Mechanisms of HIV transcription regulation by drugs of abuse

Abstract

There is a strong correlation between the use and abuse of illicit drugs and the spread of Human Immunodeficiency Virus (HIV). It is well established that illicit drugs users are a high risk population for infection with HIV with an increase rate of HIV transmission and replication. Cocaine, amphetamine, methamphetamine, heroin and morphine stand out as the most frequently abused illicit drugs and their use correlates well with HIV infection and AIDS progression. Notably, major incidences of HIV infection in illicit drug abusers are primarily due to high risk activities such as needle sharing and unprotected sex. Several studies have demonstrated that drugs of abuse increase the overall viral load by enhancing HIV replication in patients, in particular in the central nervous system (CNS). The CNS is a common target for both drugs of abuse and HIV, and their synergistic action accelerates neuronal injury and cognitive impairment. In order to generate complete genomic transcripts, HIV gene expression has to go through both the initiation and elongation phases of transcription, which requires coordinated action of different transcription factors. In this review, we will provide the latest updates of the involved molecular mechanisms that regulate HIV transcription and discuss how drugs of abuse, such as cocaine, amphetamine, methamphetamine, heroin and morphine, modulate those mechanisms to upregulate HIV transcription and eventually HIV replication.

INTRODUCTION

The drugs of abuse, morphine/heroin, cocaine and amphetamine/methamphetamine, have different mechanisms of action, as they exert their effects through different cellular receptors. Whereas, the downstream effects of these drugs are quite similar, even in different cell types. Most of these drugs enhance HIV replication, promote glial cell activation, and lead to blood brain barrier (BBB) deterioration. The underlying molecular mechanisms and factors responsible for these deleterious effects are still not well defined. The following represents the state of our knowledge regarding these drugs and, particularly, their interaction with HIV.

The CNS is the common target for drugs of abuse and HIV

According to a recent estimate, approximately 40 million people worldwide are infected with HIV (Tanne 2006, Buch, Yao et al. 2011). Illicit drug use contributes significantly to HIV infection and transmission (Nath, Hauser et al. 2002, Buch, Yao et al. 2011, Alfahad and Nath 2013). In the United States, infection in drug abusers accounts for one-third of all new cases of HIV infection (Ferris, Mactutus et al. 2008, Buch, Yao et al. 2011). Cocaine, amphetamine (AMPH), methamphetamine (METH), heroin and morphine are the most frequent drugs of abuse, and they have been implicated as major contributing factors for HIV infection, transmission and AIDS progression (Larrat and Zierler 1993, Peterson, Gekker et al. 1993, Donahoe and Vlahov 1998, Fiala, Gan et al. 1998, Webber, Schoenbaum et al. 1999, Nath, Maragos et al. 2001, Nath, Hauser et al. 2002, Chang, Ernst et al. 2005, Hauser, El-Hage et al. 2006, Scott, Woods et al. 2007, Nair, Saiyed et al. 2009, Buch, Yao et al. 2012, Gaskill, Calderon et al. 2013). Both drugs of abuse and HIV target the central nervous system (CNS). Within a few days of infection, HIV establishes a reservoir in the CNS by infecting different kinds of brain cells including microglial cells, circulating macrophages, lymphocytes and astrocytes (Kaul, Garden et al. 2001, Roth, Tashkin et al. 2002, Gekker, Hu et al. 2006, Kaul and Lipton 2006, Nath and Clements 2011, Mantri, Pandhare Dash et al. 2012, Kim, Jung et al. 2013, Napuri, Pilakka-Kanthikeel et al. 2013). Viral replication and expression of viral proteins (e.g. Tat and gp120), and the resulting immune response impair the functioning of these cells leading to the deterioration of both the immune and nervous systems (Bagasra and Pomerantz 1993, Klein, Matsui et al. 1993, Peterson, Gekker et al. 1993, Mao, Huang et al. 1996, Eisenstein and Hilburger 1998, Bansal, Mactutus et al. 2000, Nath, Maragos et al. 2001, Nath, Hauser et al. 2002, Nair, Mahajan et al. 2005, Giunta, Obregon et al. 2006, Kaul and Lipton 2006). Illicit drugs promote HIV replication and accelerate HIV-associated neuropathology (neuro-AIDS) which is clinically recognized as HIV-associated neurocognitive disorder (HAND) (Baldwin, Roth et al. 1998, Nath, Maragos et al. 2001, Hauser, El-Hage et al. 2006, Antinori, Arendt et al. 2007, Alfahad and Nath 2013). Despite the success of highly active antiretroviral therapy (HAART) in controlling circulating HIV, HAND remains a significant co-morbidity responsible for deterioration in quality of life and enhanced mortality among HIV-infected patients (Nath, Maragos et al. 2001, Pandya, Krentz et al. 2005, Alfahad and Nath 2013). The consequence of this disorder is an extraordinary health and financial burden on HIV patients and on society (Nath, Maragos et al. 2001, Cook, Burke-Miller et al. 2008).

Drugs of abuse enhance HIV replication

Numerous ex vivo studies have shown that different drugs of abuse, mainly cocaine, AMPH, METH, heroin and morphine, enhance HIV replication (Peterson, Gekker et al. 1991, Bagasra and Pomerantz 1993, Nath, Hauser et al. 2002, Ellis, Childers et al. 2003, Kumar, Torres et al. 2004, Dhillon, Williams et al. 2007, Nath 2010, Wires, Alvarez et al. 2012). Similarly, in several experimental models of HIV infection, drugs of abuse have been shown to accelerate HIV replication and the concomitant loss of CD4+ T cells (Nath, Maragos et al. 2001, Kumar, Kumar et al. 2002, Nath, Hauser et al. 2002, Roth, Tashkin et al. 2002, Kumar, Torres et al. 2004, Roth, Whittaker et al. 2005, Nath 2010, Buch, Yao et al. 2011). Most of these studies have shown that the increase in the concentration of HIV was due to an augmented rate of HIV infection and transmission following drug exposure. In addition to these studies, it has been recently argued (Wires, Alvarez et al. 2012, Sahu, Farley et al. 2015) that drugs of abuse also enhance HIV replication by increasing the rate of HIV transcription and generating higher numbers of complete HIV genomic transcripts.

