Epigenetics of μ-Opioid receptors: Intersection with HIV-1 infection of the Central Nervous System

Abstract

The abuse of intravenous drugs, such as heroin, has become a major public health concern due to the increased risk of HIV-1 infection. Opioids such as heroin were originally identified and subsequently abused for their analgesic effects. However, many investigations have found additional effects of opioids, including regulation of the immune system. As such, chronic opioid abuse has been shown to promote HIV-1 pathogenesis and facilitate HIV-1-associated neurocognitive dysfunction. Clinical opioids, such as morphine and methadone, as well as illicit opioids, such as heroin, exert their effects primarily through interactions with the μ-opioid receptor (MOR). However, the mechanisms by which opioids enhance neurocognitive dysfunction through MOR-mediated signaling pathways are not completely understood. New findings in the regulation of MOR expression, particularly epigenetic and transcriptional regulation as well as alternative splicing, sheds new insights into possible mechanisms of HIV-1 and opiate synergy. In this review, we identify mechanisms regulating MOR expression and propose novel mechanisms by which opioids and HIV-1 may modulate this regulation. Additionally, we suggest that differential regulation of newly identified MOR isoforms by opioids and HIV-1 has functional consequence in enhancing HIV-1 neurocognitive dysfunction.

Introduction

The abuse of intravenous drugs, such as heroin, contributes substantially to the global HIV-1 epidemic, with an estimated 20% of injection drug users (IDUs) infected with HIV-1. The direct contribution of intravenous (IV) drug abuse to HIV-1 infection is about 30% worldwide, excluding sub-Saharan Africa, and is the third most frequently reported risk factor for HIV-1 infection in the United States (CDC, 2009; Vlahov et al., 2010). Risk behaviors associated with IV drug abuse, primarily sharing of syringes, has driven the HIV-1 epidemic in Eastern Europe, Asia, North Africa, and North & South America. The largest reports of HIV-1 infection among IDUs are found in China, the United States, and Russia (Vlahov et al., 2010). Although the incidence of HIV-1 infection among IDUs has decreased nearly 80% since the late 1980’s, IDUs still represent a large proportion of the population with new HIV-1 diagnoses (CDC, 2009).

In addition to the devastating effects it has on the immune system, HIV-1 infection is a major cause of several neurocognitive dysfunctions, collectively known as HIV-1-associated neurocognitive disorders or HAND. Symptoms of HAND are associated with neuronal damage. Since HIV-1 does not produce active infection in neurons, neuronal damage in HIV-1 infection has been proposed to be mediated either by direct interactions of secreted viral proteins with neurons or indirectly by inflammatory molecules secreted by infected glia (Lindl et al., 2010) and/or astrocytes. Introduction of highly active antiviral therapy (HAART), now called combination antiretroviral therapy (cART), has significantly reduced the progression of HIV-1 as well as HAND (Ances and Ellis, 2007; Gray et al., 2003; Lindl et al., 2010; Sacktor, 2002). However, as a result of therapy, HIV-1-infected individuals are living longer, leading to an increased prevalence of HAND (Anthony and Bell, 2008; González-Scarano and Martín-García, 2005; McArthur, 2004; Wang et al., 2006).

Aside from providing a method of viral transmission, injection drug abuse also intrinsically affects HIV-1 pathogenesis, particularly in the central nervous system (CNS). The immunosuppressive effects of opioids enhance HIV-1 pathogenesis (Cabral, 2006) while synergistic HIV-1-opioid interactions mediate inflammation and glial dysfunction, leading to neurological complications (Hauser et al., 2006; Hauser et al., 2007). Neurotoxicity during HIV-1 infection may occur indirectly, with toxic cellular products released from infected glial cells causing dysfunction in nearby neurons (Hauser et al., 2007), or directly by neurotoxic HIV-1 viral proteins such as gp120, Tat, and Vpr (Crews et al., 2008; Lindl et al., 2010). It is suggested that opioid abuse reduces the threshold for HIV-1 neurotoxicity due to overlapping apoptotic pathways activated by opioids through the μ-opioid receptor (MOR) (Carmody, 1987) and by HIV-1 viral proteins through chemokine receptors, primarily CXCR4 and CCR5 (Hauser et al., 2006). However, specific mechanisms and common cellular pathways have yet to be fully established.

In this review, we will examine the role of opioid abuse in directly enhancing HIV-1 neurocognitive dysfunction and how newly identified regulatory mechanisms of MOR expression may affect this process. We postulate that epigenetic activation and transcriptional modulation of OPRM1 by signaling cascades mediated by HIV-1 infection and opioid abuse may alter MOR expression, thus modifying HIV-1-opioid crosstalk. Additionally, we examine how alternative splicing may alter MOR function. Common epigenetic, transcriptional, translational, and signal transduction mechanisms that regulate alternative splicing and how these mechanisms may be altered by both opioids and HIV-1 infection to differentially regulate MOR isoform expression are discussed. We examine how differential expression of MOR isoforms may result in the activation of different biochemical pathways, altering HIV-1-opioid crosstalk and neurotoxic signaling. Lastly, we identify similar biochemical pathways activated separately by opioids and HIV-1 infection that potentially overlap, resulting in a decreased threshold for HIV-1 proteins or opioid signaling and increased neurotoxicity.

