(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Recruitment of the CoREST transcription repressor complexes by Nerve Growth factor IB-like receptor (Nurr1/NR4A2) mediates silencing of HIV in microglial cells [1]

['Fengchun Ye', 'Department Of Molecular Biology', 'Microbiology', 'Case Western Reserve University', 'Cleveland', 'Ohio', 'United States Of America', 'David Alvarez-Carbonell', 'Kien Nguyen', 'Konstantin Leskov']

Date: 2022-09 

Human immune deficiency virus (HIV) infection in the brain leads to chronic neuroinflammation due to the production of pro-inflammatory cytokines, which in turn promotes HIV transcription in infected microglial cells. However, powerful counteracting silencing mechanisms in microglial cells result in the rapid shutdown of HIV expression after viral reactivation to limit neuronal damage. Here we investigated whether the Nerve Growth Factor IB-like nuclear receptor Nurr1 (NR4A2), which is a repressor of inflammation in the brain, acts directly to restrict HIV expression. HIV silencing following activation by TNF-α, or a variety of toll-like receptor (TLR) agonists, in both immortalized human microglial cells (hμglia) and induced pluripotent stem cells (iPSC)-derived human microglial cells (iMG) was enhanced by Nurr1 agonists. Similarly, overexpression of Nurr1 led to viral suppression, while conversely, knock down (KD) of endogenous Nurr1 blocked HIV silencing. The effect of Nurr1 on HIV silencing is direct: Nurr1 binds directly to the specific consensus binding sites in the U3 region of the HIV LTR and mutation of the Nurr1 DNA binding domain blocked its ability to suppress HIV-1 transcription. Chromatin immunoprecipitation (ChIP) assays also showed that after Nurr1 binding to the LTR, the CoREST/HDAC1/G9a/EZH2 transcription repressor complex is recruited to the HIV provirus. Finally, transcriptomic studies demonstrated that in addition to repressing HIV transcription, Nurr1 also downregulated numerous cellular genes involved in inflammation, cell cycle, and metabolism, further promoting HIV latency and microglial homoeostasis. Nurr1 therefore plays a pivotal role in modulating the cycles of proviral reactivation by potentiating the subsequent proviral transcriptional shutdown. These data highlight the therapeutic potential of Nurr1 agonists for inducing HIV silencing and microglial homeostasis and ultimately for the amelioration of the neuroinflammation associated with HIV-associated neurocognitive disorders (HAND).

HIV enters the brain almost immediately after infection where it infects perivascular macrophages, microglia and, to a much lesser extent, astrocytes. In previous work using an immortalized human microglial cell model, we observed that integrated HIV constantly underwent cycles of reactivation and subsequent re-silencing. We now show that HIV shutdown after proviral reactivation is mediated by the Nurr1 nuclear receptor. Both the functional activation of Nurr1 by specific agonists, and the over expression of Nurr1, resulted in rapid silencing of activated HIV in microglial cells. Nurr1 not only repressed HIV expression but also selectively down regulated genes involved in microglial homeostasis and inflammation. Thus, Nurr1 is pivotal for HIV silencing and repression of inflammation in the brain and is a promising therapeutic target for the treatment of HAND.

Funding: This study was supported by NIH grants R01 DA043159 and R01 DA049481 to J. K. and R21-AI127252 to S. V. and two Development Awards from CFAR P30-AI36219. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability: The RNA-seq data has been deposited in the NCBI Sequence Read Archive (SRA) and is available under BioProject accession PRJNA789419. All other relevant data are within the manuscript and its Supporting Information files.

Copyright: © 2022 Ye et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Here we report that Nurr1 plays a pivotal role in silencing active HIV in microglial cells by recruiting the CoREST/HDAC1/G9a/EZH2 transcription repressor complex to HIV promoter. Transcriptomic data from cells expressing various levels of Nurr1 additionally demonstrated that it promotes microglial homoeostasis and suppression of inflammation in the brain.