HIV transcription

HIV replication relies on efficient transcription to generate full genomic transcripts. HIV performs its transcription mainly by using the host cell transcription machinery with the help of its own master transactivator protein Tat. In order to generate complete HIV genomic transcripts, transcription has to successfully pass through both the initiation and elongation phases. We have recently demonstrated that cocaine promotes HIV transcription and replication by activating different transcription factors that are required for HIV gene expression, besides inducing the establishment of a transcriptionally active chromatin structures at HIV LTR (Figure 1).

Figure 1. Schematic diagram showing that drugs of abuse, specifically cocaine and methamphetamine, accelerate HIV transcription; especially cocaine enhances HIV transcription by augmenting both initiation and elongation phases.

Different drugs of abuse have been shown to activate NF-κB (discussed in text), but only the direct impact of METH and cocaine on HIV transcription has been documented so far. In a recent study, METH activated NF-κB has been shown to directly promote HIV transcription. Similarly, by performing a comprehensive analysis we define the underlying molecular mechanism which cocaine utilizes to promote both initiation and elongation phases of HIV transcription. Cocaine treatment results in the activation and recruitment of NF-κB and MSK1 at HIV LTR. MSK1 subsequently catalyze the phosphorylation of NF-κB at its serine residue 276. Phosphorylation of NF-κB at 276^th serine residue enhances the functional capability of NF-κB by augmenting the interaction of NF-κB with HATs. HATs such as p300 in turn catalyze the acetylation of core histones (mainly H3 and H4) at HIV LTR and facilitate the establishment of transcriptionally active chromatin structures, which boost the access of transcription machinery to HIV LTR and thus facilitate the initiation phase of HIV transcription. MSK1 in addition to phosphorylating NF-κB catalyzes the phosphorylation of histone H3 at serine 10 (H3S10). Phosphorylated histone H3 at 10^th residue contributes to the establishment of transcriptionally active chromatin structure at HIV LTR. Moreover, phosphorylated H3S10 also facilitates the recruitment of P-TEFb at HIV LTR through a mechanism that is not yet fully defined; usually with the help of BRD proteins which bind specifically to acetylated histones through their bromo domain. P-TEFb, an established elongation factor, subsequently catalyzes the phosphorylation of several proteins, including RNA Polymerase II (RNAP II) and negative elongation factors. These modifications enhance the processivity of RNAP II and nullify the impact of negative elongation factors, and thus promote the elongation phase of HIV transcription.

HIV transcriptional initiation

The initiation phase of HIV transcription involves the binding of transcription factors, such as SP1, TATA-box-binding protein (TBP), and TBP-associated factors (TAFs) to the core long terminal repeat (LTR) promoter, an essential and sufficient component to sustain minimal basal HIV transcription. HIV LTR core promoter consists of TATA box, initiator sequence and three SP1 binding sites (Jones, Kadonaga et al. 1986, Garcia, Harrich et al. 1989, Ross, Buckler-White et al. 1991, Olsen and Rosen 1992, Rittner, Churcher et al. 1995). However, efficient initiation of HIV transcription occurs only after the binding of transcriptional activators, primarily NF-kB, NF-AT, AP-1 and STAT5 to the enhancer sequences of HIV LTR (Nabel and Baltimore 1987, Kinoshita, Su et al. 1997, Yang, Chen et al. 1999, Selliah, Zhang et al. 2006). These factors, after binding to HIV LTR, promote efficient HIV transcription mainly by recruiting histone acetyl transferases (HATs) and through cooperation with SP1 proteins (Perkins, Edwards et al. 1993, Alcami, de Lera et al. 1995, Gerritsen, Williams et al. 1997, Garcia-Rodriguez and Rao 1998, Bosque and Planelles 2008).

Drugs of abuse and HIV transcriptional initiation

Almost all the drugs of abuse are known to enhance HIV replication. Moreover, all the examined drugs of abuse are known to activate NF-κB, a transcription factor that plays critical role in HIV transcriptional initiation, suggesting that drugs of abuse enhance the rate of HIV replication by augmenting HIV transcription. However, very little is known about the molecular mechanisms that different drugs of abuse utilize to enhance HIV gene expression. Recently, we have investigated in detail some of the molecular mechanisms that are modulated by cocaine in order to enhance HIV transcription (Sahu, Farley et al. 2015). Similar studies have been done describing the molecular mechanisms that methamphetamine use to promote HIV transcription (Wires, Alvarez et al. 2012).

NF-κB plays critical role during HIV transcription

The NF-κB/Rel proteins are transcription factors that induce expression of a number of cellular genes. Many of these genes regulate host immunity and inflammatory responses (Baldwin 1996, Ghosh, May et al. 1998). Transcription factor NF-κB consists of either homo- or heterodimers of five related proteins, p65 (Rel A), Rel B, c-Rel, p50/p105 (NF-κB) and p52/p100 (NF-κB2) (Hayden and Ghosh 2004). Of these, one of the best-characterized and functionally most active NF- κB proteins is the heterodimer comprising Rel A (p65) and p50, which is widely expressed and heavily involved in NF-κB-regulated transactivation (Schmitz and Baeuerle 1991, Nath, Maragos et al. 2001). In its inactive state, the NF-κB complexes are sequestered in the cytoplasm through their interaction with inhibitory IκB proteins (Hayden and Ghosh 2004, Sobhian, Laguette et al. 2010). However, upon cell activation through a wide array of stimuli, serine residues 32 and 36 of IκB become phosphorylated by several kinases, including IκB kinase (IKK) complex, which is composed of IKKα, IKKβ and IKKγ (Brown, Gerstberger et al. 1995, Bonizzi, Bebien et al. 2004, Bonizzi and Karin 2004, Hayden and Ghosh 2004) and Ribosomal S6 kinase1 (RSK1). This event induces the ubiquitination of IκB at lysine residues 21 and 22 and leads to its 26S proteasome degradation. Due to the dissociation of IκB, the nuclear localization signal (NLS) of NF-κB proteins become exposed and consequently NF-κB translocates into the nucleus (Kerr, Inoue et al. 1991, Brown, Gerstberger et al. 1995). Once in the nucleus, NF-κB binds to the cognate binding sites at the promoter and enhancer regions of the genes and activates their transcription, including the HIV LTR enhancer region. The HIV enhancer region is one of the best studied elements which contains two NF-kB binding motifs (Nabel and Baltimore 1987, Perkins, Edwards et al. 1993).