Exceptional diversity of opioids and their receptors: Biological significance

Opioid receptors are members of the G protein-coupled receptor (GPCR) family, which are characterized by having seven transmembrane domains and associating with G protein subunits. GPCRs bind a variety of physiological and environmental molecules, including hormones and neurotransmitters (Devi, 2010). Natural alkaloids derived from the resin of the opium poppy, called opiates, as well as endogenous opioids, such as enkephalins, endorphins, and dynorphins, and synthetic opioids, such as morphine, methadone, and heroin, all bind opioid receptors with various affinities (Carmody, 1987; Drake et al., 2007). This variation in binding affinity was originally used to classify three different receptor subtypes, designated as μ (MOR), κ (KOR), and δ (DOR) opioid receptors (Reisine, 1995). Later, genetic screening verified the existence of these three distinct subtypes (Smith and Lee, 2003). The existence of three related, yet distinct, genes coding for μ, κ, and δ receptors, and the physical separation of these genes, has allowed for the independent evolution of protein coding sequences and promoter elements, generating great diversity in receptor expression and function (Gray et al., 2006; Mrkusich et al., 2004; Peckys and Landwehrmeyer, 1999; Singh et al., 1997; Wang and Wessendorf, 2002). Of the three opioid receptor subtypes, functional diversity among the MOR is of particular interest because of its high affinity for clinically used opioids as well as illicit opioids of abuse.

Opioid compounds were originally identified for their analgesic properties and, likewise, investigations of MOR regulation and diversity have primarily focused on its role in the sensation and modulation of pain. However, the expression of MORs in multiple cell types indicates that opioids and their receptors may modulate additional physiological systems, including the immune system (Groneberg and Fischer, 2001; Reisine, 1995; Toda et al., 2009). In this regard, the role of opioids in HIV-1 pathogenesis has become an increasingly important topic. While the increased risk of HIV-1 contraction due to IDU behavior can account for most the HIV-1 pathology (CDC, 2009), the rapid progression to AIDS and other HIV-1-associated diseases, such as HAND, in IDUs suggests that opioids actively play a role in HIV-1 pathogenesis (Hauser et al., 2005; Hauser et al., 2006; Hauser et al., 2007; Nath, 2010). This may, in part, be due to the immunosuppressive effects of opioids. However, direct interactions between neurons, opioids, and HIV-1 viral proteins are now believed to also participate in increased neuronal dysfunction (Rogers, 2011). Therefore, regulation and functional variation of MOR expression has major significance not only in the modulation of pain but also in the modulation of disease pathogenesis, particularly in HIV-1 (CDC, 2009; Guo et al., 2002; Li et al., 2003; Li et al., 2002; Peterson et al., 1994; Peterson et al., 1999; Peterson et al., 1990; Suzuki et al., 2002).

Structural organization and regulation of OPRM1 gene

The OPRM1 gene, localized on chromosome 6q24-q25 spans a region of 243366 base pairs (Gene ID 4988) and comprises over 13 identified exons (Figure 1). Pre-mRNA transcripts from the OPRM1 gene undergo extensive splicing to generate 21 known isoforms of the human μ-opioid receptor (Table 1).

Figure 1. Exon mapping of the human OPRM1 gene.

Pre-mRNA transcripts from the OPRM1 gene undergo extensive splicing to generate 21 known isoforms of the human μ-opioid receptor. Shown above are the known exon regions of the OPRM1 gene. Solid arrows indicate that the exon below is derived from a region of the above exon. For example, exon 3c is generated from a segment of exon 3a. This may occur either due to incomplete incorporation of the exon or the incorporation of flanking intronic sequences. Given the complexity of the OPRM1 exon expression, transcriptional and epigenetic regulation, particularly interacting histone modifications, chromatin-binding proteins, and splicing factors at the intron-exon junctions, is critical for regulating differential splicing of the MOR. Exact mechanisms regulating alternative splicing specificity of the MOR are not known. As such, the corresponding exon names and numbers may not match those cited in earlier literature.

Table I. Isoforms of the human μ-Opioid Receptor.

Currently there are 21 known isoforms of the human μ-opioid receptor generated by alternative splicing. Here we show both the exon map and protein sequence of each individual MOR isoform as well as known localization in human tissue and downstream signaling. The exon sequence encoding for each isoform is indicated in the table. The exon map of the OPRM1 gene is shown in Figure 1. The protein sequence of each isoform is indicated in the table and conserved protein sequences are shown as textured boxes. Full protein sequences of conserved regions are indicated below. Textured boxes that contain a short amino acid sequence indicate that in that given isoform, the conserved region is truncated after the sequence shown. All exon and protein sequences were obtained from the NCBI database. As such, the exon names and numbers corresponding to a given isoform may not match those cited in earlier literature.

Variant	Exon Map	Variant	Tissue Expression	Signaling	Ref
Mor-1	3a-5a-6a-11		Immune Cells, Spinal cord, Pre-frontal Cortex, Temporal cortex, Piriform cortex, Nucleus accumbens, Pons, Cerebellum	-Gi/Go Coupled GPCR -Inhibition of adenylyl cyclase and cAMP -Inhibition of N-, P/Q-, and L-type Ca2+ channels, -Stimulate K+ channels -Activate (MAP) kinase cascade	(Law et al., 2000; Xu et al., 2009)
MOR-1A	3a-5a-6b		Cutaneous Nerve Fibers, Be(2)C Cells, CNS	-Inhibition of adenylyl cyclase and cAMP	(Bare et al., 1994; Pan et al., 2005a; Ständer et al., 2002)
MOR-1B1	3a-5a-6a-10		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)
MOR-1B2	3a-5a-6a-9b		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)
MOR-1B3	3c-5a-6a-9a		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)
MOR-1B4	3a-5a-6a-8b		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)
MOR-1B5	3a-5a-6a-8a		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)
MOR-1C	3d-5c-6a-12		Unknown	Unknown
MOR-1G1	1b-5b-6a-11		Spinal cord, Pre-frontal Cortex, Temporal cortex, Piriform cortex, Nucleus accumbens, Pons, Cerebellum	Unknown	(Xu et al., 2009)
MOR-1G2	1a-5a-6a-11		Spinal cord, Pre-frontal Cortex, Temporal cortex, Piriform cortex, Nucleus accumbens, Pons, Cerebellum	Unknown	(xu et al., 2009)
MOR-1H	1c-2-3c-5a-6a-11		Unknown	Unknown
MOR-11	1d-3e-5a-6a-11b		Pre-frontal cortex, Piriform cortex, Pons	Unknown	(Xu et al., 2009)
MOR-1JL	J-2J-3J-4		Unknown	Unknown
MOR-1K1	13a-2K-3K4		Frontal lobe, Medulla oblongata, Insula, Nucleus accumbens, Pons, Spinal cord, and Dorsal root ganglion	-GS Coupled GPCR -Activation of adenylyl cyclase and cAMP -Activation of Ca2+ channels -Increased NO production	(Gris et al., 2010)
MOR-1K2	13b-2K-3K-4		Frontal lobe, Medulla oblongata, Insula, Nucleus accumbens, Pons, Spinal cord, and Dorsal root ganglion	-GS Coupled GPCR -Activation of adenylyl cyclase and cAMP -Activation of Ca2+ channels -Increased NO production	(Gris et al., 2010)
MOR-1O	3b-5c-6a-12		Be(2)C Cells, CNS	Unknown	(Cadet, 2004; Pan et al., 2003)
MOR-1R	3d-5a-6a-7		Unknown	Unknown
MOR-1V	3f-5a-6a-9-11c		Unknown	Unknown
MOR-1W	W-5a-6b		Unknown	Unknown
MOR-1X	3b-5c-6a-7		Be(2)C Cells, CNS	Unknown	(Pan et al., 2003)
MOR-1Y	3d-5a-6a-9c		Be(2)C Cells	-Inhibition of adenylyl cyclase and cAMP	(Pan et al., 2005a)

