr (in the jejunum, basal ganglia neurons, and HCN-2 cells). One-way ANOVA (GraphPad, Prism v9) was used for the BHB assay data analysis. The Shapiro-Wilk test (GraphPad Prism v9) was used to test the normality of immunofluorescence and BHB data.
Acknowledgments
We would like to thank P. P. Aye, F. Schiro, K. P. Shannon, J. Mansfield, and the veterinary staff and pathology personnel at the Tulane National Primate Research Center for assistance in animal work and necropsy sample collection.
Untitled section
Funding: Research reported in this publication was supported by the National Institutes of Health award numbers R01DA042524 (M.M.) and R01DA052845 (M.M.), P30AI161943, P51OD011104, and P51OD111033. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The funding agency (NIH) had played no role in the study design, data collection, data analyses, interpretation, or writing of the manuscript.
Untitled section
Author contributions: Conceptualization: M.M. and L.S.P. Methodology: M.M. and L.S.P. Validation: L.S.P. Formal analysis: L.S.P. and M.M. Investigation: L.S.P., L.R., and M.M. Resources: M.M. Visualization: L.S.P. and M.M. Supervision: M.M. and L.S.P. Project administration: M.M. and L.S.P. Funding acquisition: M.M. Writing—original draft: L.S.P. Writing—review and editing: M.M. and L.S.P.
Untitled section
Competing interests: The authors declare that they have no competing interests.
Untitled section
Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Shotgun metagenomic raw sequencing data have been submitted to the Sequence Read Archive (SRA) with BioProject ID: PRJNA1260450 and can be accessed using the link https://www.ncbi.nlm.nih.gov/sra/?term=PRJNA1260450.
Supplementary Materials
This PDF file includes:
REFERENCES AND NOTES
Untitled section
References
- 1.Webel A. R., Schexnayder J., Cioe P. A., Zuñiga J. A., A review of chronic comorbidities in adults living with HIV: State of the science. J. Assoc. Nurses AIDS Care 32, 322–346 (2021).
- 2.Collins L. F., Armstrong W. S., What it means to age with HIV infection: Years gained are not comorbidity free. JAMA Netw. Open 3, e208023 (2020).
- 3.Hadavandsiri F., Shafaati M., Mohammad Nejad S., Ebrahimzadeh Mousavi M., Najafi A., Mirzaei M., Narouee S., Akbarpour S., Non-communicable disease comorbidities in HIV patients: Diabetes, hypertension, heart disease, and obstructive sleep apnea as a neglected issue. Sci. Rep. 13, 12730 (2023).
- 4.McCutcheon K., Nqebelele U., Murray L., Thomas T. S., Mpanya D., Tsabedze N., Cardiac and renal comorbidities in aging people living with HIV. Circ. Res. 134, 1636–1660 (2024).
- 5.Jia W., Xie G., Jia W., Bile acid-microbiota crosstalk in gastrointestinal inflammation and carcinogenesis. Nat. Rev. Gastroenterol. Hepatol. 15, 111–128 (2018).
- 6.Chawla A., Wang C., Patton C., Murray M., Punekar Y., de Ruiter A., Steinhart C., A review of long-term toxicity of antiretroviral treatment regimens and implications for an aging population. Infect. Dis. Ther. 7, 183–195 (2018).
- 7.Hu A., Zaongo S. D., Harypursat V., Wang X., Ouyang J., Chen Y., HIV-associated neurocognitive disorder: Key implications of the microbiota-gut-brain axis. Front. Microbiol. 15, 1428239 (2024).
- 8.McDew-White M., Lee E., Premadasa L. S., Alvarez X., Okeoma C. M., Mohan M., Cannabinoids modulate the microbiota-gut-brain axis in HIV/SIV infection by reducing neuroinflammation and dysbiosis while concurrently elevating endocannabinoid and indole-3-propionate levels. J. Neuroinflammation 20, 62 (2023).
- 9.Pernomian L., Duarte-Silva M., de Barros Cardoso C. R., The aryl hydrocarbon receptor (AHR) as a potential target for the control of intestinal inflammation: Insights from an immune and bacteria sensor receptor. Clin. Rev. Allergy Immunol. 59, 382–390 (2020).
