Cross-talk between lipid homeostasis and endoplasmic reticulum stress in neurodegeneration: insights for HIV-1 associated neurocognitive disorders (HAND)

Abstract

The dysregulation of lipid homeostasis is emerging as a hallmark of many CNS diseases. As aberrant protein regulation is suggested to be a shared pathological feature amongst many neurodegenerative conditions, such as Alzheimer’s disease (AD), Parkinson’s disease (PD), and Huntington’s disease (HD), disruptions in neuronal lipid processing may contribute to disease progression in the CNS. Specifically, given the endoplasmic reticulum (ER) dual role in lipid homeostasis as well as protein quality control (PQC) via unfolded protein response (UPR), lipid dysregulation in the CNS may converge on ER functioning and constitute a crucial mechanism underlying aberrant protein aggregation. In the current review, we discuss the diverse roles of lipid species as essential components of the CNS. Moreover, given the importance of both lipid dysregulation and protein aggregation in pathology of CNS diseases, we attempt to assess the potential downstream cross-talk between lipid dysregulation and ER dependent PQC mechanisms, with special focus on HIV-associated neurodegenerative disorders (HAND).

Introduction:

As one of the most abundant cellular molecules, lipids participate in a diverse range of processes, from functioning as major components of the cell membrane to serving as signaling molecules, energy stores and modulators of stress responses (Yeagle, 1989). These macromolecules are notable for their highly variable molecular patterns that can form more complex structures by joining together or binding to the other organic molecules (Spector and Yorek, 1985). Given their diverse cellular functions, it is crucial that the content and composition of these molecules be monitored to preserve homeostasis and prevent toxic accumulation of lipid metabolites (lipotoxic) in non-adipose tissues (Van Herpen and Schrauwen-Hinderling, 2008; Weinberg, 2006).

Although lipids can be modified within multiple compartments of a cell, the endoplasmic reticulum (ER) plays a prominent yet distinct role in lipid metabolism. The majority of lipids, such as membrane lipids, cholesterol, and neutral lipids are synthesized by the ER (Fagone and Jackowski, 2009; Lodhi and Semenkovich, 2014). By sensing and responding to lipid fluctuations, as well as preventing the accumulation of lipotoxins in cellular compartments, the ER is essential for maintaining lipid and cellular homeostasis (Han and Kaufman, 2016; Stevenson et al., 2016). In addition to its roles in lipid homeostasis, the ER maintains unique transmembrane proteins that respond to imbalances between proteins and chaperones (so called ER stress), in a process termed the ER unfolded protein response (UPR) (Volmer and Ron, 2015). The correlation between ER stress with lipid metabolism has been discussed as a crucial factor in pathogenic processes amongst peripheral tissues (Basseri and Austin, 2012; Han and Kaufman, 2016; Ho et al., 2018).

In utilizing the same stress response mechanism for both lipids and proteins, the ER may modulate the pathogenesis of neurodegenerative diseases by facilitating cross-talk between lipid homeostasis and protein quality control (PQC) systems. Along with impaired PQC, aberrant lipid metabolism and activated UPR are hallmarks of disease progression in many neurodegenerative disorders, including Alzheimer’s disease (AD), Parkinson’s disease (PD), Huntington’s disease (HD) and amyotrophic lateral sclerosis (ALS) (Hetz and Saxena, 2017; Scheper and Hoozemans, 2015). Likewise, other cellular stresses, including viral infection, are known to modulate UPR as well as the dysregulation of lipid homeostasis (Chan, 2014).

In this review, we discuss the mechanisms of lipid homeostasis in the CNS as well as their importance in the context of neuropathogenesis. Specifically, we focus on the association between disrupted lipid homeostasis and disturbances in ER-dependent PQC (i.e. UPR), suggesting lipid dysregulation could constitute a crucial mediation mechanism underlying aberrant protein aggregation amongst many neurodegenerative diseases. In particular, we highlight HIV-associated lipid dysregulation which can potentially modulate UPR activation and ultimately dysregulate neuronal homeostasis.

Interplay between the ER stress and lipid homeostasis

The endoplasmic reticulum (ER) is a multifunctional organelle that maintains neuronal homeostasis by regulating calcium balance, lipid metabolism, as well as protein trafficking and processing (Baumann and Walz, 2001; Berridge, 2002; Schwarz and Blower, 2016; Subramanian and Meyer, 1997). ER homeostasis can be disturbed by a number of genetic and environmental factors, including pharmacological perturbation, viral infection, intracellular alterations in Ca²⁺, as well as alterations in lipid content and composition (Fu et al., 2011; Han and Kaufman, 2016; Jacquemyn et al., 2017; Kaufman, 1999). Furthermore, internal or external stimuli which disturb ER homeostasis are referred to as ER stress, which in turn trigger the unfolded protein response (UPR) to enhance cell survival and proper cell function (Basseri and Austin, 2012; Welihinda et al., 1999).