MCLHRRVPSEETYSLDRFAQNPPLFPPPSLPASESRMAHPLLQRCGAARTGFCKKQQELWQRRKEAALGTRKVSVLLATSHSGARPAVST

MDSSAAPTNASNCTDALAYSSCSPAPSPGSWVNLSHLDGNLSDPCGPNRTDLGGRDSLCPPTGSPSMITAIMALYSIVCVVGLFGNFLVMYVIVRYTK

MKTATNIYIFNLALADALATSTLPFQSVNYLMGTWPFGTILCKIVISIDYYNMFTSIFTLCTMSVDRYIAVCHPVKALDFRTPRAKIINVCNWILSSAIGLPVMFMATTKYRQGSIDCTLTFSHPTWYWENLLKICVFIFAFIMPVLIITVCYGLMILRLKSVRMLSGSKEKDRNLRRITRMVLVVVAVFIVCWTPIHIYVIIKALVTIPETTFQTVSWHFCIALGYTNSCLNPVLYAFLDENFKRCFREFCIPTSSNIEQQNSTRIRQNTRDHPSTANTVDRTNHQ

MOR-1 like other G-protein coupled receptor contains seven transmembrane (TM) domains with an extracellular N-terminus and an intracellular C-terminus (Figure 2). The first three exons encode for the N-terminus and the TM domains.

Figure 2. Schematic representation of the mechanisms involved in OPRM1 gene regulation.

The OPRM1 gene expression is tightly controlled through a variety of transcription factors, alternative splicing mechanism, miRNAs, and by epigenetic mechanisms. The expression of OPRM1 by cytokines (1) and opioid agonist (2) can be regulated by a group of nuclear receptors and transcription factors. Recent evidences suggest that OPRM1 expression can also be regulated by epigenetic mechanisms, namely histone modifications (3) DNA methylation (4), and microRNAs (5). Finally, the expression of OPRM1 can also be regulated by mechanisms involving spliceosome recruitment to generate alternatively spliced MOR isoforms (6).

Originally, the OPRM1 promoter was thought to contain both distal and proximal promoter, however it now appears that this is not the case. The OPRM1 gene is unique in that there is no distinct singular promoter region, as in the mouse homologue (Choi et al., 2008). Instead, it has been found that the OPRM1 promoter region is TATA-less and therefore has multiple transcriptional initiation sites (Bedini, 2008; Xu and Carr, 2001b). Transcription of OPRM1 is controlled by a complex interplay of positive and negative regulatory elements (Xu and Carr, 2001b). These regulatory elements include NFκB(Kraus et al., 2003), SP1 & SP3(Xu and Carr), AP-1, and STAT6 (Börner et al., 2004; Hwang et al., 2010; Kraus, 2009; Kraus et al., 2001; Kraus et al., 2003). Cytokines, such as interleukin-4 (IL-4) (Kraus et al., 2001), interferon-γ (IFN-γ) (Kraus et al., 2006), insulin-like growth factor-1 (Bedini et al., 2008), and phorbol esters (Wei and Loh, 2011), among many others, also modulate OPRM1 transcription (Figure 2). As we will discuss in detail, epigenetic mechanisms may regulate OPRM1 expression by modulating these transcriptional initiation sites.

MOR-mediated signal transduction

The classical MOR is a prototypical G_i/G_o coupled G-protein coupled receptor. As such, activation of the typical MOR stimulates inhibitory pathways mediated by heterotrimeric G_i and/or G_o proteins. These pathways include inhibition of adenylyl cyclases and N-, P/Q-, and L-type Ca²⁺ channels, altering intracellular Ca²⁺ levels. Activation of the MOR is also known to stimulate K⁺ channels and regulate the mitogen-activated protein (MAP) kinase cascade (Law et al., 2000). However, GPCRs have been shown to activate additional pathways, including the anti-apoptotic phosphoinositide 3 kinase (PI3K)/Akt pathway involved in neuronal survival (Polakiewicz et al., 1998; Yin et al., 2006). This has been shown to occur through the association of the p110γ PI3K isoform with either the G_α or G_βγ subunit of the heterotrimeric G protein (Duronio, 2008; Tegeder and Geisslinger, 2004).