Drugs of abuse accelerate HIV transcriptional initiation by activating NF-kB

As mentioned above, very little is known about the molecular mechanisms that different drugs of abuse utilize to enhance HIV transcription. In a recent study we performed such analysis for cocaine (Sahu, Farley et al. 2015). Earlier studies have convincingly demonstrated the activation of NF-κB in cells following cocaine exposure and the concomitant enhanced HIV replication and increased viral load in patients (Hou, Vlaskovska et al. 1996, Ang, Chen et al. 2001, Dhillon, Williams et al. 2007, Yao, Yang et al. 2010, Sahu, Farley et al. 2015). To extend those studies further, using monocytic cell lines (THP1 and U937) and primary monocyte derived macrophages (MDMs), we found that both acute (one time) and chronic (multiple times for 3 days) treatment with cocaine efficiently activated NF-κB (Sahu, Farley et al. 2015). Cocaine not only promotes the activation (nuclear translocation) of NF-κB, but also enhances the functional activity of NF-κB by augmenting the ability of NF-κB to interact with histone acetyltransferases (HATs) (detailed below and in Figure 1). NF-κB activation by cocaine is primarily occurs via RSK1 activation, instead of IKKβ, a kinase specifically activated by TNF-α, which is one of the strongest and most specific activators of NF-κB (Ghosh, May et al. 1998, Ghosh and Hayden 2008, Sahu, Farley et al. 2015). RSK1 is a downstream kinase of the extracellular signal-regulated kinase (ERK) pathway, one of the main pathways through which cocaine exerts many of its effects (Berhow, Hiroi et al. 1996, Valjent, Corvol et al. 2000, Valjent, Pages et al. 2004). Cocaine activated RSK1 primarily phosphorylates IκBβ at Ser19 and Ser23 (our unpublished data), instead of IκBα, which is mainly phosphorylated by IKKβ (Ghoda, Lin et al. 1997). NF-κB activation in addition to activating the transcription of other genes, promotes the transcription of its own inhibitor IκBα. The newly formed IκBα interacts with NF-κB, takes it to the cytoplasm, and thus negatively regulates expression of NF-κB dependent genes. Unlike IκBα, the IκBβ promoter does not have NF-κB binding sites. Consequently, NF-κB activation does not directly lead to the synthesis of this inhibitor. Consequently, RSK1 induced NF-κB activation lasts longer and contributes to the persistence of cocaine effects [our unpublished data and (Thompson, Phillips et al. 1995)].

In addition to the activation of NF-κB, cocaine exposure enhances the functional activity of NF-κB primarily by activating mitogen- and stress-activated kinase 1 (MSK1). MSK1 becomes activated upon the stimulation of the MAPK/extracellular-signal regulated kinase (ERK) cascade (Lu, Koya et al. 2006, Zhai, Li et al. 2008). MSK1 subsequently phosphorylates the p65 subunit of NF-κB at serine residue 276 (p65S276). This post-translational modification boosts the interaction of NF-κB with histone acetyltransferases (HATs) (Perkins, Felzien et al. 1997, Zhong, May et al. 2002, Vermeulen, De Wilde et al. 2003, Sahu, Farley et al. 2015). This enhanced interaction translates into greater recruitment of HATs at HIV LTR through NF-κB binding to its consensus binding sequences at HIV LTR. HATs in turn acetylate the core histones, which reduces the electrostatic interactions between histones and DNA leading to a more relaxed/open chromatin structures around the LTR promoter. These relaxed transcriptionally active chromatin structures consequently further promote access to all the remaining components of the transcription machinery (holo-transcription machinery) and even augment the overall flow of the transcription machinery at the HIV LTR promoter. These events eventually enhance the rate of transcriptional initiation from the LTR promoter (Sahu, Farley et al. 2015). Similarly, the direct role of METH induced NF-κB in enhancing HIV transcriptional initiation has also recently been shown (Wires, Alvarez et al. 2012), Figure 1.

Additional factors induced by drugs of abuse that could promote HIV transcriptional initiation

CREB (cAMP response element binding protein): The psychostimulants (cocaine, AMPH and METH) and opiates, such as morphine increase CREB activity during both the acute and chronic drug treatments (Ma, Zheng et al. 2001, Carlezon, Duman et al. 2005, Edwards, Graham et al. 2007, Briand and Blendy 2010). CREB forms homodimers that bind to genes at cAMP response elements (CREs). HIV LTR has several well defined binding sites for CREB (Krebs, Goodenow et al. 1997, Ross, Nonnemacher et al. 2001). CREB becomes functionally active after its phosphorylation at Ser133. This phosphorylation event enhances the interaction of CREB with CREB-binding protein (CBP), a histone acetyl transferase. CBP in turn promotes the establishment of euchromatin structures at gene promoters, including HIV LTR and promotes transcription (Figure 1) (Mayall, Sheridan et al. 1997, Marzio, Tyagi et al. 1998, Ross, Nonnemacher et al. 2001, McClung and Nestler 2003, Carlezon, Duman et al. 2005, Levine, Guan et al. 2005, Briand and Blendy 2010). However, the significance of CREB induced by drugs of abuse in the context of HIV transcription is still unclear, but our preliminary results suggest that cocaine activated CREB promotes HIV transcription (unpublished data).