A growing body of evidence suggests that the UPR or part of its machinery could be involved in several neurodegenerative diseases, including AD, PD and ALS (Hetz and Saxena, 2017; Scheper and Hoozemans, 2015). Moreover, misfolded protein aggregates as well as dysregulated lipid metabolism are implicated in a variety of neurodegenerative diseases (Adibhatla et al., 2006; Adibhatla and Hatcher, 2007; Tracey et al., 2018). Therefore, through disruptions in ER dependent PQC, lipid dysregulation may constitute a crucial mechanism in the pathogenesis of CNS disorders and ensuing behavioral deficits. As indicated previously, the UPR is a signal transduction pathway known to be activated by an imbalance between unfolded/misfolded proteins and chaperones to maintain protein folding homeostasis in the ER (Volmer and Ron, 2015). Recent studies reveal a bidirectional yet dynamic relationship between lipid homeostasis and UPR activation, where activation of ER stress pathways can result from aberrant lipid metabolism and promote lipogenesis (Figure 1) (Basseri and Austin, 2012; Gentile et al., 2011; Wei et al., 2006).

Figure 1:

Summary of the bidirectional relationship between lipid metabolism disturbance and ER stress. The schematic shows the accumulation of the most abundant lipid species in the CNS induces ER-stress and ultimately UPR activation. The UPR signaling pathway initiates through three arms of UPR: ATF6, IRE1 α and PERK. The downstream mediators of these transmembrane receptors can dysregulate transcription of genes involved in lipid biosynthesis. ATF6 response to both protein and lipid accumulation, important for activation of fatty acid oxidation pathway and as a transcription factor induce ER stress regulatory genes. Activation of PERK and phosphorylation of eIF2α under lipid species accumulation-induced ER stress enhances CHOP expression which may lead to apoptosis and cell death. The UPR arm IRE-1α activation which has been reported mostly under protein aggregation can enhance lipogenic gene expression. The suggested model of crosstalk between UPR signaling pathway and lipogenesis and reciprocally between lipid accumulation and UPR induction causes neuronal dysfunctionality and CNS inflammation (Created with BioRender.com).

UPR signaling is associated with three important ER-localized transmembrane transducers, specifically inositol-requiring enzyme-1 (IRE1), activating transcription factor 6 (ATF6), and protein kinase RNA-like ER kinase (PERK) (Muneer and Shamsher Khan, 2019). Activation of UPR pathways is indicative of ER stress, and can lead to the release of apoptosis signaling components (Figure 1) (Basseri and Austin, 2012). Acute and severe ER stress can dysregulate UPR pathways, perturb lipid metabolism and consequently induce lipotoxicity in peripheral tissues by disrupting the composition of membrane phospholipids or augmenting ER-induced apoptosis (Han and Kaufman, 2016; Scheper and Hoozemans, 2015). It is particularly evident that the ATF6 pathway, which is activated as a result of protein misfolding ER stress, plays a prominent role in stress-induced lipid accumulation (Han and Kaufman, 2016). Investigations of IRE1 and PERK mutants also suggest that perturbations in the ER lipid bilayer composition may lower the threshold for UPR-mediated activation (Volmer et al., 2013).

However, evidence suggests dysfunctional lipid metabolism may activate UPR regardless of the misfolded protein stress in the ER lumen (Basseri and Austin, 2012; Han and Kaufman, 2016). As documented previously, the deletion of the genes responsible for lipid biosynthesis triggers UPR activation, and reciprocally, other evidence shows upregulation of the genes involved in lipid metabolism upon UPR induction (Jonikas et al., 2009; Thibault et al., 2012; Travers et al., 2000). Among the various branches of UPR, the function of IRE1, a master regulator of UPR, has particularly been linked to lipid homeostasis. An excessive amount of triglycerides in hepatocytes has been reported upon the deletion of IRE1 (Ho et al., 2018). Independent of sensing protein misfolding, the PERK-eIF2a pathway can also regulate lipogenesis, given that induced PERK and eIF2a phosphorylation results in increased lipid accumulation in hepatocytes (Oyadomari et al., 2008). Thus, both IRE1a and PERK sense and respond to changes in lipid bilayer composition and trigger UPR activation (Ho et al., 2018; Volmer and Ron, 2015).

Along with the toxic accumulation of lipids, pathological accumulation of protein aggregates is commonly observed in the brains of those individuals suffering from neurodegenerative diseases. Notably, accumulated proteins may not localize in ER specifically, suggesting an indirect link between ER stress and lipid dysregulation-activated UPR (Figure 1) ((Scheper and Hoozemans, 2015)). While data examining the crosstalk between lipotoxicity, misfolded proteins and stress-induced UPR are limited, their dynamic interactions may regulate the pathophysiology of many neurodegenerative disorders and facilitate subsequent behavioral deficits.

