Extracellular Vesicles and HIV-Associated Neurocognitive Disorders: Implications in Neuropathogenesis and Disease Diagnosis

Abstract

Extracellular vesicles are heterogeneous cell-derived membranous structures of nanometer size that carry diverse cargoes including nucleic acids, proteins, and lipids. Their secretion into the extracellular space and delivery of their cargo to recipient cells can alter cellular function and intracellular communication. In this review, we summarize the role of extracellular vesicles in the disease pathogenesis of HIV-associated neurocognitive disorder (HAND) by focusing on their role in viral entry, neuroinflammation and neuronal degeneration. We also discuss the potential role of extracellular vesicles as biomarkers of HAND. Together, this review aims to convey the importance of extracellular vesicles in the pathogenesis of HAND and foster interest in their role in neuroinflammatory diseases.

1. Introduction

Human Immunodeficiency Virus (HIV) infection continues to affect the global population, with approximately 37.5 million individuals infected worldwide, and 1.5 million new infections in 2020 alone (UNAIDS 2021). Although the development of combined antiretroviral therapy (cART) has significantly decreased mortality and morbidity resulting from HIV infection, 50–60% of people living with HIV (PLH) may develop a spectrum of conditions involving cognitive, motor, and behavioral abnormalities (Alford and Vera 2018; Heaton et al. 2011; Simioni et al. 2010). Collectively termed HIV-associated neurocognitive disorder (HAND), these deficiencies, which can be mild or severe, negatively affect everyday function and pose a significant medical cost for both the individual and healthcare systems overall (Cysique et al. 2011; Hogan and Wilkins 2011).

Neuroanatomically, HAND is characterized by cell death in subcortical regions including the basal ganglia and hippocampus, as well as thinning of the cerebral cortex (Albright et al. 2003; Everall et al. 2005; McArthur 2004). Moreover, these regions of the forebrain appear to be selectively vulnerable to pathological injury of synapses, namely shrinkage/retraction of the neurites, short-segment branching and decreased numbers of dendritic spines, axonal injury, and overall synaptodendritic atrophy (Avdoshina et al. 2017; Ellis et al. 2007; Masliah et al. 1997; Spudich and Gonzalez-Scarano 2012; Wenzel et al. 2019a). However, the pathological causes of HAND are still a subject of intense investigation. Thus, further characterization and understanding of the mechanisms that promote the pathogenesis of HIV within the central nervous system (CNS) may lead to new treatments for this disorder.

Recently, extracellular vesicles (EVs) have gained importance in human physiology due to their ability to package and deliver cargoes (nucleic acids, lipids, proteins) and molecular signals, thereby facilitating cell-to-cell communication and tissue homeostasis (van Niel et al. 2018; Yáñez-Mó et al. 2015), or to promote the pathogenesis in numerous diseases (Caobi et al. 2020; Rezaie et al. 2021; Sampey et al. 2014). Lately, more focus has been directed toward uncovering the role of EVs in HIV infection in the body and the CNS. This review will discuss the role of EVs in the neuroinflammatory aspects of HAND by overviewing their contribution in the process of HIV entry and subsequent pathogenesis within the CNS. We will also discuss how EVs may be used in the disease diagnosis of HAND. The overarching goal is to emphasize the importance of this growing field, and to create an interest in EVs and their role in both HAND and other neuroinflammatory diseases.

2. Overview of Extracellular Vesicles in the Brain

EVs are typically categorized into three subtypes based on their differences in size, structure, and formation (Figure 1). These include exosomes, microvesicles and apoptotic bodies.

Figure 1. Biogenesis and characteristic markers of extracellular vesicles.

EVs can be classified as exosomes, microvesicles, or apoptotic bodies. Each has a unique pathway of biogenesis, subsequent markers for characterization, and packaged cargo. Exosomes are the smallest of extracellular vesicles (30–150 nm) and are created through the formation of multivesicular bodies within the endosomal pathway. Fusion of these multivesicular bodies with the plasma membrane releases exosomes into the extracellular space. Microvesicles are 150–1000 nm extracellular vesicles formed through outward budding of the plasma membrane. This process often involves cholesterol and/or lipid rafts. The biogenesis of apoptotic bodies occurs during programed cell death, leading to large 500–5000 nm vesicular structures. Subsequent fusion Created with BioRender.com

2.1. Exosomes.

Exosomes are the smallest and most well characterized EVs, with a size ranging from 30–150 nm. They are intraluminal vesicles (ILVs) formed by membrane invaginations of multivesicular endosomal bodies (MVBs) in the late endosomal pathway (Colombo et al. 2013; Colombo et al. 2014). When MVBs are transported to the plasma membrane, fusion occurs and ILVs are released into the extracellular space and thus referred to as exosomes (van Niel et al. 2018). The release process requires microtubules and associated molecular motors (kinesins and myosins), molecular switches (small GTPases), and fusion machinery (SNAREs and tethering factors) (Cai et al. 2007). Under homeostatic conditions, exosomes can be released by diverse cell types in the periphery as well as the CNS. These include neurons, astrocytes, oligodendrocytes and microglia.