AP-1 (activator protein-1): AP-1 is another transcription factor that contributes to HIV transcription and is efficiently activated by different drugs of abuse. AP-1 is mainly formed by heterodimerization between Fos family proteins (c-Fos, FosB, Fra-1, or Fra-2) and Jun family proteins (c-Jun, JunD, or JunB) (Chinenov and Kerppola 2001). AP-1 also cooperates functionally with NF-κB (Rao, Luo et al. 1997, Yang, Chen et al. 1999).The levels of some of the AP-1 subunits (c-Fos, FosB, and JunB) are very rapidly induced by a single acute stimulus of cocaine, but return to normal values within only 8 to 12 hours (Hope, Nye et al. 1994, Moratalla, Elibol et al. 1996, Kuzmin and Johansson 1999). METH has also been shown to activate AP-1 (Wang, Watanabe et al. 1992, Sheng, Ladenheim et al. 1996). All Fos family proteins are induced transiently by acute drug exposure, however chronic administration specifically induces long-lasting expression of δFosB (Nestler 2008, Perrotti, Weaver et al. 2008).The levels of δFosB protein, a naturally occurring, alternatively spliced variant of FosB, which lacks the C-terminal 101 amino acids (Nakabeppu and Nathans 1991, McClung and Nestler 2003, Nestler 2008) remains uniquely elevated for even weeks after the last chronic drug stimulus because of the unusual stability of the δFosB protein (Hope, Nye et al. 1994, Moratalla, Elibol et al. 1996, Ulery, Rudenko et al. 2006, Alibhai, Green et al. 2007, Carle, Ohnishi et al. 2007). δFosB competes for AP-1 binding in the promoter region of responsive genes (McClung and Nestler 2003). In certain genes it acts as an inhibitor by recruiting HDACs (Nakabeppu and Nathans 1991, Yen, Wisdom et al. 1991, Renthal, Carle et al. 2008), while in others it activates their expression by recruiting SWI/SNF machinery (Bibb, Chen et al. 2001, McClung and Nestler 2003, Kumar, Choi et al. 2005, Renthal, Carle et al. 2008, Renthal and Nestler 2008). We have found that cocaine induced AP-1 is an activator of HIV transcription, whereas over expression of δFosB exerts mild inhibition on HIV transcription (unpublished findings).

HIV transcriptional elongation

The efficiency of the elongation phase of HIV transcription is predominantly dependent on the HIV encoded protein transactivator of transcription (Tat). Prior to Tat protein generation, the elongation of HIV transcription proceeds slowly. The inefficient phase of HIV transcription in the absence of Tat is primarily attributed to the presence of non-processive or elongation defective RNA polymerase II (RNAP II) and due to the binding of two general inhibitory factors at HIV LTR: negative elongation factor (NELF) and DRB sensitivity-inducing factor (DSIF) (for recent reviews refer to (Cho, Schroeder et al. 2010, Nechaev and Adelman 2011, Tyagi and Bukrinsky 2012, Taube and Peterlin 2013, Jadlowsky, Wong et al. 2014, Mbonye and Karn 2014). However, once Tat is synthesized and is accumulated in the cell beyond a certain threshold, it positively feedbacks the entire system. Tat binds to an RNA stem loop structure called the trans-activation response (TAR) element, which is present at the 5′ extremity of all HIV transcripts. Tat brings positive transcription elongation factor b (P-TEFb), thus enhances the recruitment of P-TEFb at HIV LTR (Herrmann and Rice 1995, Karn 1999, Peterlin and Price 2006). The cyclin dependent kinase 9 (CDK9) subunit of P-TEFb subsequently hyper-phosphorylates the C-terminal domain (CTD) of RNA polymerase II (RNAPII) and converts the pausing RNAPII into a processive or elongation proficient polymerase (Parada and Roeder 1996, Kim, Bourgeois et al. 2002). Besides RNAPII, P-TEFb also phosphorylates inhibitory factors DSIF and NELF. These modifications either dissociate the negative factors or convert them into a positive transcription factor (Ivanov, Kwak et al. 2000, Bourgeois, Kim et al. 2002, Fujinaga, Irwin et al. 2004). P-TEFb recruitment at LTR is also essential to reactivate latent provirus in primary T cells (Tyagi, Pearson et al. 2010, Budhiraja, Famiglietti et al. 2013). Besides recruiting P-TEFb, Tat also brings additional elongation factors, such as ELL2, AFF4, ENL and AF9 to HIV LTR; together they form a Super Elongation Complex (SEC) (Sobhian, Laguette et al. 2010, He, Chan et al. 2011, Chou, Upton et al. 2013). The enhanced rate of HIV transcription results in generating more Tat protein and higher Tat levels further accelerate HIV transcription. HIV transcription then enters into the fast Tat-dependent phase that eventually accelerates HIV transcription several hundred fold (Karn 1999, Taube, Fujinaga et al. 1999). HIV transcription thus differs from the normal transcription of cellular genes as it is auto-regulated by the Tat protein. These events subsequently relieve all restrictions to HIV transcription, resulting in efficient production of unspliced HIV genomic transcripts which get packaged and lead to the generation of new viral particles (Figure 1) (Kim, Byrn et al. 1989, Pomerantz, Trono et al. 1990).

Drugs of abuse and HIV transcriptional elongation

Drugs of abuse augment the rate of HIV replication, which can be done both by enhancing the rate of infection/transmission and also by increasing the rate of the generation of full-length HIV genomic transcripts. The later mechanism has not yet been well investigated pertaining drugs of abuse. As mentioned earlier that to generate the full genomic transcript, HIV transcription has to pass through both the initiation and elongation phase of HIV transcription. Thus, in addition to NF-κB activation, this implies that drugs of abuse need to activate P-TEFb, an essential factor required for the elongation phase of HIV transcription. In that regard we investigated the role of cocaine in activating and recruiting P-TEFb at LTR.