Lipid composition and organization in the CNS

The brain is often recognized as one of the most lipid-rich organs, while lipid composition is itself known to influence a variety of clinically relevant behavioral indices (Tracey et al., 2018). Broadly speaking, lipids can be classified into eight major subcategories. These include fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, saccharolipids, polyketides, sterol lipids, and prenol lipids (Fahy et al., 2011). The relative conformation and modification of these macromolecules are commonly associated with variations in the structure and function of lipid bilayers (Casares et al., 2019). In neuronal membranes, lipids govern the arrangement of membrane proteins, including those ionic channels necessary for maintaining electrostatic potentials (Shamim et al., 2018). However, the CNS lipid composition is notable for influencing a variety of additional functions, such as secondary messenger cascades, energy conservation, and neuronal signaling (Tracey et al., 2018). However, lipid dysregulation and their accumulation as nano-droplets can lead to cellular dysfunction and compromised viability, commonly referred to as lipotoxicity, a phenomenon which has generally been defined by increased level of fatty acids (Hidalgo‐Lanussa et al., 2019; Weinberg, 2006). The crucial impact of altered lipid metabolism is demonstrated in many neurological disorders and neurodegenerative disease, including bipolar disorders, schizophrenia, Alzheimer’s, Parkinson’s, Niemann-Pick and Huntington disease (Adibhatla et al., 2006; Wenk, 2005).

Fatty acids, a diverse group of lipids synthesized by chain elongation of an acetyl-CoA group, are fundamental lipid components of the CNS (Hussain et al., 2019). Fatty acids are especially important in the brain, both because of their abundance as well as their vulnerability to oxidative stress, which can subsequently induce neuroinflammation (Fahy et al., 2011). They are classified into three main classes of esters: triglycerides, phospholipids and cholesterol esters (Naudí et al., 2015). By binding to hydroxyl groups of glycerol molecules, fatty acids are typically utilized to generate phospholipids for neuronal membrane formation and ensure proper functioning of neurons (Fahy et al., 2009; Hamilton et al., 2007; Tracey et al., 2018). In turn, such phospholipids maintain neuronal homeostasis through the regulation of membrane proteins, receptors, enzymes, and ion channels, as well as serving as bioenergetic reservoirs and precursors of secondary messengers for signal transduction, including inflammatory responses (Shamim et al., 2018). Considering their important role in membrane formation and protein rearrangement in lipid bilayers, they are not only critical for providing structural integrity in neurons but are also crucial for neuronal plasticity (Harayama and Riezman, 2018).

The most abundant fatty acyl components in the CNS are polyunsaturated fatty acids (PUFAs), which are the most susceptible fatty acids to oxidative damage and form numerous toxic products as a result of non-enzymatic lipid peroxidation (Nowak, 2013). Docosahexaenoic acid (DHA) is one notable PUFA given its modulation of several homeostatic processes in neurons, including inflammatory signaling cascades and cellular viability (Avallone et al., 2019; Tanaka et al., 2012). Specifically, DHA is incorporated into lipid species that are crucial for neuronal membrane integrity, including phosphatidylethanolamine (PE), phosphatidylserine (PS) and phosphatidylcholine (PC) (Wassall et al., 2018). Furthermore, DHA mediates the release of neurotransmitters, myelination and neuronal growth through activation of signaling pathways (de Bus et al., 2019). Therefore, DHA plays a critical role in neuronal development and protects neurons from consequential injuries, including aging (Hamilton et al., 2007; Wassall et al., 2018). Although other major classes of fatty acids, including Saturated (SAFAs) and Monounsaturated (MUFA) fatty acids, are proportionally lower than PUFAs; however, their dysfunctional metabolism can generate neuronal damage, and precipitate neuronal apoptosis (Hamilton et al., 2007; Hidalgo‐Lanussa et al., 2019; Mahley, 2016; Shamim et al., 2018).

In addition to fatty acids, sterol lipids are similarly essential for efficient physiological functioning of neurons given their diverse roles, including lipid raft formation and signal transduction mechanisms(Martín et al., 2014). Cholesterol is the most abundant form of sterols, which is known for its important roles in neuronal plasticity and synaptic remodeling(Mesa-Herrera et al., 2019). For instance, it enhances presynaptic differentiation and facilitates synapse formation during the functional development of the brain (Goritz et al., 2005; Hussain et al., 2019). As one of the most important components of neuronal membranes, cholesterol concentrations also regulate membrane fluidity, and thereby control membrane trafficking and transmembrane signaling (Casares et al., 2019; Dietschy, 2009). In addition, given that the vast majority of cholesterol in the brain is incorporated into myelin sheaths that surround axons, it is inherently critical for the efficient synaptic transmission of electrochemical potentials (Jung et al., 2017). Thus, of all tissues throughout the body, the CNS maintains the highest concentrations of cholesterol (Saher et al., 2011).