If β-arrestins compete with G proteins for the receptor before G protein activation, G protein-independent β-arrestin-related pathways will be activated. If β-arrestins compete with G proteins for the receptor after G protein activation, then G protein-dependent β-arrestin-related pathways will be activated. Regardless of this competition, translocation of the β-arrestins to the receptor disrupts G protein coupling, dampens the signal transduction processes, and promotes receptor internalization by targeting the receptor to clathrin-coated pits (Chu et al., 2010; Connor et al., 2004; Zhang et al., 2009). Intracellular signaling stimulated by β-arrestins then designate whether the receptor is recycled back to the membrane or is targeted for degradation by lysosomes. As such, β-arrestins can act as negative regulators of GPCR-mediated signaling.

In addition to desensitizing GPCR signaling, β-arrestins can also function as a scaffolding molecule for the recruitment of signaling molecules. β-arrestins recruit a variety of endocytic proteins and signaling molecules to the receptor, thereby linking GPCRs to additional signaling pathways. Signaling components that are recruited via β-arrestin include components of the JNK3 and ERK1/2 MAP kinase cascades, p53 and NFκB pathways, Akt, and the phosphodiesterase 4D3 and 4D5 isoforms of cAMP (Gintzler and Chakrabarti, 2006; Ma and Pei, 2007; Moorman et al., 2009). Cytoplasmic β-arrestins may also interact with regulators of transcription factors, such as the E₃ ubiquitin ligase MDM2 and the NFκB-inhibitor IκBα, indirectly regulating transcription. Indeed, previous studies have shown that opioid receptor activation stimulates the cytoplasmic retention of MDM2 in a β-arrestin/MDM2/receptor complex, reducing nuclear expression of MDM2 and resulting in increased p53-dependent transcription and apoptosis. Similarly, interaction with β-arrestin 2 prevents phosphorylation and degradation of IκBα, thereby attenuating activation of NFκB and transcription of NFκB target genes. Direct epigenetic regulation by β-arrestins has also been shown by opioid receptor activation. Following activation of DOR and KOR, but not MOR, β-arrestin 1 translocates into the nucleus where it accumulates at the p27 and FOS promoters. It then recruits the histone acetyltransferase p300, promoting histone H4 hyperacetylation at the promoter sites, activating their transcription (Ma and Pei, 2007). It is thought that the distinction between DOR, KOR, and MOR activation is due to the high affinity of DOR and KOR, but not MOR, for β-arrestin 1. This is due to the fact that while β-arrestin 1 and β-arrestin 2 are highly similar, β-arrestin 2 possesses a strong nuclear export signal in its C terminus, which hinders its retention in the nucleus. Given this fact, β-arrestin 1 may play a more important role in GPCR-mediated nuclear signaling and preferential activation of particular β-arrestin-dependent signaling and epigenetic pathways is dependent on which member of the β-arrestin family is recruited by a given GPCR (Kang et al., 2005; Ma and Pei, 2007). Overall, an important distinction to be made in β-arrestin signaling is that while most G protein-dependent pathways are rapid and transient, the majority of β-arrestin-dependent pathways are slower in onset and prolonged. Therefore, β-arrestin binding of GPCRs is not only a regulatory mechanism of G protein signal transduction but also serves as a critical initiator of a second-wave of prolonged signal transduction as well as epigenetic and transcriptional regulation (Ma and Pei, 2007).

The competition between G proteins and β-arrestins determines the selection of the downstream pathways. Receptor phosphorylation can increase the affinity of the receptor for β-arrestins, promoting the activation of β-arrestin dependent pathways over G protein pathways (Zheng et al., 2010). Specific residues of the intracellular loops and C-terminal tail of GPCRs, including the MOR, are phosphorylated in response to agonist-binding by one of two types of serine/threonine protein kinases, G protein-coupled receptor kinases (GRKs) or second messenger-activated kinases, which include PKA, PKC, Src, MAPK, and Ca²⁺/calmodulin-dependent protein kinase II (Chu et al., 2010; Clayton et al., 2010; Feng et al., 2011). GRK-mediated phosphorylation of the agonist-occupied receptor subsequently increases the affinity of the agonist–receptor complex for β-arrestins. The functions of second-messenger-mediated phosphorylation are still unclear (Feng et al., 2011). Therefore, C-terminal phosphorylation site specificity contributes to the specificity of β-arrestin binding. As such, MOR signaling pathways are dependent on the extent to which the agonist-receptor complex is phosphorylated and the subsequent recruitment of β-arrestins or other secondary messengers. Interestingly, morphine exposure generally results in less phosphorylation of the MOR than other, more efficacious agonists, such as DAMGO. This may be due to specific agonist-induced phosphorylation at different residues (Johnson et al., 2005). Therefore, weak phosphorylation-inducing agonists, such as morphine, might desensitize the receptor via pathways other than those involving GRK and β-arrestins (Chu et al., 2010; Zheng et al., 2010). Likewise, these agonists will not induce β-arrestin specific signaling components.