Cocaine promotes HIV transcriptional elongation primarily by activating MSK1

Cocaine activates MSK1 which, besides catalyzing phosphorylation of the p65 subunit of NF-κB at serine 276, catalyzes the phosphorylation of histone H3 at serine 10 (P-H3S10) at the HIV LTR promoter (Sahu, Farley et al. 2015). Similar findings have been reported for cellular promoters (Soloaga, Thomson et al. 2003). Notably, cocaine also enhances the phosphorylation of p65 and of histone H3 locally at HIV LTR by augmenting the recruitment of MSK1 at LTR (Sahu, Farley et al. 2015). Although several other enzymes besides MSK1 are known to catalyze H3S10 phosphorylation, the predominant role of MSK1 during cocaine exposure has been well established (Brami-Cherrier, Roze et al. 2009, Nestler 2012, Walker, Cates et al. 2014). P-H3S10 is a euchromatic mark which facilitates the establishment of transcriptionally active euchromatin structure (Wei, Garber et al. 1998, Soloaga, Thomson et al. 2003). Accordingly, MSK1 induced H3S10 phosphorylation promotes the establishment of transcriptionally active chromatin structures at HIV-1 LTR following cocaine treatment (Sahu, Farley et al. 2015).

In addition to facilitating the establishment of transcriptionally active chromatin structures, P-H3S10 promotes the recruitment of P-TEFb to HIV LTR analogous to other gene promoters (Ivaldi, Karam et al. 2007, Hu, Lu et al. 2014, Sahu, Farley et al. 2015). As noted above, P-TEFb plays an essential role in supporting the elongation phase of HIV-1 transcription by catalyzing several above mentioned phosphorylation events, RNAP II CTD, NELF, DSIF (Parada and Roeder 1996, Wei, Garber et al. 1998, Ivanov, Kwak et al. 2000, Bourgeois, Kim et al. 2002, Kim, Bourgeois et al. 2002, Fujinaga, Irwin et al. 2004, Peterlin and Price 2006, Karn 2011). The resultant, complete unspliced HIV transcripts are then packaged and lead to the generation of new viral particles; i.e., enhanced HIV replication. Thus, cocaine boosts HIV-1 gene expression by inducing both the initiation and elongation phases of HIV-1 transcription. Our current studies are devoted to verifying that METH, heroin and morphine also promote both the initiation and elongation phases of HIV transcription.

Overview of epigenetic landscape

A majority of the eukaryotic DNA is densely packaged in the form of chromatin structures. The chromatin structures are characterized by their fundamental subunits, nucleosomes. A nucleosome consists of an octamer, a pair of four core histones (H3, H4, H2A and H2B), which are enwrapped by 147 base pairs of DNA (Kouzarides 2007, Tyagi and Bukrinsky 2012, Tessarz and Kouzarides 2014). A number of residues in the tails of histones are post-translationally modified in numerous ways and these covalent modifications inscribe a complex “code” that eventually dictates the accessibility of gene’s promoter to the transcriptional machinery. These core histones undergo various kinds of post-translational modifications such as acetylation, methylation (mono-, di-, or tri-methylation), sumoylation, phosphorylation, ubiquitinylation, ADP-ribosylation etc. (Kouzarides 2007). These modifications are termed epigenetic modifications, as they modulate gene expression without changing the DNA sequence, and can be transmitted to offspring. The relatively stable nature of epigenetic modifications makes them ideal mediators for drug-induced brain maladaptation, which frequently leads to drug addiction. The collective nature of all the epigenetic modifications defines the specific nature of the chromatin structures. Chromatin structures, especially in the vicinity of the promoter region of a gene, regulate its expression. Thus, the type and amount of epigenetic modifications play a decisive role in defining specific chromatin structure and subsequent regulation of gene expression (Wolffe 1994, Narlikar, Fan et al. 2002, Felsenfeld and Groudine 2003). The open or relaxed chromatin structure which promotes access to the transcription machinery at the promoter region of a gene is called transcriptionally active or euchromatin structure. On the other hand, the closed or compact chromatin structure which inhibits the access of transcription machinery to the promoter region of a gene, is called transcriptionally repressive or heterochromatin structure (Narlikar, Fan et al. 2002, Kouzarides 2007).

The level of acetylation of nucleosomal core histones is regulated by the actions of histone acetyltransferases (HATs) and histone deacetylases (HDACs). HATs catalyze the reaction that promotes acetylation of core histones at their lysine residues, while HDACs catalyze the reaction that removes the acetyl group from histones. The negative charge of acetyl groups nullifies the positive charge of the lysine residues of histones. This event reduces the interaction between the positively charged histones with the negatively charged DNA, opens up or relaxes the chromatin structures and supports the entry of transcription machinery. Thus, histone acetylation induces euchromatin structures and facilitates transcription. In contrast, histone deacetylation promotes the establishment of heterochromatin structure and represses gene expression. Hence, the recruitment and levels of HATs and HDACs at and around gene promoter play critical role in controlling the expression of a gene (Kouzarides 2007, Bannister and Kouzarides 2011, Tyagi and Bukrinsky 2012, Tessarz and Kouzarides 2014).

Methylation of core histones is another major epigenetic modification that determines the overall nature of chromatin structure of a gene. However, histone methylation can lead to either transcriptional activation or repression depending on the methylation of particular histone residues and the extent of methylation. For example, accumulation of epigenetic modifications H3K4me1, H3K4me3 and H3k36me3 facilitates the establishment of transcriptionally active euchromatin structures and usually occurs at actively transcribing genes, whereas enrichment of H3K9me2, H3K9me3 or H3K27me3 epigenetic marks indicates the presence of heterochromatic structures and negatively correlates with transcription (Su and Tarakhovsky 2006, Kouzarides 2007, Bannister and Kouzarides 2011, Tessarz and Kouzarides 2014). Histone methylation can take place at both lysine and arginine residues. The equilibrium between histone methyltransferases and histone demethylases eventually dictate the histone methylation level. Similarly other histone tail modifications which regulate transcription, such as phosphorylation, ubiquitination, sumoylation, ADP-ribosylation, etc. are also regulated by complementary enzyme sets that catalyze reversed reactions.