- 10.Sathyasaikumar K. V., Blanco-Ayala T., Zheng Y., Schwieler L., Erhardt S., Tufvesson-Alm M., Poeggeler B., Schwarcz R., The tryptophan metabolite indole-3-propionic acid raises kynurenic acid levels in the rat brain in vivo. Int. J. Tryptophan Res. 17, 11786469241262876 (2024).
- 11.McDew-White M., Lee E., Alvarez X., Sestak K., Ling B. J., Byrareddy S. N., Okeoma C. M., Mohan M., Cannabinoid control of gingival immune activation in chronically SIV-infected rhesus macaques involves modulation of the indoleamine-2,3-dioxygenase-1 pathway and salivary microbiome. EBioMedicine 75, 103769 (2022).
- 12.Kumar V., Torben W., Mansfield J., Alvarez X., Vande Stouwe C., Li J., Byrareddy S. N., Didier P. J., Pahar B., Molina P. E., Mohan M., Cannabinoid attenuation of intestinal inflammation in chronic SIV-infected rhesus macaques involves t cell modulation and differential expression of micro-RNAs and pro-inflammatory genes. Front. Immunol. 10, 914 (2019).
- 13.Premadasa L. S., Lee E., McDew-White M., Alvarez X., Jayakumar S., Ling B., Okeoma C. M., Byrareddy S. N., Kulkarni S., Mohan M., Cannabinoid enhancement of lncRNA MMP25-AS1/MMP25 interaction reduces neutrophil infiltration and intestinal epithelial injury in HIV/SIV infection. JCI Insight 8, e167903 (2023).
- 14.Premadasa L. S., McDew-White M., Romero L., Gondo B., Drawec J. A., Ling B., Okeoma C. M., Mohan M., Epigenetic modulation of the NLRP6 inflammasome sensor as a therapeutic modality to reduce necroptosis-driven gastrointestinal mucosal dysfunction in HIV/SIV infection. Cell Commun. Signal 23, 199 (2025).
- 15.Roager H. M., Licht T. R., Microbial tryptophan catabolites in health and disease. Nat. Commun. 9, 3294 (2018).
- 16.Godzien J., Kalaska B., Rudzki L., Barbas-Bernardos C., Swieton J., Lopez-Gonzalvez A., Ostrowska L., Szulc A., Waszkiewicz N., Ciborowski M., García A., Kretowski A., Barbas C., Pawlak D., Probiotic Lactobacillus plantarum 299v supplementation in patients with major depression in a double-blind, randomized, placebo-controlled trial: A metabolomics study. J. Affect. Disord. 368, 180–190 (2025).
- 17.Li C., Stražar M., Mohamed A. M. T., Pacheco J. A., Walker R. L., Lebar T., Zhao S., Lockart J., Dame A., Thurimella K., Jeanfavre S., Brown E. M., Ang Q. Y., Berdy B., Sergio D., Invernizzi R., Tinoco A., Pishchany G., Vasan R. S., Balskus E., Huttenhower C., Vlamakis H., Clish C., Shaw S. Y., Plichta D. R., Xavier R. J., Gut microbiome and metabolome profiling in Framingham heart study reveals cholesterol-metabolizing bacteria. Cell 187, 1834–1852.e19 (2024).
- 18.Moore M. P., Shryack G., Alessi I., Wieschhaus N., Meers G. M., Johnson S. A., Wheeler A. A., Ibdah J. A., Parks E. J., Rector R. S., Relationship between serum β-hydroxybutyrate and hepatic fatty acid oxidation in individuals with obesity and NAFLD. Am. J. Physiol. Endocrinol. Metab. 326, E493–E502 (2024).
- 19.Shah P. A., Park C. J., Shaughnessy M. P., Cowles R. A., Serotonin as a mitogen in the gastrointestinal tract: Revisiting a familiar molecule in a new role. Cell. Mol. Gastroenterol. Hepatol. 12, 1093–1104 (2021).
- 20.Quintero-Villegas A., Valdés-Ferrer S. I., Role of 5-HT7 receptors in the immune system in health and disease. Mol. Med. 26, 2 (2019).