Alternative splicing of the μ-opioid receptor: Functional consequences

Multiple studies in mice, rats, and humans have shown that the MOR pre-mRNA undergoes extensive splicing to give multiple isoforms (Bare et al., 1994; Kvam et al., 2004). Both C-terminal (Bare et al., 1994; Doyle et al., 2007; Pan et al., 1999; Pan et al., 2005b; Pan et al., 2003; Pan et al., 2000) and N-terminal (Pan, 2002; Pan et al., 2001; Xu et al., 2009) variants have been identified, all shown to have distinct regional, cellular, and subcellular distributions (Abbadie et al., 2000a; Abbadie et al., 2000b; Abbadie et al., 2000c; Abbadie et al., 2004; Abbadie et al., 2001; Abbadie et al., 2002; Pasternak, 2010; Ständer et al., 2002; Zhang et al., 2006). Some general studies have shown differences in ligand binding affinity between these receptors (Bolan et al., 2004; Oldfield et al., 2008; Pan et al., 2005a; Pasternak et al., 2004; Pasternak, 2004; Ravindranathan et al., 2009). Furthermore, these receptors seem to have different rates of [³⁵S]GTPγS binding, internalization, and resensitization in response to different μ-selective ligands (Abbadie and Pasternak, 2001; Koch et al., 2001; Koch et al., 1998; Mizoguchi et al., 2003; Tanowitz et al., 2008). Studies in mice suggest that these isoforms have functional significance. Some studies on exon-specific KO studies in mice have found that sequences encoding for particular exons are an important molecular target for the activation of G-proteins by MOR agonists and are required for morphine-induced nociception in the spinal cord (Mizoguchi et al., 2002a; Mizoguchi et al., 2002b; Mizoguchi et al., 2003). However, there are discrepancies between studies, with some studies showing retention of nociception with heroin and M6G (a morphine metabolite) in exon KO mice (Schuller et al., 1999). These differences arise from different knockout methods and it is suggested that other exons near the KO exon were also disrupted (Mizoguchi et al., 2003). As such, functional differences observed among different exon KO mice strains are attributed to alterations of additional exons, suggesting functional roles for individual isoforms incorporating these exons. Therefore, closer examinations of the functional significance of individual exons of OPRM1, as well as the isoforms that incorporate them, are needed.

Despite the functional significance of individual exons, examination of the downstream functional consequences of individual MOR isoforms is scarce. However, one functional study of the truncated variant MOR-1K showed interesting results. As previously described, the classical MOR, or MOR-1, is a G_i/G_o coupled 7-transmembrane receptor that is expressed on the cell membrane, activation of MOR-1 resulted in inhibition of adenylyl cyclase and Ca²⁺ channels, among other results (Law et al., 2000). The alternatively spliced MOR-1K receptor does not fit this model in that it is a 6-transmembrane receptor sequestered in the intracellular compartment. Despite this lack of membrane expression, MOR-1K is still functional and is primarily coupled to G_s proteins, in contrast to MOR-1. As such, activation of MOR-1K was shown to increase adenylyl cyclase and Ca²⁺ channel activity (Gris et al., 2010). Unfortunately, no investigations on the downstream signal transduction pathway stimulated by MOR-1K activation has been conducted.

Although MOR-1K is only a single example of an MOR isoform with alternative functions, it has a major consequence on our understanding of opioid signaling since opioids may now have an excitatory or inhibitory effect based on which isoform is stimulated. Given that each region of a given GPCR has important functional roles, it is likely that alternatively spliced isoforms are functionally different from each other in numerous ways. Furthermore, the differential expression of MOR isoforms may have functional consequences, given that agonist selectivity and C-terminal phosphorylation site availability contributes to the specificity of receptor phosphorylation and the recruitment of β-arrestins or other secondary messenger proteins. Therefore, alternative splicing resulting in an isoform with altered agonist selectivity or an alteration of the C-terminal phosphorylation sites will affect how the isoform is phosphorylated, the competition between G proteins and β-arrestins, and ultimately which signaling pathways are stimulated. Currently, however, there is little to no data on the signal transduction mediated by individual MOR isoforms or how differential regulation of their expression alters the overall signal transduction of opioids (Table I).

Regulation of MOR alternative splicing by signal transduction

Normally, the regulation of splicing is achieved by SR proteins, which bind to areas within and flanking alternative exons of pre-mRNA, promoting or inhibiting assembly at nearby splice sites. Changing the activity of these proteins by posttranslational modifications alters splicing regulation. Interestingly, a study in human blood lymphocytes found that the MOR isoform MOR-1A is specifically and significantly up regulated in methadone maintenance treated individuals while MOR-1O is specifically and significantly down regulated (Vousooghi et al., 2009). Regulation of OPRM1 promoter activity alone does not account for this differential regulation of MOR isoform expression. Therefore, in addition to regulating epigenomic and translational activity, opioids may also regulate alternative splicing mechanisms.

To date, there have been no studies investigating the mechanisms involved in the regulation of MOR alternative splicing and, as such, the effects of opioids on this process are completely unknown. Interestingly, however, one mechanism of post-translational modification of SR proteins is phosphorylation by Akt, which alters the role of SR proteins in both splicing and translation (Blencowe and Graveley, 2007). Since the activation of the MOR, as well as other GPCRs, has been shown to modulate Akt, it is possible that the MOR may self-regulate its own splicing via modulation of Akt-dependent phosphorylation of regulatory SR proteins. Many additional cellular signaling pathways stimulated by MOR activation, such as Ras/MEK/ERK, MAPK, Ca²⁺/calmodulin/CaMK IV, and Rac/JNK/p38 have also been shown to regulate alternative splicing (Tarn, 2007). However, none of these mechanisms have been implicated specifically in regulating alternative splicing of the MOR pre-mRNA (Figure 2).

Epigenetic regulation of OPRM1 expression

Epigenetic regulation involves reversible, heritable changes and chromatin modifications without DNA alterations and results in altered gene activity. Differential expression of the opioid receptors during development and adulthood requires epigenetic regulation. DNA methylation and histone modifications have been found in the regulatory regions of all three genes encoding for the different opioid receptor subtypes (Figure 2). During embryonic development, the OPRM1 gene is completely silenced while expression of genes encoding for the KOR and the DOR are gradually increased. During neuronal differentiation there is a shift in this expression, with the KOR and DOR encoding genes being silenced and a robust activation of OPRM1. This is followed by a reactivation of the KOR and DOR genes (Wei, 2008).

Role of DNA methylation

Epigenetic regulation of OPRM1 expression in mice has been shown to be mediated by MeCP2, through its association with the chromatin-remodeling factor Brg1 and the DNA methyltransferase Dnmt1. This regulation occurs through DNA methylation, demethylation, and subsequent chromatin remodeling of the OPRM1 promoter in various CNS regions. When methylated, CpG regions of the promoter facilitate MeCP2 binding. MeCP2 then recruits repressor proteins that deacetylate associated histones, forming a compact structure around the promoter region and silencing OPRM1 transcription. Demethylation of these CpG regions causes MeCP2 and associated repressor proteins to dissociate from the promoter region, facilitating OPRM1 transcription (Hwang et al., 2010; Hwang et al., 2007; Hwang et al., 2009; Meaney and Ferguson-Smith, 2010).