Cholesterol synthesis and recycling is a critical pathway for maintaining normal neuronal physiology. Conversely, abnormal cholesterol metabolism in the brain is associated with several neurodegenerative diseases, including AD, HD, PD and ALS (Martín et al., 2014). Notably, its metabolism in the brain is independent of peripheral circulation, due to the blood brain barrier’s inhibition of lipoprotein-bound cholesterol passage (Dietschy and Turley, 2001). Within the CNS, cholesterol is locally synthesized by astrocytes, microglia, oligodendrocytes and to lower extent by neurons (Dietschy, 2009). Redistribution and recycling of cholesterol throughout the brain employs lipoproteins, complex particles that consist of a core of triacylglycerols (Mahley, 2016). Specifically, their structure is characterized by a rich source of fatty acids, cholesterol esters and fat-soluble vitamins surrounded by an envelope of phospholipids and proteins constitutes apolipoproteins (Hamilton et al., 2007; Mahley and Huang, 2012). The unique lipoprotein metabolism pathway associated with cholesterol transfer in the brain reflects the importance of this mechanism to maintain lipid homeostasis in CNS (Mahley, 2016).

In addition to cholesterol transport, lipoproteins also translocate lipid precursors between organs through the bloodstream and facilitate the transport of phospholipids and cholesterol between glia and neurons in the brain (Mahley, 2016). As might be expected, the association of certain lipoproteins with brain disorders has been established in pathological studies; for instance, apoE4 has been detected as the major risk factor of the onset of AD (Bu, 2009; Shamim et al., 2018). ApoE lipoproteins are considered primarily responsible for regulating CNS lipid metabolism by redistributing cholesterol and phospholipids to cells for repair and remodeling (Mahley, 2016). Cholesterol itself also serves as a dynamic spacer that maintains raft assembly, while its removal from rafts causes dissociation of most proteins (Simons and Ehehalt, 2002). Similar to the backbone of sphingolipids (ceramide), cholesterol is synthesized in ER and is translocated to the Golgi apparatus, where raft assembly takes place (Simons and Ehehalt, 2002; van Meer, 1989).

In addition to the diverse classes of fatty acyls and cholesterol esters, the brain contains a large portion of sphingolipids (Pamplona, 2008). Sphingolipids are primarily synthesized in the ER and play a vital role in membrane integrity as well as cellular signaling (Shamim et al., 2018). Within neuronal membranes, sphingolipids exert a number of important biological functions that are attributed to their intrinsic polar nature(Shamim et al., 2018). This group of lipids are classified as sphingolipids or glycosphingolipids, and are designated based on fatty-acyl groups attached to the second carbon that forms ceramide. Glycosphingolipids are further characterized by their one or more carbohydrates attached to the C1 of ceramides (Merrill, 2011). In the CNS, sphingolipids are mostly found in the central and peripheral myelin structure and participate in the signaling pathways which control survival, migration and differentiation of neurons (Hussain et al., 2019; Naudí et al., 2015). Moreover, sphingolipids are important in synaptic transmission and modulate neuron-glia interaction (Tracey et al., 2018). Collectively, dysregulated sphingolipids metabolism is correlated with the pathogenesis of multiple neurodegenerative disorders, including AD, PD, ALS, HD and MS (Adibhatla and Hatcher, 2008; Mesa-Herrera et al., 2019; Puglielli et al., 2003).

Dysregulation of sphingolipid metabolism specifically has been shown to lead to modified membrane organizations that participate in a diverse range of pathogenic events (Adibhatla and Hatcher, 2008). These alterations are most commonly reflected in alteration of lipid rafts (Sonnino et al., 2014), the unique components of plasma membrane enriched in cholesterol, and sphingolipids (Simons and Ehehalt, 2002). These components form microdomains in plasma membranes that in neurons are essential for cell adhesion, axonal guidance, and synapse communication (Petro and Schengrund, 2009). In addition, lipid rafts are dynamic molecular structures which include receptors, channels, recognition molecules, and enzymes that are involved in signal transduction and intracellular trafficking of proteins and lipids (Brown and London, 1998; Korade and Kenworthy, 2008). In neurons, lipid rafts are particularly essential for neuronal adhesion and synapse communication (Petro and Schengrund, 2009).

As a multifunctional organelle that has specific sensors for regulating the biosynthesis of different lipid classes (e.g. sterols, phospholipids) as well as their distribution to the other cellular compartments, the ER is essential for mitigating lipotoxicity (Jacquemyn et al., 2017). Lipid droplets which arises from ER, are also a dynamic lipid structures that involved in a variety of cellular functions and showed a significant role in cellular signaling and inflammation (Farmer et al., 2020). Additionally, given that lipid dysregulation is common amongst neurodegenerative diseases, it is relevant to consider the potential cross-talk between ER functioning and those pathogenic processes shared amongst such conditions, namely ER stress.

