Bioenergetics and neuroimaging research: a neuropathophysiological linkage in the setting of cocaine use amongst persons with HIV

Despite innovations in antiretroviral therapy (ART) that have transformed HIV infection from an acute illness with high mortality risk into a chronic, largely manageable disease, the viral reservoir that persists in brain continues to pose a risk for neurocognitive impairment and other deleterious clinical outcomes. ART regimens can inhibit viral integration and suppress replication to nondetectable levels in plasma and cerebrospinal fluid (CSF) but do not eliminate viral reservoirs, including that in brain [1]. Moreover, HIV transcripts within CSF cells have been associated with brain injury despite suppressive ART [2]. Comorbid HIV and cocaine use exacerbates brain atrophy and neurocognitive decline despite viral suppression [3–5]. Intersecting factors disrupted by chronic cocaine use among people with HIV (PWH) contribute to HIV-associated neuropathology, including neurotransmitter signaling (particularly dopamine), neuroinflammation, blood–brain–barrier (BBB) integrity, and energy metabolism. Further, the neuropathological severity associated with HIV and cocaine is spatially heterogenous [6–8]. The healthy brain is energetically expensive and complex with region-specific, unique functional roles [9–11]. Further, compartmentalization of HIV infection in brain contributes to this heterogeneity [12]. Mechanistic links between HIV and cocaine require additional characterization to assess region-dependent contributions to develop therapeutic interventions for cocaine use disorder comorbid with HIV.

Mamidi et al.[13] focused on associations between chronic cocaine use and HIV on glucose uptake. Using ¹⁸F-FDG PET/CT in a 2 × 2 experimental design with HIV (present/absent) and cocaine (present/absent) (N = 63), they showed the lowest uptake with both HIV and cocaine. One factor – HIV or cocaine – showed intermediate uptake, and neither factor showed the highest uptake. The pronounced impact of cocaine on HIV-associated neuropathology is, in part, due to disruption of dopaminergic neurotransmission. The dopamine system is linked to inflammation and immunological function. Brain regions with high basal dopamine levels, such as the striatum and substantia nigra, are amongst the most vulnerable to HIV [14]. Dopamine exposure to human macrophages results in elevated production of pro-inflammatory cytokines and chemokines [14]. Acutely, elevated dopamine concentrations due to cocaine use increase oxidative stress, exacerbated by Tat [15,16]. Chronically, cocaine use is associated with dopamine depletion, demonstrated by PET scanning research [17]. In addition, HIV itself is associated with dopamine depletion as well as neurocognitive impairment and depression [18] not investigated here. This constellation suggests a synergistic effect of HIV and cocaine on dopaminergic transmission. To the extent that dopaminergic neurotransmission impacts glucose uptake, only additive effects of HIV and cocaine were reported here. No interaction of HIV and cocaine was observed.

A major hallmark of chronic HIV is elevated pro-inflammatory cytokine and chemokine production. Suppressed PWH still have elevated neuroinflammation in the parietal and occipital cortex and the globus pallidus. Neuroinflammation is associated with decreased neurocognitive performance and increased white matter damage supported by a PET study with [¹¹C] PBR28 and neuropsychological testing [8]. Viral proteins, Tat and gp120, both facilitate the production of pro-inflammatory cytokine and chemokines that decrease BBB tight junction protein expression and are directly neurotoxic [19]. Loss of BBB integrity allows free virus and HIV-infected monocytes to enter brain, exacerbating neuronal damage [20]. Similarly, cocaine increases neuroinflammatory markers by activating microglial cells and disrupting BBB integrity – decreasing tight junction protein expression in human pericytes [21]. When measuring chronic cocaine-induced microglial activation in vivo, rhesus macaques displayed increased TSPO PET expression in dopamine-rich regions via [³H] PK-11195 [6,7]. However, humans with a history of chronic cocaine use assessed with TSPO PET via [¹¹C] PBR28 displayed no significant changes [6,22]. Of note, increased TSPO expression using current tracers does not distinguish between microglial and astrocytic activation. Further, there are other limitations with the utility of both PK-11195 and PBR28 tracers. Hence, PET scanning studies are currently inconclusive, though studies using other methodologies support neuroinflammatory effects associated with cocaine. Cocaine has been linked with increased TNF-α expression and is well known to stimulate HIV replication through induction of NF-κB and activation of transcription through the HIV LTR. The increased expression of TNF-α induced by HIV might exacerbate that by cocaine. Pro-inflammatory cytokine production has been associated with dopamine depletion outside of HIV infection. This suggests an intrinsic link between chronic HIV despite suppression, ongoing neuroinflammation, and persistent dopamine depletion, which is associated with depression and neurocognitive symptoms. This linkage may also reflect the results reported here and suggests the possible clinical utility of TNF-α inhibitors and dopaminergic agonists for the treatment of depressive and neurocognitive symptoms in virally suppressed PWH, supporting normalization of brain glucose uptake.

In adults, the brain's immense energetic demands require roughly 20% of all glucose and constitute approximately the same proportion of total oxygen consumption during resting conditions [23,24]. Maintenance of brain metabolic homeostasis is particularly sensitive to metabolic coupling between types of brain cells that contribute to clinical disorders when disrupted [25–27]. Viral–host interactions after an infection like HIV shift bioenergetics for incompletely understood reasons. Changes in energetic metabolism have been reported to occur in vitro using cultured astrocytes, neurons, and microglia due to Tat and gp120 [28–30], cytokines and chemokines [31], oxidative stress [32], and ART [33]. In vivo, virally suppressed PWH display decreased glucose uptake in the frontal cortex and the anterior cingulate cortex via FDG-PET [34,35]. Altogether, these changes suggest a shift from metabolism of primarily glucose to other oxidative substrates. Moreover, in vitro, cocaine is associated with a similar metabolic shift [29].

As suggested above, energetic demands vary across brain regions. Recent studies suggest that the brain also uses other substrates, such as fatty acids, lactate, pyruvate, glutamate, glutamine, and ketone bodies, more frequently than previously considered [10,11,36]. The composition of substrates used may shift under various factors such as age, diet, brain activity or injury, cognitive reserve, and the presence of viral infections like HIV [26,37]. Hence, future studies should expand from the general study of glucose uptake as the primary substrate to other substrates and associated changes in oxidative stress and mitochondrial function. Clinical research suggests the importance of associated interacting comorbidities, such as cardiovascular disease, with HIV [38] and cocaine [39]. It should be noted that age, ethnicity, and education and concomitant opioid use were not able to be separately analyzed here. Of note, older age is also associated with dopamine depletion, suggesting a more prominent effect amongst older PWH.

In addition to future studies examining other energy substrate outcome measures; improved control of extraneous factors; and integration of clinical outcomes of cocaine use among PWH, neuroimaging studies can be particularly helpful in examining spatial heterogeneity in energetic effects induced by toxic HIV protein and transcript burden as well as pro-inflammatory cytokine secretion associated with cocaine use. Yet, these methods incompletely capture metabolic changes in brain. It can be concluded that there remains much to explore as to how the bioenergetic shifts occurring due to HIV and cocaine may be mechanistically linked to clinical outcomes.

Acknowledgements

Conflicts of interest

There are no conflicts of interest.