In addition to size, exosomes are identified by common markers, including CD63, CD9, HSP70, ALIX, and others (Théry et al. 2001). ALG-2-interacting protein X (ALIX) is a protein associated with endosomal sorting, and thus it is incorporated during exosome biogenesis (Willms et al. 2016). CD63 and CD9 are two members of the tetraspanin superfamily of membrane-spanning proteins that are involved in several biological processes including cell adhesion, motility, membrane fusion, and membrane protein organization. They are often incorporated into the membrane of exosomes and contain functional microdomains that facilitate cargo packaging and exosome secretion (Perez-Hernandez et al. 2013). Heat shock protein 70 (HSP70) is a chaperone protein involved in cellular protein homeostasis and regulates the cellular response under conditions of stress by modulating protein degradation and stability.

In terms of cargo, exosomes can package proteins, lipids and nucleic acids, and a large variety of small non-coding RNA species. The genetic cargo of exosomes can be translated into proteins by target cells, thus, when released, exosomes play a range of roles in synaptic transmission, neurite outgrowth, trophic support, inflammation, and blood brain barrier integrity (Saeedi et al. 2019). In addition, several neuropathogenic proteins, such as prions (Fevrier et al. 2004), β-amyloid peptide (Rajendran et al. 2006) and α-synuclein (Emmanouilidou et al. 2010) are released from cells in association with EVs. These secreted vesicles are thought to participate in disseminating pathogenesis through interaction with recipient cells. Importantly, exosomes can be detected in the cerebrospinal fluid. Thus, exosomes can be used as biomarkers for numerous neurodegenerative diseases.

2.2. Microvesicles.

Microvesicles are the second class of EVs, with a characteristic size of 150–1000 nm. These vesicles are formed from outward budding and fission of the plasma membrane of the producer cell (Muralidharan-Chari et al. 2010; Yáñez-Mó et al. 2015). Because of this, microvesicles are usually characterized by size, and by markers of the producer cells. Microvesicles carry many of the same cargo as exosomes including lipids, proteins and nucleic acids (Bolukbasi et al. 2012; Shen et al. 2011; Yang and Gould 2013). Cholesterol is one lipid component abundant in microvesicles and when depleted, decreases their biogenesis. In the CNS, many studies focusing on microvesicles involve the glial cell population. Microglia-derived microvesicles are particularly important as the first line of defense in the case of CNS infection. For example, lipopolysaccharide increases microvesicle release in microglia, whose content is enriched in proinflammatory cytokines and microRNA that drive proinflammatory responses (Jablonski et al. 2016; Kumar et al. 2017; Paolicelli et al. 2019). Moreover, released microvesicles represents an efficient way to modulate the activity of neighboring neurons. For instance, microvesicles secreted from microglia can contain the endocannabinoid N-arachidonylethanolamine, which can activate the type 1 cannabinoid receptor (CB1) expressed by GABAergic neurons (Gabrielli et al. 2015). Additionally, astrocytes secrete microvesicles carrying Fibroblast Growth Factor-2 (FGF-2) and Vascular Endothelial Growth Factor (VEGF), two peptides with neurotrophic properties (Proia et al. 2008; Upadhya et al. 2020), supporting the role that astrocytes play in brain development and function. Indeed, microvesicles can improve functional deficit following brain ischemia (Lee et al. 2016). Thus, microvesicles can alter synaptic transmission, mediate neuron-glia interactions (Antonucci et al. 2012), and play a role in inflammation and disease states (Beneventano et al. 2017; Joshi et al. 2014).

2.3. Apoptotic bodies.

Apoptotic bodies are the last characterized EVs. These vesicles are the largest, found up to 5000 nm in diameter (Hristov et al. 2004). Formation of apoptotic bodies occurs during programed cell death or apoptosis (Akers et al. 2013; Elmore 2007). Thus, the cargo of apoptotic bodies are degraded proteins, segmented organelles, as well as cleaved and condensed DNA. In the CNS, the apoptosis of neurons is an organized process, where toxic mediators and damage is contained or controlled by glial cells in order to maintain a suitable environment for neighboring neurons (Zia et al. 2021).