Interplay between lipid alterations and UPR in neurodegenerative disease

The pathogenesis of CNS diseases is most commonly associated with impaired neuronal homeostasis as well as the underlying mechanisms that facilitate such dysregulated homeostasis, including alterations in membrane lipid composition (Adibhatla and Hatcher, 2008). In particular, impaired lipid metabolism, maladaptive regulation of intraneuronal proteins and a disruption of neuronal homeostasis are increasingly recognized as hallmarks of CNS disease progression (Hetz and Saxena, 2017; Scheper and Hoozemans, 2015; Smith and Mallucci, 2016; Tracey et al., 2018). In this current section, we discuss the possible cross-talk between altered membrane lipid composition and disrupted PQC in the progression of neurodegenerative disease.

The ER is a hub of homeostatic control by sensing the stress signals induced through misfolded protein accumulation or membrane lipid perturbation (Volmer et al., 2013). Lipids are specifically important biomolecules of CNS because of their abundance and their central role in dictating membrane structure and governing their biophysical properties (Sastry, 1985). Dysfunction in lipid homeostasis is a known risk factor of many neurodegenerative diseases, including those marked by extensive alterations in neuronal development, gliosis and ultimately neuronal apoptosis (Adibhatla and Hatcher, 2007; Hussain et al., 2019; Mesa-Herrera et al., 2019).

Along with the accumulation of aggregated proteins (e.g. Aβ, Tau), alteration in ER functioning and dysregulated UPR are associated with neurodegenerative disease, suggesting that all effective factors which induce UPR activation, including abnormal lipid homeostasis, may facilitate the pathogenesis of neurodegenerative diseases (Cirone, 2020; Lindholm et al., 2006; Yoshida, 2007). The exact role of UPR in the pathophysiology of CNS disease has not been fully understood, but it has been shown that the ER stress is the hallmark of most known neurodegenerative disease (Lindholm et al., 2006; Xiang et al., 2017). In addition, a robust association is observed throughout the clinical population between cholesterol transport mechanisms deficits and the diagnosis of neurological disorders such as AD, Niemann–Pick type C disorder, and HD (Corder et al., 1993; Lambert et al., 2013; Petrov et al., 2016). Thus, given its processing is ER-dependent, variation in cholesterol regulation may contribute to abnormal ER functioning and aberrant UPR activation.

Indeed, the ER is a major site for the synthesis of sterols as well as sphingolipids, and it has a significant role in controlling membrane lipid composition. Impaired ER-Golgi trafficking of either proteins or lipids is associated with ER stress (Biden et al., 2014). In a stressed ER, excessive production of reactive oxygen species (ROS) can cause oxidative stress, a pathological hallmark of the neurodegenerative diseases (Reed, 2011). Furthermore, increase in ROS production has been identified as an important mechanism by which neuronal plasticity is compromised during aging and after CNS injury (Salim, 2017). Elevated oxidative stress is a common correlate of pathogenesis of neurodegenerative disease, suggesting ROS modifications to lipid structures could influence ER functioning and UPR in neurodegenerative conditions (Barnham et al., 2004; Chen et al., 2012; Malhotra and Kaufman, 2007). (By peroxidizing lipids, elevated oxidative stress is known to generate unstable byproducts that compromise neuronal functioning and viability (Pamplona, 2008). Notably, such increased levels of lipid peroxidation are associated with significantly shorter life expectancies across a variety of mammalian species, and correlate with the onset of age-associated neuropathology (de Diego et al., 2019; Jové et al., 2013). In addition, elevated oxidative modifications to cholesterol are suggested to facilitate the accumulation of toxic Aβ, provided that oxysterols (e.g 24-OH Chol & 27-OH Chol) are known to mediate the enzymatic processing of APP and influence Aβ production (Famer et al., 2007; Marwarha et al., 2013). Evidence also indicates a positive feedback loop between ROS generation and ER dysfunction, suggesting the adverse consequences of oxidative stress may be exacerbated by the ER, and further contribute to ER dysregulation (e.g. chronic UPR, Ca+ overload etc.) (Bhandary et al., 2012). Therefore, the dynamic relationship between ER processing, UPR and lipid metabolism in the CNS may be perturbed by elevated oxidative stress in neurodegenerative disease and contribute to pathogenesis.

As the most abundant component of neuronal membrane and lipid rafts, cholesterol is a vulnerable component to oxidation and can affect cellular functions through its biologically active oxidized product-oxysterol (Zhang et al., 2015). Oxidative stress, which causes lipid peroxidation specifically in phospholipids, sphingomyelin and its primary precursor, ceramide, has also been implicated in many pathological studies (Angelova and Abramov, 2018; Barnham et al., 2004; Reed, 2011). As a cholesterol-rich platform with vulnerable membrane domains to oxidative stress, lipid rafts may modulate membrane ER protein activity, protein folding or trafficking through changes in local membrane structure which indirectly may transduce UPR (Volmer and Ron, 2015). It has been known for about a decade that ER stress can perturb lipid metabolism and lipogenic gene regulation (Bobrovnikova-Marjon et al., 2008; Oyadomari et al., 2008; Werstuck et al., 2001). However, the underlying mechanisms are not well understood. There is a bidirectional relationship between lipid metabolism and ER stress, while aberrant lipid metabolism can also cause the ER stress and ultimately UPR activation (Fu et al., 2011; Han and Kaufman, 2016; Wei et al., 2006; Werstuck et al., 2001). A recent study indicated that increased proportions of phosphatidylcholine (PC), in comparison to phosphatidylethanolamine (PE), along with saturated fatty acids, triglyceride, and ceramide accumulation can stimulate ER stress (Basseri and Austin, 2012).