Promoter methylation of the human OPRM1 gene has also been demonstrated to play a role in its regulation. It was demonstrated that the OPRM1 promoter region is CpG methylated in non-expressing neuronal cells while it is unmethylated in expressing neuronal cells (Andria and Simon, 1999). Hypermethylation of two CpG sites in the OPRM1 gene promoter has been demonstrated in peripheral lymphocytes of Caucasian methadone maintained former heroin addicts (Nielsen et al., 2009). The same group has also demonstrated ethnic diversity in the DNA methylation of the OPRM1 gene promoter. In African-Americans, the degree of methylation was significantly decreased, while in Hispanics, the degree of methylation was increased in former heroin addicts (Nielsen et al., 2010). Although these findings do not definitively prove that opioids alter the epigenetic regulation of the OPRM1 gene, they are interesting given the known roles of KORs and DORs in β-arrestin dependent epigenetic regulation (Kang et al., 2005; Ma and Pei, 2007).

Role of Histone modification

The other major form of epigenetic regulation is mediated by posttranslational modification of DNA-associated histone proteins in chromatin. Notably, histone acetylation by histone acetyltransferase (HAT) and deacetylation by histone deacetylases (HDACs) play a crucial role in the regulation of gene expression. Indeed, it has been demonstrated that HDAC1 and HDAC2 are recruited to the proximal region of the OPRM1 promoter in human neuronal NMB cells. Additionally, the HDAC inhibitor trichostatin A (TSA) affected the OPRM1 gene expression both at the transcriptional and post-transcriptional levels (Lin et al., 2008). A recent study demonstrated that cycloheximide induced the murine OPRM1 gene in an epigenetic fashion with increased recruitment of acetylated histone H3 and methylated H3-K4 as well as a concomitant decrease in HDAC2 binding and H3-K9 methylation on the promoter (Kim et al., 2011).

Role of MicroRNA

MicroRNAs (miRNAs) are noncoding RNAs, generally 20–30 nucleotides in length, that are processed and assembled into ribonucleoprotein (RNP) complexes called micro-RNPs (miRNPs) or into miRNA-induced silencing complexes (miRISCs) (Carthew and Sontheimer, 2009; Filipowicz et al., 2008; Zhang and Su, 2009). miRNA complexes primarily reside in cytoplasmic RNA processing bodies (P-bodies) where they inhibit gene expression at the post-transcriptional level (Zhang and Zeng, 2010). This is mediated by the binding of miRNA complexes to miRNA binding sites generally located within the 3’-untranslated region (3’-UTR) region of the mRNA transcript. The mechanism of regulation is determined by the degree of miRNA-mRNA complementarity, with perfect complementarity promoting cleavage and incongruity promoting repression of the mRNA. Whether this cleavage or repression occurs at translation initiation or post-initiation is still subject to debate. Alternatively, miRNAs complexes may also regulate gene expression at the epigenetic level. Epigenetic regulating miRNAs, dubbed epi-miRNAs, modulate gene expression indirectly by targeting components of epigenetic machinery, including DNA methyltransferases and enzymes regulating histone acetylation (Carthew and Sontheimer, 2009; Iorio et al., 2010).

The presence of a long ′3 -UTR in ORPM1 mRNA (Han et al., 2006; Ide et al., 2005) suggests that it may contain regions that could play a role in the regulation of receptor expression. The role of miRNA in the regulation of murine MOR is seen in the fact that miRNA23b blocks the association of MOR1 mRNA with polysomes, leading to inhibition of MOR1 protein translation (Wu et al., 2008). Additionally, the miRNA let-7 promotes translocation and sequestration of MOR mRNA to P-bodies, leading to translation repression. As such, anti-let-7 treatment was shown to decrease brain let-7 levels and partially attenuate opioid antinociceptive tolerance in mice, suggesting a direct role of miRNA regulation in MOR function (He et al., 2010).

Regulation of OPRM1 and MOR expression by Opioids and HIV-1

Given that the activity of the MOR is tightly regulated by the activation and transcription of OPRM1 (Xu and Carr, 2001b), a particularly important question is what environmental factors alter the epigenome and transcription of OPRM1. The up-regulation of MORs by pro-inflammatory cytokines like IL-1, IL-6 and TNFα, as well as anti-inflammatory cytokines, like IL-4, in neurons suggests that these factors mediate epigenetic and transcriptional regulation of the OPRM1 gene(Kraus, 2009). A study in Jurkat T cells showed that IL-4 stimulation leads to phosphorylation and activation of STAT6. Remodeling of the chromatin structure, including histone modifications, is then signaled by IL-4 activated STAT6. This remodeling is most likely necessary to switch the OPRM1 gene from a heterochromatic, or inactive, state to a euchromatic, or active, state. Additionally, IL-4 induces the disassociation of MeCP2 from the OPRM1 promoter, demethylating the promoter. Following these events, STAT6 binds to a regulatory DNA element and facilitates its transcription (Kraus et al., 2010). Additional regulatory mechanisms involve IL-1 and TNFα, mediated by AP-l and/or NF-κB, and IL-6, mediated by STAT1 and/or STAT3 (Kraus, 2009).