- 21.Zhu H., Xiao X., Chai Y., Li D., Yan X., Tang H., MiRNA-29a modulates visceral hyperalgesia in irritable bowel syndrome by targeting HTR7. Biochem. Biophys. Res. Commun. 511, 671–678 (2019).
- 22.De Vadder F., Grasset E., Mannerås Holm L., Karsenty G., Macpherson A. J., Olofsson L. E., Bäckhed F., Gut microbiota regulates maturation of the adult enteric nervous system via enteric serotonin networks. Proc. Natl. Acad. Sci. U.S.A. 115, 6458–6463 (2018).
- 23.Huang Z.-B., Hu Z., Lu C.-X., Luo S.-D., Chen Y., Zhou Z.-P., Hu J.-J., Zhang F.-L., Deng F., Liu K.-X., Gut microbiota-derived indole 3-propionic acid partially activates aryl hydrocarbon receptor to promote macrophage phagocytosis and attenuate septic injury. Front. Cell. Infect. Microbiol. 12, 1015386 (2022).
- 24.Sun J., Zhang Y., Kong Y., Ye T., Yu Q., Kumaran Satyanarayanan S., Su K. P., Liu J., Microbiota-derived metabolite Indoles induced aryl hydrocarbon receptor activation and inhibited neuroinflammation in APP/PS1 mice. Brain Behav. Immun. 106, 76–88 (2022).
- 25.Pallister T., Jackson M. A., Martin T. C., Zierer J., Jennings A., Mohney R. P., MacGregor A., Steves C. J., Cassidy A., Spector T. D., Menni C., Hippurate as a metabolomic marker of gut microbiome diversity: Modulation by diet and relationship to metabolic syndrome. Sci. Rep. 7, 13670 (2017).
- 26.Binda C., Lopetuso L. R., Rizzatti G., Gibiino G., Cennamo V., Gasbarrini A., Actinobacteria: A relevant minority for the maintenance of gut homeostasis. Dig. Liver Dis. 50, 421–428 (2018).
- 27.Vujkovic-Cvijin I., Sortino O., Verheij E., Sklar J., Wit F. W., Kootstra N. A., Sellers B., Brenchley J. M., Ananworanich J., Loeff M. S. V., Belkaid Y., Reiss P., Sereti I., HIV-associated gut dysbiosis is independent of sexual practice and correlates with noncommunicable diseases. Nat. Commun. 11, 2448 (2020).
- 28.Klase Z., Ortiz A., Deleage C., Mudd J. C., Quiñones M., Schwartzman E., Klatt N. R., Canary L., Estes J. D., Brenchley J. M., Dysbiotic bacteria translocate in progressive SIV infection. Mucosal Immunol. 8, 1009–1020 (2015).
- 29.Sharma G., Garg N., Hasan S., Shirodkar S., Prevotella: An insight into its characteristics and associated virulence factors. Microb. Pathog. 169, 105673 (2022).
- 30.Buendia J., Sears S., Mgbere O., Prevalence and risk factors of high cholesterol and triglycerides among people with HIV in Texas. AIDS Res. Ther. 19, 43 (2022).
- 31.Jiménez-Osorio A. S., Jaen-Vega S., Fernández-Martínez E., Ortíz-Rodríguez M. A., Martínez-Salazar M. F., Jiménez-Sánchez R. C., Flores-Chávez O. R., Ramírez-Moreno E., Arias-Rico J., Arteaga-García F., Estrada-Luna D., Antiretroviral therapy-induced dysregulation of gene expression and lipid metabolism in HIV+ patients: Beneficial role of antioxidant phytochemicals. Int. J. Mol. Sci. 23, 5592 (2022).
- 32.Zhu H., Chen L., Chen Z.-X., Lin H., Liu J., Kwek E., Ma K. Y., He W., Wang G., Chen Z.-Y., Synthesis of cholesterol analogues and comparison on their effect on plasma cholesterol with β-Sitosterol. Food Chem. 461, 140820 (2024).
- 33.Li T., Apte U., Bile acid metabolism and signaling in cholestasis, inflammation, and cancer. Adv. Pharmacol. 74, 263–302 (2015).
- 34.Santiago P., Scheinberg A. R., Levy C., Cholestatic liver diseases: New targets, new therapies. Therap. Adv. Gastroenterol. 11, 1756284818787400 (2018).