In sum, lipid perturbation as well as protein aggregation is associated with the dysfunctional UPR and ultimately apoptosis, suggesting a possible lipid-protein interaction in neurodegenerative disease progression (Figure 1). In the next section, we discuss HANDs as a comprehensive model of neurodegeneration that shows overlaps with non-viral neurodegenerative disease. In addition, clinically relevant examples where both protein misfolding and dysregulated lipid activation of UPR are further considered.

HIV/Neuro-HIV associated dysregulation of neuronal homeostasis

Human Immunodeficiency virus-type1 (HIV-1) infection often causes diverse complications and neurological symptoms, including cognitive and motor disorders termed neuro-HIV (Price et al., 1988). HIV-1 enters the brain of the infected individual in the early stages of the infection resulting in HANDs progression (Price et al., 1988). HANDs is described as various neurological disorders that accompany mental slowness, trouble with memory and concentration, speech problems, and behavioral changes. Although the HIV infection initially appears in the human immune system, the disease progression and symptoms appear after toxin infiltration from monocytes and macrophages to the central nervous system (CNS) (Bagashev and Sawaya, 2013; Mediouni et al., 2015). Consequently, HIV-1 infection disrupts neuronal homeostasis by dysregulating multiple pathways, many of which are modulated by the HIV-1 trans-activator of transcription (Tat) protein which impairs neuronal homeostasis at different levels (Clark et al., 2017). Furthermore, neuronal toxicity and loss of function remain at the heart of HIV associated neuropathy. Indeed, neurocognitive impairments are reported in about half of HIV patients despite treatment with antiretroviral therapies (ART) (Deeks et al., 2012; Mediouni et al., 2015; Richman et al., 2009). Although the cause of HAND in patients with well-controlled viremia is uncertain, some factors thought to contribute to HAND progression include aging, the persistence of HIV reservoirs in the brain, and long term CNS pharmaco-toxicity (Cross et al., 2013; Wallet et al., 2019). The toxic effect of viral proteins such as Tat, Nef and Env in altering PQC, lipid metabolism and cell survival pathways has been studied extensively in neurons(Ahooyi et al., 2019; Fan and He, 2016; Lindwasser and Resh, 2001; Mayne et al., 1998; Shi et al., 1998; Zheng et al., 2001). However, deficiencies in lipid metabolism following HIV infection and its potential crosstalk with neuronal PQC pathways requires further consideration.

A growing body of literature confirms lipid alterations upon HIV-1 infection and subsequently ART treatment; specifically, alterations are seen amongst the most vulnerable lipid species to oxidative stress, including, ceramide, sphingomyelin and cholesterol (Bonfanti et al., 2007; Haughey et al., 2004; Oka et al., 2012; Rose et al., 2008). Similar to observations in aging and multiple neurodegenerative disease, elevations in membrane-associated oxidative stress is a common correlate of Neuro-HIV (Ivanov et al., 2016). For example, increased levels of pro-inflammatory lipid species such as oxidized LDL and HDL are established markers of immune activation in response to HIV (Funderburg and Mehta, 2016). Similarly, TNFa and FasL, two inflammatory mediators and potent inducers of ceramide generation, are significantly increased in HIV-Associated Dementia (HAD) (Haughey et al., 2004; Mayne et al., 1998). Synergistic neurotoxicity between viral proteins gp120, Tat and TNFa has been also suggested to have occurred in the level of ceramide (Haughey et al., 2004). The observation in HIV patients with mild and moderate or severe dementia suggested the perturbation of sphingomyelin metabolism occurs early in HIV dementia progression which is preceded by the formation of ceramide (Haughey et al., 2004). Interestingly, it has also been shown that the severity of cognitive impairments of HIV is correlated with the level of sphingolipid and ceramide accumulation (Haughey et al., 2004). In addition, considerable evidence indicates elevated ROS in HIV can influence ceramide production (Bezombes et al., 2001; Haughey et al., 2004; Singh and Kishimoto, 1980). Indeed, the CNS of HIV patients can be characterized by increased sphingomyelinase activity and elevation in ceramide that can subsequently result in neuronal death (Haughey et al., 2004; Shi et al., 1998). An intimate relationship between ROS, ceramide and apoptosis in neurons has been studied extensively, supporting the hypothesis that lipid metabolism dysfunctionality is a critical player in HAD progression (Arora et al., 1997; Lee et al., 2006; Quillet-Mary et al., 1997).