Drugs of abuse induce epigenetic modifications

The regulation of expression of any eukaryotic gene involves chromatin reorganization, usually near the promoter region. Accordingly, gene expression modulated by drugs of abuse includes induction of different epigenetic modifications which subsequently define the type of chromatin structure at and near the promoter of that particular gene. Histone phosphorylation and acetylation (phospho-acetylation) are the most studied epigenetic modifications in the context of drugs of abuse, and are usually induced simultaneously by a number of drugs of abuse (Robison and Nestler 2011, Cadet, Jayanthi et al. 2013, Nestler 2014, Godino, Jayanthi et al. 2015). Extended literature has demonstrated the correlation between drugs induced histone phospho-acetylation with the effects on cellular gene expression, behavioral changes and addiction (Martin, Jayanthi et al. 2012, Feng and Nestler 2013, Nestler 2014). Numerous studies have clearly demonstrated the important role of drug of abuse induced various epigenetic modifications in stably modifying the expression of several genes in the nervous system (Kumar, Choi et al. 2005, Levine, Guan et al. 2005, Renthal, Maze et al. 2007, Nestler 2012, Feng and Nestler 2013, Nestler 2014, Godino, Jayanthi et al. 2015).

Interestingly, the mode of drugs of abuse exposure (acute or chronic) also defines the activation and expression duration of genes by inducing a selective set of epigenetic modifications (Kumar, Choi et al. 2005, Renthal and Nestler 2008, Renthal, Kumar et al. 2009, Renthal and Nestler 2009, Maze and Nestler 2011, Feng and Nestler 2013, Nestler 2014). Genome wide ChIP analysis after exposure to different drugs of abuse showed both subtle and persistent induction of genes due to selective histone H3 and H4 acetylation (Tsankova, Kumar et al. 2004, Kumar, Choi et al. 2005, Renthal, Kumar et al. 2009, Cadet, Jayanthi et al. 2013, Godino, Jayanthi et al. 2015). Interestingly, following cocaine, amphetamine (AMPH) and methamphetamine (METH) exposure, H3 acetylation was found to be associated with chronically induced genes, while H4 acetylation was associated with acutely induced genes. However, in a subset of other cellular genes, drug exposure promotes both H3 and H4 acetylation (Tsankova, Kumar et al. 2004, Brami-Cherrier, Valjent et al. 2005, Kumar, Choi et al. 2005, Renthal and Nestler 2008, Renthal and Nestler 2009, McQuown and Wood 2010, Feng and Nestler 2013, Nestler 2014, Godino, Jayanthi et al. 2015), suggesting further need for a more detailed investigation. Specifically, data regarding METH and AMPH are very rudimentary and non-conclusive. For example, a study found that a single METH injection enhances a global time dependent increase in acetylated H4K5 and H4K8, but found a corresponding global decrease in H3K9, H3K18, and H4K16 acetylation in the NAc (Martin, Jayanthi et al. 2012, Cadet, Jayanthi et al. 2013). This study also noted that METH exposure decreases the levels of HDAC1, but increases HDAC2 levels; for more detail please refer to a recent review (Godino, Jayanthi et al. 2015).

Another important finding from different analyses was that usually most drugs of abuse promote gene expression by inducing histone phospho-acetylation, whereas suppress genes by inducing histone H3 methylation at lysine 9 (Black, Maclaren et al. 2006, Shapshak, Duncan et al. 2006, Kouzarides 2007, Chiappelli, Shapshak et al. 2008, Shapshak, Chiappelli et al. 2008, McQuown and Wood 2010, Nestler 2014). This clearly demonstrated the inhibitory effect of cocaine on histone deacetylases (HDACs); however this is not true for all the HDAC classes, as detailed in the following section (Renthal, Kumar et al. 2009, Maze and Nestler 2011). Another factor that appears to play major role in gene regulation is the presence of dominant epigenetic modifications, as found in different analyses of reward regions of brain (Renthal and Nestler 2009, McQuown and Wood 2010, Feng and Nestler 2013, Nestler 2014). Epigenetic modifications induced by drugs of abuse also modulate the expression of certain genes that are linked to the phenotypic manifestations such as BDNF (brain-derived neurotrophic factor) and Cdk5 (cyclin-dependent kinase 5), CaMKIIα (Ca²⁺/calmodulin-dependent protein kinase IIα), c-Fos, δFosB etc. Thus, fluctuations in the expression of these genes are more visible in the form of altered behavioral consequences following drug intake. Based on the area of reward regions and occurred predominant histone modifications over there, different genes are more activated at different areas of reward regions of brain (Feng and Nestler 2013, Nestler 2014). Recent studies have confirmed that cocaine specifically inhibits class II HDACs, mainly HDAC-4 and -5, but activates Class III HDACs, SIRT1 and SIRT2. However, morphine exposure selectively induces SIRT1 expression only (Feng and Nestler 2013, Ferguson, Koo et al. 2013). These HDACs also deacetylate numerous non-histone proteins as well, hence the apparent outcome of drugs of abuse is the combined effect of a complex direct and indirect epigenetic and non-epigenetic cues (Renthal, Maze et al. 2007, Maze and Nestler 2011). There are lots of factors that define the ultimate nature of epigenetic modifications induced by drugs of abuse and their phenotypic manifestations, including different dosing regimens, diverse behavioral testing paradigms, different reward region areas or multiple biochemical targets in the brain. Hence, comprehensive analyses are needed to fully characterize regulation of histone modifications by different drugs of abuse and understand their precise contributions to addiction and behavioral consequences.

The investigations pertaining epigenetic modifications induced by drugs of abuse and their involvement in supporting their deleterious effects are still in their infancy. Moreover, almost nothing is known about the role of drugs of abuse induced epigenetic modifications on HIV gene expression. Experiments in rodents confirmed that specific epigenetic marks induced by drugs of abuse are transmitted to offspring, which exhibited behavioral abnormalities, such as drug tolerability and reduced response to conditioned fear (Itzhak, Ergui et al. 2015). These results suggest that the deleterious effects of drugs of abuse and/or HIV can be transmitted to the next generation and interfere with the epigenetic reprogramming that occurs during embryonic development (Itzhak, Ergui et al. 2015).