- 35.Chiang J. Y. L., Ferrell J. M., Bile acid metabolism in liver pathobiology. Gene Expr. 18, 71–87 (2018).
- 36.Han Y.-M., Ramprasath T., Zou M.-H., β-hydroxybutyrate and its metabolic effects on age-associated pathology. Exp. Mol. Med. 52, 548–555 (2020).
- 37.Akimov M. G., Kudryavtsev D. S., Kryukova E. V., Fomina-Ageeva E. V., Zakharov S. S., Gretskaya N. M., Zinchenko G. N., Serkov I. V., Makhaeva G. F., Boltneva N. P., Kovaleva N. V., Serebryakova O. G., Lushchekina S. V., Palikov V. A., Palikova Y., Dyachenko I. A., Kasheverov I. E., Tsetlin V. I., Bezuglov V. V., Arachidonoylcholine and other unsaturated long-chain acylcholines are endogenous modulators of the acetylcholine signaling system. Biomolecules 10, 283 (2020).
- 38.Kinchen J. M., Mohney R. P., Pappan K. L., Long-chain acylcholines link butyrylcholinesterase to regulation of non-neuronal cholinergic signaling. J. Proteome Res. 21, 599–611 (2022).
- 39.Vorkas P. A., Isaac G., Anwar M. A., Davies A. H., Want E. J., Nicholson J. K., Holmes E., Untargeted UPLC-MS profiling pipeline to expand tissue metabolome coverage: Application to cardiovascular disease. Anal. Chem. 87, 4184–4193 (2015).
- 40.Vorkas P. A., Shalhoub J., Isaac G., Want E. J., Nicholson J. K., Holmes E., Davies A. H., Metabolic phenotyping of atherosclerotic plaques reveals latent associations between free cholesterol and ceramide metabolism in atherogenesis. J. Proteome Res. 14, 1389–1399 (2015).
- 41.Zeleznik O. A., Poole E. M., Lindstrom S., Kraft P., Van Hylckama Vlieg A., Lasky-Su J. A., Harrington L. B., Hagan K., Kim J., Parry B. A., Giordano N., Kabrhel C., Metabolomic analysis of 92 pulmonary embolism patients from a nested case-control study identifies metabolites associated with adverse clinical outcomes. J. Thromb. Haemost. 16, 500–507 (2018).
- 42.Vujkovic-Cvijin I., Swainson L. A., Chu S. N., Ortiz A. M., Santee C. A., Petriello A., Dunham R. M., Fadrosh D. W., Lin D. L., Faruqi A. A., Huang Y., Apetrei C., Pandrea I., Hecht F. M., Pilcher C. D., Klatt N. R., Brenchley J. M., Lynch S. V., McCune J. M., Gut-resident Lactobacillus abundance associates with IDO1 inhibition and Th17 dynamics in SIV-infected macaques. Cell Rep. 13, 1589–1597 (2015).
- 43.Chen C. M., Wu C. C., Huang C. L., Chang M. Y., Cheng S. H., Lin C. T., Tsai Y. C., Lactobacillus plantarum PS128 promotes intestinal motility, mucin production, and serotonin signaling in mice. Probiotics Antimicrob. Proteins 14, 535–545 (2022).
- 44.Spohn S. N., Bianco F., Scott R. B., Keenan C. M., Linton A. A., O’Neill C. H., Bonora E., Dicay M., Lavoie B., Wilcox R. L., MacNaughton W. K., De Giorgio R., Sharkey K. A., Mawe G. M., Protective actions of epithelial 5-hydroxytryptamine 4 receptors in normal and inflamed colon. Gastroenterology 151, 933–944.e3 (2016).
- 45.Chang L., Sultan S., Lembo A., Verne G. N., Smalley W., Heidelbaugh J. J., AGA clinical practice guideline on the pharmacological management of irritable bowel syndrome with constipation. Gastroenterology 163, 118–136 (2022).
- 46.Madia V. N., Messore A., Saccoliti F., Tudino V., De Leo A., De Vita D., Bortolami M., Scipione L., Pindinello I., Costi R., Di Santo R., Tegaserod for the treatment of irritable bowel syndrome. Antiinflamm. Antiallergy Agents Med. Chem. 19, 342–369 (2020).