These mechanisms of OPRM1 regulation are interesting since both morphine and HIV-1 have been shown to alter cytokine release (Bonnet et al., 2008; Turchan-Cholewo et al., 2009). It has previously been shown that HIV-1 infection increases MOR mRNA expression (Beltran et al., 2006). However, regulation of MOR expression through an IL-4/STAT6-dependent pathway has not been shown in HIV-1 infection. Rather, HIV-1 infection has been shown to regulate OPRM1 expression in an SP1-dependent manner, although the mechanism of this regulation is not completely understood (Liu et al., 2009). Likewise, MOR ligands have been shown to alter MOR expression (Zadina et al., 1993). It was previously shown that DAMGO increases binding of SP1 and SP3 to the OPRM1 promoter region. It is likely that this is mediated through the cAMP-PKA pathway, however an exact mechanism has not been determined. Regardless, this DAMGO-mediated increase in SP1/SP3 binding may modulate transcription (Xu and Carr, 2001a). In addition to this mechanism, a study in CEM × 174 cells suggests a role for PI3K and Akt in morphine-induced initiation of OPRM1 transcription. Here the authors suggest that the transcription factor E2F1, a downstream effector of Akt, along with additional transcription factors Sp1 and YY1, interact with the promoter region of the OPRM1. In short, morphine treatment was shown to promote nuclear translocation of E2F1 and enhance E2F1 gene expression through a PI3K/Akt dependent pathway. This enhanced E2F1 interacted directly and indirectly with Sp1 and YY1, respectfully, to bind to the OPRM1 promoter and facilitate transcription. Therefore, in addition to epigenetic regulation by cytokines and interleukins during HIV-1 infection, morphine may directly regulate transcription of the OPRM1 gene through either a cAMP/PKA/SP1/SP3 or PI3K/Akt/E2F1 pathway (Li et al., 2008; Liu et al., 2010).

Opioid abuse may also modulate MOR mRNA transcript levels specifically and directly by miRNA mechanisms. A recent study demonstrated that morphine significantly up regulates let-7 miRNA expression in SH-SY5Y neuronal cells(He et al., 2010). Additionally, morphine treatment of primary human monocyte-derived macrophages leads to differential regulation of miRNA expression. Among the miRNAs differentially regulated were hsa-miR-15b and hsa-miR-181b, both of which have several targets in pro-inflammatory pathways (Dave and Khalili, 2010). Therefore, opioid abuse may regulate MOR expression through both direct miRNA-dependent inhibition of MOR mRNA or indirectly by miRNA modulation of pro-inflammatory or other signal transduction pathways (Figure 2).

Viruses, such as HIV-1, are dependent on intracellular machinery to facilitate their own replication. As such, many viruses, including HIV-1, have developed mechanisms to hijack cellular processes to facilitate replication while evading immune responses. Those viruses that cause latent infection benefit from heritable, epigenetic changes in host transcription and, therefore, have evolved mechanisms of modulating host epigenetic, transcriptional, and translation regulation. HIV-1 has been shown to increase cellular DNA methyltransferase activity to produce a general increase in cellular DNA methylation (Paschos and Allday, 2010). Furthermore, a miRNA generated by the HIV-1 TAR element specifically recruits HDAC-1 in order to silence its own transcription (Klase et al., 2007; Ouellet et al., 2009). While this HIV-1-generated miRNA may silence transcription for the purpose of driving latency, HIV-1 replication may be actively suppressed by cellular miRNAs. HIV-1 mRNA expression is modulated by RISCs and P bodies, and a depletion of P bodies, as well as cellular anti-HIV-1-miRNAs, enhances HIV-1 production and infection both in vitro and in vivo (Nathans et al., 2009; Wang et al., 2009). Interestingly, of the five known anti-HIV-1-miRNAs highly expressed in human monocytes, morphine treatment reduces the expression of four. This effect was specific to anti-HIV-1-miRNAs, as no change was seen in the expression of known anti-HCV-miRNA (Wang et al., 2011). Therefore, in addition to SP1-dependent modulation of OPRM1 expression, the existence of HIV-1-generated miRNAs, and their interaction with the epigenetic, transcriptional, and translational machinery, leave open the possibility of direct or indirect modulation of OPRM1 and MOR expression by HIV-1. Furthermore, there is clear evidence that opioids modulate HIV-1 replication by inhibiting anti-HIV-1-miRNAs. As such, alterations in OPRM1 and MOR expression may have functional consequences in HIV-1 pathogenesis due to changes in anti-HIV-1-miRNA regulation.

Opioid abuse enhances HIV-1-associated neurocognitive dysfunction by stimulating overlapping biochemical responses in the CNS

Nervous system disorders caused by HIV-1 infection, such as HAND, remain a significant problem among HIV-1 infected individuals, occurring in approximately 30–50% of HIV-1-infected persons in the United States(Power et al., 2009; Rappaport and Berger, 2010). While any cognitive ability can be compromised in HAND, disturbance in memory, particularly in learning and retrieval of new information, tends to be prominent. Many individuals also exhibit psychomotor slowing, problems in attention, and disturbance in executive functions(Grant, 2008). While it is well accepted that HAND results from HIV-1-mediated neuronal damage, exact mechanisms have yet to be fully established.

Currently, there are two models for the development of neurocognitive dysfunction in HIV-1. The indirect model proposes that neuronal dysfunction is mediated by the inflammatory and excitotoxic responses mounted by infected and uninfected glial cells in response to HIV-1 infection or HIV-1 viral proteins. Microglia, the resident macrophage of the CNS, are productively infected with HIV-1 and are the major source of toxic cellular proteins, including TNF-α, interleukins, and nitric oxide, in addition to serving as viral reservoirs (Fischer-Smith and Rappaport, 2005; Garden, 2002; Lindl et al., 2010; Louboutin et al., 2010; Yadav and Collman, 2009). Additionally, HIV-1 infection increases cytokine and reactive oxygen species (ROS) release by astrocytes, despite the absence of infection in these cells (Kaul et al., 2001; Turchan-Cholewo et al., 2009). As such, HIV-1-mediated neurotoxicity may be secondary to the inflammatory response mounted by glial cells, resulting in a simultaneous loss of trophic support and an increasingly neurotoxic environment (Hauser et al., 2006). Alternatively, the direct model proposes that HIV-1-mediated neuronal dysfunction is mediated by viral proteins, such as gp120, Tat, and Vpr, secreted by actively infected glial cells, such as macrophages and microglia. These secreted viral proteins directly interact with uninfected neurons to promote neuronal apoptosis (Lindl et al., 2010). Although these models present two separate and distinct pathways of HIV-1-mediated neurotoxicity, they are not mutually exclusive. Therefore, neuronal dysfunction present in HIV-1 infection is most likely due to a combination of both direct and indirect insults.