Role of epigenetics in the regulation of HIV transcription

Similar to other retroviruses, HIV integrates into the host cell genome, preferentially within the intronic regions of actively transcribing genes. This feature is due to the selective binding preference of lens epithelium-derived growth factor (LEDGF), a protein that plays an important role during HIV integration. LEDGF shows specificity towards transcriptionally active open chromatin structures (Schroder, Shinn et al. 2002, Han, Lassen et al. 2004, Lewinski, Yamashita et al. 2006, Brady, Agosto et al. 2009, Meehan, Saenz et al. 2009, Vatakis, Kim et al. 2009). Analogous to cellular genes, the expression of integrated HIV genome is facilitated by the establishment of transcriptionally active chromatin structures around the LTR promoter (Verdin, Paras et al. 1993).

The HIV LTR promoter is precisely flanked by two well-placed nucleosomes, independent of the site of integration in the cellular genome The nucleosome-0 (Nuc-0) is located upstream of the LTR promoter and nucleosome-1 (Nuc-1) is assembled downstream from the LTR promoter (Verdin 1991, Verdin, Paras et al. 1993). The epigenetic modifications of these two nucleosomes (Nuc-0 and Nuc-1) play major role in defining the overall chromatin structure at LTR and consequently controlling HIV gene expression (Verdin, Paras et al. 1993, Jordan, Defechereux et al. 2001, Jordan, Bisgrove et al. 2003).

The chromatin structures at HIV LTR are remodeled mainly by two kinds of protein complexes. One of these protein complexes modifies chromatin structures by inducing various post-translational epigenetic modifications at the N-terminal tails of histones. The second group of chromatin-modifying complexes uses energy to change the structures of nucleosomes in order to promote opening-up or relaxing nucleosomal structure. The SWI/SNF is one of such complexes that changes the location and reorganization of nucleosomes via an ATP dependent mechanism. The SWI/SNF remodeling complexes facilitates the opening of the nucleosomal structures and promotes access of the LTR promoter to transcription factors (for detail, please refer to (Henderson, Holloway et al. 2004, Agbottah, Deng et al. 2006, Mahmoudi, Parra et al. 2006, Treand, du Chene et al. 2006, Hakre, Chavez et al. 2011, Hargreaves and Crabtree 2011, Liu, Balliano et al. 2011, Rafati, Parra et al. 2011, Van Duyne, Guendel et al. 2011). These events eventually accelerate the overall rate of HIV transcription.

It is worth noting that most of the enzymes that catalyze epigenetic modifications do no bind directly to the DNA. As a result, the epigenetic enzymes must be recruited to HIV LTR by a variety of DNA binding proteins. HIV transcriptional repressors, such as CBF-1, YY1/LSF1, P50 homodimer, AP4, CTIP2, and thyroid hormone receptor, recruit chromatin modifying enzymes along with several other proteins as multiprotein complexes to HIV LTR to induce transcriptionally repressive heterochromatin structures (Coull, Romerio et al. 2000, Hsia and Shi 2002, Imai and Okamoto 2006, Williams, Chen et al. 2006, Marban, Suzanne et al. 2007, Tyagi and Karn 2007). We have demonstrated that CBF-1-induced repressive chromatin structures play an important role in restricting HIV transcription and promote HIV latency in primary CD4+ T cells (Tyagi, Pearson et al. 2010). The role of repressive epigenetic modifications in restricting HIV transcription during latency is quite evident due to the fact that their removal or inhibition leads to the reactivation of latent proviruses (Choudhary and Margolis 2011, Hakre, Chavez et al. 2011, Margolis 2011, Mbonye and Karn 2011, Tyagi and Bukrinsky 2012, Mbonye and Karn 2014).

In addition to the post-translational modification of nucleosomal histones, other epigenetic modifications, such as DNA methylation at the CpG islands flanking the transcription start site have also been implicated in regulating the HIV gene expression (Blazkova, Trejbalova et al. 2009, Kauder, Bosque et al. 2009, Chavez, Kauder et al. 2011). Hyper-methylation of DNA usually represses gene transcription, as hyper methylation of DNA can sterically hinder the transcriptional machinery, promote the recruitment of corepressor complexes and/or modify nucleosome structure. A number of studies have confirmed the impact of different drugs of abuse in modulating DNA methylation and components of involved protein machineries, but unfortunately no concrete conclusions have been derived from them (Robison and Nestler 2011, Nestler 2012, Feng and Nestler 2013, Nestler 2014). Moreover, the impact of drugs of abuse induced DNA methylation on HIV transcription has not yet been specifically assessed.

Moreover, the roles of certain noncoding RNAs, primarily microRNAs (miRNAs), have been implicated in HIV transcription and latency (Budhiraja and Rice 2013, Suzuki, Ahlenstiel et al. 2015). The miRNAs are single-stranded small RNA molecules of roughly 25 bp, which normally after binding to specific complementary sequences on the target mRNA inhibit its translation; however, miRNA binding occasionally also results in the degradation of specific mRNA (Bartel 2004). Viral and cellular miRNAs have a demonstrable role in regulating HIV gene expression and latency. In particular, cellular miR-28, miR-125b, miR-150, miR-223 and miR-382, which are enriched in metabolically silent, resting CD4+ T lymphocytes, suppress HIV translation by targeting its mRNA (Pomerantz, Feinberg et al. 1991, Lassen, Ramyar et al. 2006, Huang, Wang et al. 2007). A number of recent papers provide further details of miRNA-mediated regulation of HIV latency (Bennasser, Yeung et al. 2007, Huang, Wang et al. 2007, Klase, Kale et al. 2007, Triboulet and Benkirane 2007, Yeung, Benkirane et al. 2007, Corbeau 2008, Kumar and Jeang 2008, Sung and Rice 2009, Narayanan, Kehn-Hall et al. 2011, Sun and Rossi 2011). Therefore, innovative methodologies designed to manipulate the action of involved miRNAs could be proved useful in designing treatments targeting HIV transcription, replication and latency.