3. Extracellular vesicles and their role in HIV neuropathogenesis

3.1. General pathway of HIV neuropathogenesis.

There are multifactorial mechanisms leading to HIV neuropathogenesis (Valcour et al. 2011) (Figure 2). Typically, HIV enters the CNS through peripheral monocytes and/or macrophages, which have crossed the blood brain barrier and infiltrated the brain parenchyma (Eugenin et al. 2006; Leon-Rivera et al. 2021). Infectious virions are then spread to the glial population, including microglia and end-feet astrocytes. Microglia themselves can be productively infected with HIV and release mature virions and HIV viral proteins (D’Aversa et al. 2005; Garcia-Mesa et al. 2016; Wallet et al. 2019). Astrocytes can be infected by HIV both in vitro and in vivo (Conant et al. 1994; Lutgen et al. 2020) and become a HIV reservoir (Valdebenito et al. 2021). Once inside the CNS, HIV promotes neuronal damage by a variety of mechanisms through “direct” or “indirect” neurotoxicity (Kaul et al. 2001). Direct toxicity implies that HIV viral proteins including tat, gp120, nef, and others released from infected cells or shed from the virus have the intrinsic ability to induce neurodegeneration in both receptor-dependent and independent manners (Bansal et al. 2000; Campbell et al. 2015; Kim et al. 2008). These include aberrant intracellular calcium (Meucci and Miller 1996; Norman et al. 2008), mitochondria toxicity (Norman et al. 2007; Rozzi et al. 2018), impaired axonal transport (Wenzel et al. 2019b), synaptic simplification (Dickens et al. 2017), loss of dendritic spines (Nath and Steiner 2014), deposition of amyloid-beta plaques (Hategan et al. 2017) and activation of pro-apoptotic pathways (Meucci et al. 1998). These effects, although originally described in neurons in culture, have also been confirmed in animal models of HAND, including gp120 transgenic mice (Toggas et al. 1994), Tat transgenic mice (Kim et al. 2003) or HIV-transgenic rats (Reid et al. 2001). On the contrary, the indirect mechanism(s) relies on a constitutive release of pro-inflammatory factors (e.g. Interleukin 1-β, Tumor Necrosis Factor-α, reactive oxygen species) from glial cells along with a reduction in neurotrophic support (e.g. BDNF, GDNF) all of which contribute to reduced neuronal survival (Bachis et al. 2012; Nath et al. 2012; Nosheny et al. 2004; Walsh et al. 2014). Importantly, EVs are shown to be intricately involved in these neurotoxic pathways, which are discussed below.

Figure 2. The contribution of extracellular vesicles on the neuroinflammatory pathways of HIV.

EVs are released from various cell types including neuroepithelia microglia, astrocytes, and neurons. During HIV infection, these EVs can potentiate the neuroinflammatory processes in the brain, which are highlighted in red text. Neuroepithelia release EVs involved in breakdown of the blood-brain barrier. Astrocytes and microglia release EVs that could damage neurons, either through increased inflammatory response, decrease neurotrophic support, or direct neurotoxicity. Finally, damaged neurons can release neuron-specific EVs that may be used as biomarkers for HAND. Created with BioRender.com

3.2. Extracellular vesicles, the blood brain barrier, and neuroepithelia.

The invasion of infected monocytes and macrophages through the blood brain barrier and into the brain parenchyma is thought to initiate the pathological process of HAND. EVs, in particular exosomes derived from HIV infected cells, can cause disruption in the blood brain barrier, facilitating viral entry into the brain. For example, purified exosomes, produced from latently infected peripheral immune cells (J-Lat 9.2 lymphocytes, U1 monocytes), promote the disruption of mitochondrial dynamics and a loss of e-NOS function in primary human brain microvascular endothelial cells (Chandra et al. 2021). The brain neuroepithelium may also be compromised when HIV infection interacts with other pathogenic molecules, such as amyloid beta (Aβ). The neuroepithelium has been characterized as a structure capable of synthesizing and releasing Aβ. HIV infection potentiates the release of exosomes containing Aβ from brain endothelial cells, which subsequently interact with surrounding astrocytes and pericytes, the integral support structures of the blood brain barrier. This phenomenon leads to a disruption of blood brain barrier integrity, resulting in the infiltration of peripheral Aβ and other neurotoxins (András et al. 2017). Indeed, whole proteomic analysis confirms that HIV and Aβ interact to change the EV composition of human brain microvascular endothelia cells, with an increase in proteins involved in exocytosis, vesicle formation, and immune activation (András et al. 2020b). Finally, the detrimental changes of the neuroepithelium caused by EVs, Aβ, and HIV may alter the growth and maturation of neural progenitor cells (Cho et al. 2021), promoting premature differentiation and activation of the inflammasome within the exposed cells (András et al. 2020a). Overall, data suggest that EVs, specifically exosomes, can mediate maladaptive alterations to the brain microvascular system, which may potentiate the transmigration of HIV and other pathogens into the brain.