Along with ceramide and sphingomyelin, the dysregulation of cholesterol metabolism in both neuronal and non-neuronal cells following HIV infection has been studied extensively (Cotto et al., 2018; DeLucia et al., 2018; Mujawar et al., 2006; Sviridov et al., 2020) (Zangerle et al., 1994). HIV infected individuals display increased levels of cholesterol esters in the CSF as well as cholesterol accumulation in the brain, suggesting neuro-HIV is associated with cholesterol dysregulation ((Bandaru et al., 2013).

The combined effect of HIV-1 Tat and cocaine on astrocytes in culture confirmed disrupted cholesterol regulation through decreased expression of the lipid carrier and lipid transporter (i.e. APOE), followed by declines in astrocyte derived cholesterol (Cotto et al., 2018).In the same study, it has been shown that the HMGCR (HMG-CoA reductase), a rate limiting enzyme in cholesterol biosynthesis, is reduced by 70% under ER stress. Moreover, the increased activity of a protein that is necessary to stimulate cholesterol biosynthesis, that is SREBP2 (sterol regulatory element-binding protein 2), also been detected in astrocytes exposed to combined Tat/cocaine (Cotto et al., 2018). Such Tat/cocaine induced ER stress has been reported by several other studies (Fan and He, 2016; Ma et al., 2016; Periyasamy et al., 2016; Shin et al., 2007; Sil et al., 2019). Published data from our laboratory also confirmed such an alteration in cholesterol metabolism pathways in neurons upon exposure to Tat or cocaine. Here, we illustrated the accumulation of cholesterol and activation of the cholesterol esterification enzyme (SOAT1/ACAT1) in primary neurons upon HIV-1 Tat/cocaine treatment, which contributed to the elevation of total cholesterol in neurons (Ahooyi et al., 2018). Tat-induced upregulation of cholesterol biosynthesis genes such as sterol regulatory element‐binding proteins (SREBPs) and lipoprotein receptor-related protein (Lrp1) detected in this study suggested Tat/cocaine contributions to disrupted cholesterol biosynthesis in neurons (Ahooyi et al., 2018). However, the downstream effects of cholesterol accumulation in neurons in HIV-1 infection has yet to be examined directly.

Moreover, HIV infection can alter the composition of fatty acids and elevate the level of free fatty acids in the serum of HIV⁺ individuals (Bowman et al., 2019). For instance, increased levels of free saturated fatty acids and lysophosphatidylcholine (LPC) fatty acids contribute to the HIV-associated inflammation and the activation of the immune response ((Bowman et al., 2019; Funderburg and Mehta, 2016).  Additionally, investigations of lipid profiles in HIV infected individuals versus non-infected persons exhibit global alterations in the lipidome in ART-treated HIV individuals, which may be indicative inflammation and disease progression (Bowman et al., 2019). Despite the remarkable success of combined antiretroviral treatment (ART) in mitigating the severity of HIV-associated dementia, the incidents of cognitive dysfunction remain highly prevalent in ART treated HIV-1 patients (Matinella et al., 2015). The side effects of ARTs and their neurotoxicity has been increasingly studied; it has been shown that these drugs trigger oxidative stress as well as alterations in lipid and protein metabolism, which can lead to inflammatory responses and ER stress (Apostolova et al., 2015; Nooka and Ghorpade, 2017). The HAND relevant stimuli also interfere with ER calcium homeostasis fluctuation which triggers ER stress and UPR activation (Hu, 2016). In primary human astrocytes, elevated levels of the calcium-binding chaperones such as BiP, along with increased level of three UPR activation markers (PERK, IRE1a and ATF6) in the present of HAND-relevant stimuli indicate the importance of ER stress in HIV dementia progression (Nooka and Ghorpade, 2017). Although dyslipidemia and decline in total cholesterol in the serum of HIV+ patients has been recognized prior to exposure to ART (Grunfeld et al., 1989), evidence suggests ARV-induced dyslipidemia in PWH is termed HIV-associated lipodystrophy (Feeney and Mallon, 2011).