Cocaine induced histone modifications at HIV LTR

There is relatively little knowledge regarding the role of drugs of abuse induced epigenetic modifications in regulating HIV transcription; only recently we investigated the effect of cocaine-induced epigenetic modifications on HIV transcription (Sahu, Farley et al. 2015). Our results demonstrated that cocaine treatment greatly enhances the recruitment of histone acetyltransferase (HAT), p300, at HIV LTR and leads to the dissociation of histone deacetylase, HDAC3, from LTR. These events result in hyper-acetylation of both core histones, H3 and H4 (Sahu, Farley et al. 2015). Besides histone acetylation, cocaine treatment induces phosphorylation of histone H3 at serine 10 (p-H3S10). Acetylation and phosphorylation of core histones partially neutralizes the positive ionic charge of histones, and thus both of these epigenetic modifications contribute to the establishment of transcription-promoting euchromatin structures at HIV LTR. Interestingly, upon cocaine treatment, we also observed the loss of heterochromatic epigenetic modifications such as trimethylation of histone H3 at lysine 9 (H3K9me³) and lysine 27(H3K27me³) from HIV LTR. As indicated in Figure 1, our results demonstrated that cocaine exposure converts the transcriptionally repressive heterochromatin structures into transcriptionally active euchromatic structures at HIV LTR (Sahu, Farley et al. 2015). Accordingly, we found higher recruitment of RNA polymerase II at HIV LTR, further validating the fact that euchromatin structures facilitate the access to transcription machinery at gene promoters.

Drugs of abuse induce miRNA that regulates HIV transcription

A number of recent studies have reported the modulation of miRNAs in addiction-related behaviors in animal models, and several specific miRNAs, whose expression is altered by drugs of abuse in brain reward regions, have been shown to regulate the expression of several proteins strongly linked to addiction. Cocaine has been reported to up- and downregulate multiple miRNAs, such as upregulation of miR-181a and downregulation of miR-124 and let-7d (Chandrasekar and Dreyer 2009, Sartor, St Laurent et al. 2012). Accordingly, cocaine has been shown to enhance HIV-1 replication in CD4+ T cells by down-regulating miR-125b (Mantri, Pandhare Dash et al. 2012). Overall, very limited information is available regarding the drugs of abuse modulated miRNA, other non-coding RNA, their target mRNA and definitive role of miRNAs in regulating drugs induced behavioral and/or addiction consequences, besides their impact on HIV gene expression and replication.

Therapeutic implications to counter the impacts of drugs of abuse

Development of HAART regimens with improved penetration of blood brain barrier (BBB)

One of the difficulties associated with current HAART regimens in has been to achieve adequate concentrations of drugs in several anatomical sanctuary sites, such as the central nervous system. In these sanctuary sites, there is a perpetual low level of viral replication (Palmer, Josefsson et al. 2011, Cory, Schacker et al. 2013). Therefore, use of drugs of abuse further burdened HIV patients by accelerating ongoing HIV replication, primarily in the CNS (Chiasson, Stoneburner et al. 1990, Anthony, Vlahov et al. 1991, Ioannidis, Havlir et al. 2000, Nath, Maragos et al. 2001, Parikh, Dampier et al. 2014). Hence, the development of HAART regimens that better penetrate the CNS, will improve cognitive performance which has been demonstrated during HAART intensification with better BBB crossing antiretroviral agents (Smurzynski, Wu et al. 2011, Rappaport and Volsky 2015).

Inclusion of drugs in HAART regimens that restrict HIV transcription

While HAART efficiently restricts the HIV replication and infection of new cells, it is unable to prevent the transcription of viral proteins. The viral proteins that are produced may cause several adverse effects, such as chronic inflammation, by entering cells and interacting with cellular proteins. The CNS is particularly sensitive to inflammation induced by viral proteins, such as Gp120, Tat and Nef. The inflammatory cytokines and cytotoxic products that are secreted in response to HIV-induced inflammation by brain cells, especially microglia and perivascular macrophages and to a lesser extent astrocytes subsequently induce neuropathic changes. (Lipton and Gendelman 1995, Nath 2002, Kaul and Lipton 2006). Therefore, it implies that illicit drug mediated enhanced HIV transcription elevates the levels of viral proteins even in the presence of HAART regimens and contribute to chronic brain inflammation and deterioration of CNS function.

CONCLUSION AND PERSPECTIVES

The studies of the effects of drugs of abuse on HIV transcription are still in their infancy. Our work was the first to describe the salient underlying molecular mechanisms that cocaine utilizes in order to enhance HIV transcription. With the exception of the role of METH and cocaine on HIV transcriptional initiation, there is very little known about how other drugs of abuse affect HIV transcription. Thus, there is considerable work needed to define the underlying molecular mechanisms that different drugs of abuse utilize to enhance HIV transcription and replication.

The transcription factors discussed above are few of the many that affect HIV gene expression. The main goal of future research is to obtain a comprehensive view on the transcription factors induced by different drugs of abuse and their effect on the initiation and elongation phases of expression of relevant genes, especially those that affect HIV gene expression and brain function.

We have found that cocaine induces chromatin remodeling at HIV LTR via MSK1 and histone H3 phosphorylation. Similar epigenetic modifications at various gene promoters, especially in brain cells have been implicated in some of the long-lasting behavioral consequences of cocaine. In order to further understand the cocaine-induced epigenetic modifications at HIV LTR, it will be necessary to involve high throughput approaches, such as ChIP-Seq, to identify numerous other post-translational modifications in both histone and non-histone proteins.

There remains a huge body of work necessary to precisely characterize and define the vital role of epigenetic modifications induced by drugs of abuse as well as to determine the correlation between epigenetic changes with transcriptional regulation. It has been observed that single modification is not sufficient to regulate the expression of a gene; instead it requires the involvement of numerous epigenetic modifications in concert in order to establish euchromatin or heterochromatin structure. Thus, deciphering such a code, or chromatin signatures, will be a very difficult, yet a highly important goal for future research.

We also need to increase our understanding of the signaling pathways through which synaptic transmission and neural activity are connected to the activation of transcription factors and different epigenetic modifications.

In summary, there is a pressing need to examine the molecular mechanisms that are involved during the interactions between HIV and drugs of abuse at the level of gene expression in order to determine how they may be affecting HIV replication. A better understanding of these mechanisms may reveal new drug targets and open up new avenues for better pharmaceutical interventions in HIV-infected drug abusers.