- 47.Sayuk G. S., Tack J., Tegaserod: What’s old is new again. Clin. Gastroenterol. Hepatol. 20, 2175–2184.e19 (2022).
- 48.Shah E. D., Lacy B. E., Chey W. D., Chang L., Brenner D. M., Tegaserod for irritable bowel syndrome with constipation in women younger than 65 years without cardiovascular disease: Pooled analyses of 4 controlled trials. Am. J. Gastroenterol. 116, 1601–1611 (2021).
- 49.Stakenborg N., Labeeuw E., Gomez-Pinilla P. J., De Schepper S., Aerts R., Goverse G., Farro G., Appeltans I., Meroni E., Stakenborg M., Viola M. F., Gonzalez-Dominguez E., Bosmans G., Alpizar Y. A., Wolthuis A., D’Hoore A., Van Beek K., Verheijden S., Verhaegen M., Derua R., Waelkens E., Moretti M., Gotti C., Augustijns P., Talavera K., Vanden Berghe P., Matteoli G., Boeckxstaens G. E., Preoperative administration of the 5-HT4 receptor agonist prucalopride reduces intestinal inflammation and shortens postoperative ileus via cholinergic enteric neurons. Gut 68, 1406–1416 (2019).
- 50.Stasi C., Sadalla S., Milani S., The relationship between the serotonin metabolism, gut-microbiota and the gut-brain axis. Curr. Drug Metab. 20, 646–655 (2019).
- 51.Peltenburg N. C., Schoeman J. C., Hou J., Mora F., Harms A. C., Lowe S. H., Bierau J., Bakker J. A., Verbon A., Hankemeier T., Boonstra A., Persistent metabolic changes in HIV-infected patients during the first year of combination antiretroviral therapy. Sci. Rep. 8, 16947 (2018).
- 52.Wong A. C., Devason A. S., Umana I. C., Cox T. O., Dohnalová L., Litichevskiy L., Perla J., Lundgren P., Etwebi Z., Izzo L. T., Kim J., Tetlak M., Descamps H. C., Park S. L., Wisser S., McKnight A. D., Pardy R. D., Kim J., Blank N., Patel S., Thum K., Mason S., Beltra J. C., Michieletto M. F., Ngiow S. F., Miller B. M., Liou M. J., Madhu B., Dmitrieva-Posocco O., Huber A. S., Hewins P., Petucci C., Chu C. P., Baraniecki-Zwil G., Giron L. B., Baxter A. E., Greenplate A. R., Kearns C., Montone K., Litzky L. A., Feldman M., Henao-Mejia J., Striepen B., Ramage H., Jurado K. A., Wellen K. E., O’Doherty U., Abdel-Mohsen M., Landay A. L., Keshavarzian A., Henrich T. J., Deeks S. G., Peluso M. J., Meyer N. J., Wherry E. J., Abramoff B. A., Cherry S., Thaiss C. A., Levy M., Serotonin reduction in post-acute sequelae of viral infection. Cell 186, 4851–4867.e20 (2023).
- 53.Hung L. Y., Alves N. D., Del Colle A., Talati A., Najjar S. A., Bouchard V., Gillet V., Tong Y., Huang Z., Browning K. N., Hua J., Liu Y., Woodruff J. O., Juarez D., Medina M., Posner J., Tonello R., Yalcinkaya N., Israelyan N., Ringel R., Yang L., Leong K. W., Yang M., Sze J. Y., Savidge T., Gingrich J., Shulman R. J., Gershon M. D., Ouellet A., Takser L., Ansorge M. S., Margolis K. G., Intestinal epithelial serotonin as a novel target for treating disorders of gut-brain interaction and mood. Gastroenterology 168, 754–768 (2025).
- 54.Edinoff A. N., Akuly H. A., Hanna T. A., Ochoa C. O., Patti S. J., Ghaffar Y. A., Kaye A. D., Viswanath O., Urits I., Boyer A. G., Cornett E. M., Kaye A. M., Selective serotonin reuptake inhibitors and adverse effects: A narrative review. Neurol. Int. 13, 387–401 (2021).