3.3. Extracellular vesicles and the indirect pathway of HIV neurotoxicity.

The “indirect” mechanism hypothesis of HAND suggests that once glial cell populations are infected by HIV, they contribute to neuronal loss by releasing pro-inflammatory cytokines/chemokines and/or decreasing neurotrophic support (Kaul et al. 2001). Experimental evidence confirms that brain EVs enhance the release and subsequent modulation of neuroinflammatory processes by HIV infection. For example, monocyte derived macrophages (MDM) that are infected with HIV show an increased release of EVs containing cathepsin B, a lysosomal protease that is upregulated in postmortem brains from individuals with HAND (Cantres-Rosario et al. 2015; Rodriguez-Franco et al. 2012). Direct exposure of neuronal cultures to cathepsin B leads to increased caspase-3 cleavage and a loss of synaptophysin, which suggests neuronal toxicity or impairment (Cantres-Rosario et al. 2019). Cathepsin B positive exosomes, derived from HIV+ MDM, also promote neuronal toxicity, thereby implicating exosome biogenesis as an alternative mechanism for cathepsin B release (Cantres-Rosario et al. 2019).

3.4. Extracellular vesicles and microRNAs.

EVs can also mediate cross-talk between “infected” astrocytes and microglia through the release of microRNAs. These are small non-coding RNAs that are abundant in the CNS and act as post-transcriptional regulators of gene expression important for brain development (Fineberg et al. 2009). However, some microRNAs promote neuroinflammation and are involved in several neurological diseases, including Parkinson’s and Alzheimer’s diseases, and Schizophrenia (Soria et al. 2017). For instance, the HIV protein Tat, which regulates HIV transcription, simulates astrocytomas to release EVs containing microRNA-9 (miR-9). The EVs released can be directly internalized by microglia, which, because of miR-9, is transformed to a migratory phenotype, potentially promoting movement to brain areas with active Tat release (Yang et al. 2018). Considering that Tat protein can be actively produced in the CNS despite cART, astrocyte derived EVs may potentiate inflammation through microglia recruitment. The interactions of Tat, astrocytes, and microRNAs are also exemplified in studies showing that Tat promotes the packaging and EV-mediated release of miR-7 from stimulated human primary astrocytes (Hu et al. 2020). miR-7, is a non-coding RNA that participates in the pathology of neurodegeneration (Je and Kim 2017; McMillan et al. 2017; Yue et al. 2020). The delivery of miR-7 to neuronal cultures results in decrease synaptic and dendritic proteins (e.g. PSD96, GAD65, vGlut) (Hu et al. 2020). A reduction of these neuronal markers suggests a breakdown in neuronal architecture

Another microRNA involved with synaptic impairment is miR-29b. This is a small non-coding RNA associated with neuronal growth and maturation (Ma et al. 2020; Napoli et al. 2020; Swahari et al. 2021). miR-29b levels are increased in simian immunodeficiency virus (SIV) infected non-human primates following morphine treatment (Hu et al. 2012). Moreover, in vitro experiments revealed that Tat and morphine stimulate the packaging of miR-29b in astrocytic exosomes (Hu et al. 2012). Together, evidence points to EVs as mediators of numerous inflammatory processes resulting from HIV infection, with an overall effect in glial cell migration, decreased trophic support, and neuronal dysfunction.