It is therefore important to determine the factors responsible for inducing HIV reactivation and inflammation and explore cellular mechanisms that antagonize these factors in order to develop a molecular understanding of the etiology of HAND. Three members of the Nerve Growth Factor IB-like nuclear receptor family, which includes nuclear receptor 77 (Nur77, NR4A1), nuclear receptor related 1 (Nurr1, NR4A2), and neuron-derived receptor 1 (Nor1, NR4A3), are strong candidates for factors that mediate HIV silencing in microglial cells. These receptors play complementary roles in neurons and microglia to limit inflammatory responses. In neurons, these receptors act as positive transcriptional regulators that control expression of dopamine transporter and tyrosine hydroxylase for differentiation of dopamine neuron, as well as other key genes involved in neuronal survival and brain development [ 38 – 41 ]. By contrast, these nuclear receptors can also act as negative transcriptional regulators in microglia cells and suppress expression of inflammatory cytokines such as TNF-α and IL-1β [ 42 ]. Because of these combined mechanisms, Nerve Growth Factor IB-like nuclear receptors play a critical role in protection of the brain during neurodegenerative diseases such as Parkinson’s disease and Alzheimer’s disease [ 43 – 48 ].

The pro-inflammatory environment in the brain of HIV-infected patients is therefore likely to be highly permissive for HIV reactivation from latency [ 21 ]. However, microglia have developed counterbalancing mechanisms to prevent exaggerated activation. In the normal CNS environment, healthy neurons provide signals to microglia via secreted and membrane bound factors such as CX3CL1 and neurotransmitters that restore microglial homeostasis [ 28 ] and induce HIV-silencing [ 19 – 21 ]. For example, using a co-culture of (iPSC)-derived human microglial cells (iMG) that were infected with HIV and neurons, we demonstrated that HIV expression in iMG was repressed when co-cultured with healthy neurons but induced when co-cultured with damaged neurons [ 21 ]. Secretion of HIV proteins such as transactivator of transcription (Tat), negative regulatory factor (Nef), envelope glycoprotein gp120, and viral RNA reinforce the activation cycle because they are directly neurotoxic and also contribute to inflammation in the brain by activating uninfected microglial cells [ 29 – 37 ].

A unique feature of HIV infection of microglial cells is that the virus is able to efficiently establish latency [ 18 – 21 ]. In microglial cells, transcription initiation is primarily regulated by NF-κB. In resting microglia, NF-κB is sequestered in the cytoplasm [ 18 – 20 ]. However, unlike memory T-cells, pTEFb is not disrupted [ 22 , 23 ]. The provirus is also silenced epigenetically through the CoREST and polycomb repressive complex 2 (PRC2) histone methyltransferase machinery [ 13 , 24 – 27 ]. Activation of microglia by pro-inflammatory signals, such as tumor necrosis factor-alpha (TNF-α) and interleukin-1 beta (IL-1β), reversed these molecular restrictions, inducing the viral production which could potentially lead to neuropathology.

It has been challenging to associate the direct effects of HIV infection of microglia with the development HAND. Initial studies indicated, paradoxically, that HAND did not correlate with the number of HIV-infected cells or viral antigens in the central nervous system (CNS) [ 15 , 16 ], but instead correlates strongly with systemic inflammation and CNS inflammation [ 17 ]. However, the early studies neglected the potential side effects of anti-HIV drugs on neuronal damage, which could mask the benefits of reduced HIV expression by ART and the impact of HIV latency which can obscure the true levels of HIV in the CNS.

HIV invades the brain soon after primary infection [ 9 ]. The virus primarily infects perivascular macrophages and microglial cells, as well as a few astrocytes, but it does not infect neurons [ 10 , 11 ]. Because microglial cells are much longer-lived than astrocytes and perivascular macrophages, and can support productive HIV replication, they serve as long-lived cellular reservoirs of HIV-1, even in well-suppressed patients receiving ART [ 12 – 14 ].

A large number of HIV infected patients develop HIV-associated neurocognitive disorders (HAND) [ 1 ]. Symptoms seen in well-suppressed people with HIV (PWH), range from the mild neurocognitive disorder (MND) to asymptomatic neurocognitive impairment (ANI) [ 2 – 4 ]. Although combination antiretroviral therapy (ART) dramatically lowers the levels of HIV RNA in the brain [ 5 – 7 ], it does not reduce the incidence of HAND [ 4 , 8 ]. Thus, with the availability of widespread ART, HIV-associated dementia is virtually eliminated, but the frequencies of MND and ANI have actually increased [ 4 , 8 ].