- 55.Ghare S., Singhal R., Bryant V., Gautam S., Tirumala C. C., Srisailam P. K., Reyes-Vega A., Ghooray D., McClain C. J., Hoffman K., Petrosino J., Bryant K., Govind V., Cohen R., Cook R. L., Barve S., Age-associated gut dysbiosis, marked by loss of butyrogenic potential, correlates with altered plasma tryptophan metabolites in older people living with HIV. J. Acquir. Immune Defic. Syndr. 89, S56–S64 (2022).
- 56.Tagawa R., Kawano Y., Minami A., Nishiumi S., Yano Y., Yoshida M., Kodama Y., β-hydroxybutyrate protects hepatocytes against endoplasmic reticulum stress in a sirtuin 1-independent manner. Arch. Biochem. Biophys. 663, 220–227 (2019).
- 57.Rutsch A., Kantsjö J. B., Ronchi F., The gut-brain axis: How microbiota and host inflammasome influence brain physiology and pathology. Front. Immunol. 11, 604179 (2020).
- 58.Iacob S. A., Iacob D. G., Non-alcoholic fatty liver disease in HIV/HBV patients—A metabolic imbalance aggravated by antiretroviral therapy and perpetuated by the hepatokine/adipokine axis breakdown. Front. Endocrinol. 13, 814209 (2022).
- 59.Te H. S., Cholestasis in HIV-infected patients. Clin. Liver Dis. 8, 213–228 (2004).
- 60.Otto A. O., Rivera C. G., Zeuli J. D., Temesgen Z., Hepatotoxicity of contemporary antiretroviral drugs: A review and evaluation of published clinical data. Cells 10, 1263 (2021).
- 61.Wang B., Abbott L., Childs K., Taylor C., Agarwal K., Cormack I., Miquel R., Suddle A., Dolutegravir-induced liver injury leading to sub-acute liver failure requiring transplantation: A case report and review of literature. Int. J. STD AIDS 29, 414–417 (2018).
- 62.Neff G. W., Jayaweera D., Sherman K. E., Drug-induced liver injury in HIV patients. Gastroenterol. Hepatol. 2, 430–437 (2006).
- 63.Rivero A., Mira J. A., Pineda J. A., Liver toxicity induced by non-nucleoside reverse transcriptase inhibitors. J. Antimicrob. Chemother. 59, 342–346 (2007).
- 64.Benedicto A. M., Fuster-Martínez I., Tosca J., Esplugues J. V., Blas-García A., Apostolova N., NNRTI and liver damage: Evidence of their association and the mechanisms involved. Cells 10, 1687 (2021).
- 65.Rossotti R., Maggioni M., Merli M., Orcese C., Iavarone M., Puoti M., Severe cholestatic hepatitis related to abacavir/lamivudine/dolutegravir antiretroviral treatment in a HIV-1 infected subject. AIDS 32, 1727–1729 (2018).
- 66.Christensen E. S., Jain R., Roxby A. C., Abacavir/dolutegravir/lamivudine (Triumeq)-induced liver toxicity in a human immunodeficiency virus-infected patient. Open Forum Infect. Dis. 4, ofx122 (2017).
- 67.Jeyarajan A. J., Chung R. T., Insights into the pathophysiology of liver disease in HCV/HIV: Does it end with HCV cure? J. Infect. Dis. 222, S802–S813 (2020).
- 68.Lindor K. D., Dickson E. R., Baldus W. P., Jorgensen R. A., Ludwig J., Murtaugh P. A., Harrison J. M., Wiesner R. H., Anderson M. L., Lange S. M., Lesage G., Rossi S. S., Hofmann A. F., Ursodeoxycholic acid in the treatment of primary biliary cirrhosis. Gastroenterology 106, 1284–1290 (1994).
- 69.European Association for the Study of the Liver , EASL Clinical Practice Guidelines: Management of cholestatic liver diseases. J. Hepatol. 51, 237–267 (2009).
- 70.Hirschfield G. M., Heathcote E. J., Gershwin M. E., Pathogenesis of cholestatic liver disease and therapeutic approaches. Gastroenterology 139, 1481–1496 (2010).