It is thought that opioid abuse enhances HIV-1-mediated neurotoxicity by sensitizing neurons to both direct and indirect insults mediated by HIV-1 infection. Opioids modify multiple aspects of microglial function, including phagocytosis, chemotaxis, and chemokine production (El-Hage et al., 2006). Likewise, opioids increase cytokine and chemokine production in astrocytes (El-Hage et al., 2008). In this manner, opioid abuse may exacerbate the inflammatory environment common in HIV-1 infection by further increasing chemokine, cytokine, and ROS production by glial cells. However, opioid enhancement of HIV-1-mediated neurotoxicity is more commonly attributed to direct interactions between HIV-1 viral proteins, opioids, and neurons. Multiple pathways activated by opioids and HIV-1 viral proteins intersect as various levels, allowing for HIV-1-opioid crosstalk. This includes the recruitment of β-arrestin, which augments both MOR and CXCR4 activation of p38 and ERK and regulates downstream MAPK signaling (Hauser et al., 2006). Another point of HIV-1 and opioid intersection is the PI3K/Akt/GSK3β pathway, which, in addition to modulating transcriptional and alternative splicing mechanisms, facilitates apoptotic signaling. PI3K-activated Akt enhances neuronal stability and survival during stress, particularly in inflammation, by diminishing neuronal damage caused by interleukins, nitric oxide, TNF-α, and other inflammatory molecules. Although Akt mediates this process through multiple mechanisms, it has most notably been shown to inactivate the pro-apoptotic proteins glycogen synthase kinase-3 (GSK3) and Bad by phosphorylating them. Activation of Akt also removes IKK inhibition of NF-κB, activating anti-apoptotic genes. Additionally, Akt can block caspase activation within the CNS both directly, by inhibition of caspase 9, or indirectly through the modulation of caspase substrates (Chong et al., 2005; Song et al., 2005). Morphine has been shown to dephosphorylate Akt, inactivating it and resulting in microglial apoptosis (Xie et al., 2010). Interestingly, it has been shown that both gp120 and Tat activate GSK3β in neurons, promoting apoptosis (Hauser et al., 2006). Therefore, in addition to MAPK signaling, opioid enhancement of HIV-1-mediated neurotoxicity may, in part, be due to the combination of Akt inhibition by morphine and GSK3β activation by HIV-1 viral proteins, resulting in increased neuronal apoptosis.

Conclusion

It is generally accepted that neurological complications caused by HIV-1 infection, such as HAND, are enhanced by opioid abuse. Although this may occur indirectly by HIV-1 and opioid mediated glial dysfunction, opioid enhancement of HAND is regularly attributed to the direct, combinational effect of HIV-1 viral proteins and opioids stimulating overlapping apoptotic pathways in neurons. While exact mechanisms of this interaction are the subject of intense investigation, an equally important mechanism in this interaction, the regulation of MOR expression by HIV-1 and opioids, is severely lacking.

The activity of the MOR is tightly regulated by the transcription of OPRM1, of which the epigenetic and transcriptional regulatory mechanisms have been partially characterized. Many of these mechanisms overlap with known opioid and HIV-1 viral protein signaling cascades as studies have shown that both modulate MOR expression. Exact mechanisms of OPRM1 regulation are not fully characterized but may include miRNA regulation as well as transcriptional, translational, and epigenetic modulation.

Extensive alternative splicing of the MOR pre-mRNA has been shown to give multiple isoforms of the MOR. Given the importance of multiple GPCR regions in receptor function, alterations of these regions by alternative splicing may generate functionally distinct isoforms. This is particularly true for C-terminal variants, since site-specific phosphorylation at conserved residues regulates the recruitment of second messenger proteins and the signal transduction mechanism. Some differences have been observed between MOR isoforms, including both excitatory and inhibitory functions, suggesting that MOR isoforms individually contribute to opioid signaling, including opioid enhancement of HIV-1 neuropathogenesis. Unfortunately, the signal transduction mediated by individual isoforms is not understood.

In conclusion, opioid abuse has become a major public health concern due to the increased risk of HIV-1 infection as well as the exacerbation of HIV-1 neuropathogenesis. Complicating the understanding of HIV-1 and opioid interactions is the dynamic interplay between opioid abuse, HIV-1 infection, OPRM1 regulation, and the generation and differential expression of functionally distinct MOR isoforms by alternative splicing. Modulation of the mechanisms regulating the epigenetic and transcriptional expression of OPRM1, and/or splicing of MOR pre-mRNA, will alter the MOR isoforms expressed. This may, in turn, alter the overall physiological affect of opioids. Although mechanisms regulating all of these processes have been partially characterized, our understanding of how opioids or HIV-1 modulate these mechanisms is incomplete. Furthermore, there has been no investigation to date examining the signal transduction mechanisms mediated by individual MOR isoforms or their individual role in opioid driven pathogenesis, such as HAND. As such, extensive studies are needed first to understand how opioid abuse and pathological conditions, such as HIV-1 infection, modulate the epigenetic, transcriptional, and post-transcriptional regulation of MOR expression. Second, how MOR isoforms individually contribute to the signal transduction of different opioids needs to be investigated. Finally, studies are needed to determine how the differential regulation of MOR isoforms alters the physiological response to opioids in both homeostatic and pathological conditions.