3.5. Extracellular vesicles and the direct pathway of neurotoxicity.

HIV viral proteins including Tat, gp120, and nef promote direct neurotoxicity and cell death. Some of these proteins can be package into exosomes as reviewed elsewhere (Patters and Kumar 2018). The Tat and nef proteins have garnered interest due to their association and/or packaging into EVs. Tat has been detected in exosomes secreted from astrocytes (Rahimian and He 2016). Nef is a 27–35 kDa protein which supports viral outgrowth by inhibiting host inflammatory processes and can cause neuronal toxicity through pathways involving IP-10 or cytoskeletal remodeling (Del Río-Iñiguez et al. 2018; Stumptner-Cuvelette et al. 2001; Tan et al. 2013; van Marle et al. 2004). Glial cells can package and release exosomes containing nef. In one study, a transfection of a nef-GFP plasmid was performed in 3 cell types- CHME-5 (microglia), primary astrocytes, and SH-SY5Y neuroblastoma cells, and ALIX positive exosomes were collected. However, only exosomes from CHME-5 microglia successfully packaged nef, which were then shown to disrupt blood brain barrier integrity (Raymond et al. 2016). Contrary to this, another study that used an adenovirus expressing nef to transduce primary human astrocytes, showed that exosomes express nef intracellularly (Sami Saribas et al. 2017). The astrocyte-derived EVs containing nef can be taken up by neurons and cause oxidative stress and neuronal dysfunction, including axonal and dendritic degeneration, upregulation of tau, and impaired action potential (Sami Saribas et al. 2017). While more studies are needed to confirm that a variety of CNS cells release EV-containing viral proteins in vivo, production of neurotoxic EVs have a clinical significance because it may explain neuroinflammation, neurodegeneration, and HAND.

4. Extracellular vesicles as biomarkers of neurocognitive impairment in HAND

As the ability to characterize, purify, and manipulate EVs evolve, so has their potential to be used in the diagnosis of HIV-mediated neuropathologies. For example, a study probed EVs derived from the cerebrospinal fluid (CSF) from HIV+ subjects with or without neurocognitive impairment and found an increased concentration of EVs in the sample group that had clinical manifestations of HAND. Further proteomic analysis of these EVs showed increased markers for glial activation, inflammation, stress responses, and others (Guha et al. 2019). Another group examined EVs prepared from plasma, specifically focusing on neuronal EVs which can be identified and sequestered using neuronal cell adhesion molecule L1 (L1CAM). When comparing samples from neuropsychologically normal (NPN) to neuropsychologically impaired (NPI) HIV+ subjects, they found that L1CAM EVs in the NPI group expressed significantly elevated levels of high mobility group box 1 (HMGB1), neurofilament light, and amyloid beta, which are all associated with neuronal degeneration (Pulliam et al. 2019). The use of EVs as a potential biomarker has also been discussed in comorbid conditions such as substance abuse and HIV infection. In one study examining tobacco and alcohol use in PLH, plasma levels of L1CAM EVs were elevated in cigarette smokers. However, HIV infection alone had no effect on the elevation of L1CAM EVs. Incidentally, EVs containing astrocytic marker glial fibrillary acidic protein were elevated in samples from PLH with a history of alcohol use (Kodidela et al. 2020). Some of the previous findings can be replicated in small animal models of HAND. The HIV transgenic rat model harbors a non-infectious HIV provirus that produces HIV viral proteins- Tat, env, rev, nef, vif, throughout its lifetime, and is often used to uncover mechanisms of HIV-mediated neurodegeneration. In these animals, aged (12–18 month old) females exhibit an elevation of L1CAM EVs in the brain extracellular space as well as the plasma (Dagur et al. 2020). Thus, this model may be useful to study and discover other potential EV-related biomarkers in a basic lab setting. Overall, EVs, and in particular neuronal derived L1CAM EVs may be a useful biomarker to determine the trajectory of HIV infection within the brain.

5. Conclusions

EVs, membrane-enclosed nanoscale particles, are released by virtually all prokaryotic and eukaryotic cells as part of their normal physiology but also in response to injury. Although initially perceived as “trash” disposed by cells, EVs have gained an important role as key components of an orchestrated cellular response to internal and external cues. Indeed, it appears that EVs play a large role in the neuropathogenesis of several neurodegenerative diseases including HAND. However, there are some gaps in knowledge that may be a focus for future experiments. For instance, in HAND, little is known about the role of microglia-derived EV versus that of neurons. Are microglia-derived EVs necessary to dispose unwanted constituents or do they promote viral spreading, thereby accelerating inflammation and neuronal loss? Therefore, more studies on the implications of microvesicles and apoptotic bodies in HAND would be a welcome addition to the field. Additionally, a standardization of sample type used to identify EVs may be beneficial because, as noted in the biomarkers section, neuronal L1CAM EVs exhibit a tissue specific expression. Nevertheless, this growing field has great potential for diagnostic and therapeutic strategies for people living with HIV (Mahajan et al. 2021).