- 71.Kowdley K. V., Luketic V., Chapman R., Hirschfield G. M., Poupon R., Schramm C., Vincent C., Rust C., Parés A., Mason A., Marschall H. U., Shapiro D., Adorini L., Sciacca C., Beecher-Jones T., Böhm O., Pencek R., Jones D., A randomized trial of obeticholic acid monotherapy in patients with primary biliary cholangitis. Hepatology 67, 1890–1902 (2018).
- 72.Ishizaka A., Koga M., Mizutani T., Parbie P. K., Prawisuda D., Yusa N., Sedohara A., Kikuchi T., Ikeuchi K., Adachi E., Koibuchi T., Furukawa Y., Tojo A., Imoto S., Suzuki Y., Tsutsumi T., Kiyono H., Matano T., Yotsuyanagi H., Unique gut microbiome in HIV patients on antiretroviral therapy (ART) suggests association with chronic inflammation. Microbiol. Spectr. 9, e0070821 (2021).
- 73.Monaco C. L., Gootenberg D. B., Zhao G., Handley S. A., Ghebremichael M. S., Lim E. S., Lankowski A., Baldridge M. T., Wilen C. B., Flagg M., Norman J. M., Keller B. C., Luévano J. M., Wang D., Boum Y., Martin J. N., Hunt P. W., Bangsberg D. R., Siedner M. J., Kwon D. S., Virgin H. W., Altered virome and bacterial microbiome in human immunodeficiency virus-associated acquired immunodeficiency syndrome. Cell Host Microbe 19, 311–322 (2016).
- 74.Guo P., Zhang K., Ma X., He P., Clostridium species as probiotics: Potentials and challenges. J. Anim. Sci. Biotechnol. 11, 24 (2020).
- 75.Mboumba Bouassa R.-S., Comeau E., Alexandrova Y., Pagliuzza A., Yero A., Samarani S., Needham J., Singer J., Lee T., Bobeuf F., Vertzagias C., Sebastiani G., Margolese S., Mandarino E., Klein M. B., Lebouché B., Routy J. P., Chomont N., Costiniuk C. T., Jenabian M.-A., Effects of oral cannabinoids on systemic inflammation and viral reservoir markers in people with HIV on antiretroviral therapy: Results of the CTN PT028 pilot clinical trial. Cells 12, 1811 (2023).
- 76.Qiao Y.-N., Li L., Hu S.-H., Yang Y.-X., Ma Z.-Z., Huang L., An Y.-P., Yuan Y.-Y., Lin Y., Xu W., Li Y., Lin P.-C., Cao J., Zhao J. Y., Zhao S.-M., Ketogenic diet-produced β-hydroxybutyric acid accumulates brain GABA and increases GABA/glutamate ratio to inhibit epilepsy. Cell Discov. 10, 17 (2024).
- 77.Rothhammer V., Mascanfroni I. D., Bunse L., Takenaka M. C., Kenison J. E., Mayo L., Chao C. C., Patel B., Yan R., Blain M., Alvarez J. I., Kébir H., Anandasabapathy N., Izquierdo G., Jung S., Obholzer N., Pochet N., Clish C. B., Prinz M., Prat A., Antel J., Quintana F. J., Type I interferons and microbial metabolites of tryptophan modulate astrocyte activity and central nervous system inflammation via the aryl hydrocarbon receptor. Nat. Med. 22, 586–597 (2016).
- 78.Cannon A. S., Nagarkatti P. S., Nagarkatti M., Targeting AhR as a novel therapeutic modality against inflammatory diseases. Int. J. Mol. Sci. 23, 288 (2021).
- 79.Tian Y., Rimal B., Bisanz J. E., Gui W., Wolfe T. M., Koo I., Murray I. A., Nettleford S. K., Yokoyama S., Dong F., Koshkin S., Prabhu K. S., Turnbaugh P. J., Walk S. T., Perdew G. H., Patterson A. D., Effects of early life exposures to the aryl hydrocarbon receptor ligand TCDF on gut microbiota and host metabolic homeostasis in C57BL/6J mice. Environ. Health Perspect. 132, 87005 (2024).
- 80.Boggs D. L., Nguyen J. D., Morgenson D., Taffe M. A., Ranganathan M., Clinical and preclinical evidence for functional interactions of cannabidiol and Δ9-tetrahydrocannabinol. Neuropsychopharmacology 43, 142–154 (2018).