Along with significant dysregulation of lipid homeostasis in neurons, disturbance in Ca²⁺ signaling and elevated oxidative stress are well known stimuli that trigger UPR activation in HIV infected individuals (Fan and He, 2016; Shi et al., 1998). Consistent with this postulation, several studies illustrate the role of HIV infection in perturbation of stress-induced PQC pathway (Ahooyi et al., 2019; Borsa et al., 2015; Fan and He, 2016; Nooka and Ghorpade, 2017). Published data from our laboratory also indicated that HIV-1 Tat protein perturbs synapsis homeostasis in rat primary neurons through suppression of BAG3 protein, the regulator of the heat-shock protein (Hsp70/Hsc70) which controls PQC through eliminating misfolded/unfolded proteins (Ahooyi et al., 2019). The potential for HIV-1 Tat to induce UPR activation directly in primary neurons has yet to be investigated. However, UPR activation has been confirmed in astrocytes upon exposure to HIV-1 Tat, which is in turn suggested to lead to astrocyte mediated neuro-toxicity(Fan and He, 2016). In addition, chemical chaperon mediated inhibition of UPR/ER stress can mitigate such glia cytotoxicity otherwise induced by Tat (Fan and He, 2016). Similarly, another investigations found significant elevations in the critical UPR markers (e.g. BIP and ATF6) in HIV target cells, both in vitro and in vivo (Borsa et al., 2015) (Figure 2).

Figure 2:

the schematic shows the HIV-induced UPR activation in neurons. HIV-mediated neurotoxins through infected microglia and astrocytes increase the level of ER-stress inducers including: free radicals, imbalance Ca²⁺, accumulation of free fatty acids (FA) as well as triglyceride (TG) that leads to UPR activation. Moreover, accumulation of unfolded/misfolded proteins and toxic lipid species such as ceramides is associated with HIV toxicity in neurons. Although there is no direct evidence indicating the link between UPR and neuronal apoptosis in HAND, it has been suggested that accumulation of HIV-induced ER-stress markers (BiP and ATF6) can cause cell death. The luminal side of ATF6 and PERK are bound to the BiP which keep them inactivated. Upon UPR activation BiP releases from this complex to assist with refolding proteins and in response to FA and TG accumulation. Activated ATF6 passes through the nucleus and upregulates the ERSE (ER stress response Elements) regulated genes expression such as CHOP which activate the apoptosis pathway. IRE1-α also activates in response to increased amount of saturated FA and sterol, as well as unfolded protein accumulation. In sum, this schematic briefly shows the overlaps between HIV-derived lipid dysregulation and accumulation of unfolded proteins in UPR activation(Created with BioRender.com).

Together, these data support the modulation of multiple neuronal survival pathways during HIV infection and elucidate the novel consideration of a potential interaction between a toxic accumulation of critical biomolecules, such as lipids and proteins (Figure 2).

Concluding remarks and future perspectives:

Lipid dysregulation and lipotoxicity associated UPR activation is an emerging topic in studying neurocognitive disorders and dementia. Alterations in the lipid content can be detrimental to neuronal survival, given their capacity to induce cell death and ultimately neuronal dysfunctionality. Meanwhile, cellular responses to stress remain a vital homeostatic mechanism that is regulated through UPR signaling. In this regard, the UPR activation has been discussed in response to a variety of ER-stress inducers, such as protein aggregates and lipid metabolism alterations, that are reflected in AD, viral infection, and other environmental stimuli that cause neuronal injuries. However, the link between each of these pathways and lipid metabolism is not entirely understood.

HIV-1 infection has been recently introduced as a UPR stimulus, although it has not been studied extensively. In this review, we highlight the importance of lipid metabolism in the progression of HIV-dementia, and suggest a novel approach to assess the toxicity of HIV-1 infection through studying the underlying pathways which may link lipid dysregulation to PQC disturbance in HIV dementia. Examining the link between HIV associated lipid dysregulation and UPR activation a novel approach in the field, and can shed light on the development of new therapeutic approaches. Furthermore, considering the particular increase of cognitive dysfunction in the aged HIV-infected population (Fazeli et al., 2014; Sacktor et al., 2016; Valcour et al., 2004), identification of common hallmarks in neurocognition disorders and HIV dementia can lead to a better understanding of the underlying pathways, and provide us with better therapeutic approaches for patient outcomes.

In addition, the interpreted post-ART ER stress and UPR activation can provide a research model to examine the underlying pathways involved in dementia procedure and block the disease progression (Gills et al., 2007; Nooka and Ghorpade, 2017; Robertson et al., 2010; Schönthal, 2012). With regard to the functional significance of lipid composition in ER integrity and the common pathological features of HAND and other neurodegenerative disease, it is likely that targeting ART-elevated dyslipidemia and the mechanism by which they interact with PQC possess more beneficial outcomes (Dave et al., 2016; Feeney and Mallon, 2011; Kelesidis and Currier, 2014; Souza et al., 2013).

As more than 90% of identified proteins and pathways in HIV-associated dementia and non-viral neurodegeneration overlap (Zhou et al., 2010), the findings of pathway studies can possibly be generalized to more than an individual disease. Despite that lipodystrophy is a major concern in many patients maintained on ARTs long-term (Mutimura et al., 2007) its effects on the CNS require further study. Investigating the underlying mechanisms of ART-induced dyslipidemia in neurons will lead to the introduction of more lipid-beneficial therapeutic options, which can improve the life quality of PWH as well as introduce a new avenue of treatment in ER stress-driven neurodegeneration.