Results

Nurr1 overexpression enhances HIV silencing To further examine how the nuclear receptors contribute to HIV silencing, we constructed lentiviral vectors expressing either N-terminal 3X-FLAG-tagged Nur77, Nurr1, or Nor1 under the control of a CMV promoter. Infection of HC69 cells with the different lentiviruses generated cell lines that stably expressed either FLAG-tagged Nur77, Nurr1, Nor1, or the empty vector (Fig 3). Overexpression of Nurr1 was rigorously confirmed by mapping of normalized RNA-Seq reads to the Nurr1 gene locus (Fig 3A). The overexpression of each of the NR4A1 (Nur77), NR4A2 (Nurr1) and NR4A3 (Nor1) nuclear receptors was further verified by western blots (Fig 3B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Overexpression of Nurr1 in HC69 cells enhances HIV silencing. A, RNA-Seq confirmation of overexpression (OE) of Nurr1 in HC69 cells. Sequence read histograms for the Nurr1 locus is shown for control (vector) and Nurr1 overexpression. Annotated genes for the shown locus are indicated on the top, and the position of the locus on chromosome 2 is shown both at the top and the bottom. A read scale for each row is shown on the right, with the values for the overexpression studies drawn on a log2 scale. B, Verification of Nur77, Nurr1, and Nor1 overexpression by Western blot analysis in HC69 cell lines stably expressing 3X-FLAG-tagged Nur77, Nurr1, and Nor1 respectively by using a mouse monoclonal anti-FLAG M2 antibody (Sigma, Cat# F1804) and the nuclear receptor specific antibodies described in Fig 2B. HC69 cells stably carrying the 3X-FLAG-empty vector were used as a reference for comparison. The level of β-tubulin was used as a loading control. Notably, the levels of endogenous Nur77 and Nor1 in HC69 cells were very limited. In contrast, Nurr1 was constitutively expressed in HC69 cells. C, Schematic depicting the TNF-α stimulation and chase studies. The four cell lines described in B were either untreated or treated with high dose (400 pg/ml) TNF-α for 24 hr. To examine HIV silencing, one set of TNF-α induced cells were used in a chase experiment by continuous culture of the cells in the absence of TNF-α for an additional 48 hr. The time points at which TNF-α is added or removed are shown by arrows on the top. D, Expression of HIV Nef protein in the different cell lines before and after TNF-α stimulation and at the end of the chase experiment was measured by Western blot analysis. The level of β-tubulin was used as a loading control. E, Expression level of HIV mRNA (black bar graph) and Nurr1 (red rectangles and lines) in transcripts per million cellular transcripts are shown for each of the treatment steps shown in panel C in both vector-infected cells (on the left) and Nurr1 overexpressing cells (on the right half of the graph). For the 24 hr TNF-α stimulation step, both a low dose (20 pg/ml) and a high dose (400 pg/ml) are used. The values shown are the average of three replicate RNA-Seq samples with two standard deviations as error bars. The expression values for HIV and Nurr1 are shown on Y axes to the left and right, respectively. https://doi.org/10.1371/journal.ppat.1010110.g003 To examine how overexpression of each of these nuclear receptors modulates HIV proviral silencing after reactivation, chase experiments were performed following TNF-α stimulation (Fig 3C). In this protocol, we stimulated all four cell lines with a high dose (400 pg/ml) of TNF-α for 24 hr to induce HIV transcription through activation of NF-κB [19], followed by a 48 hr chase during which TNF-α was removed by washing the cells with PBS followed by the addition of media lacking TNF-α (Fig 3C). As shown by the western blot in Fig 3D, TNF-α strongly induced the expression of HIV Nef protein, which we used as a marker of HIV reactivation, in all cell lines at 24 hr. Notably, Nef expression decreased in all four cell lines during the 48 hrs after TNF-α withdrawal. However, the reduction in Nef expression was much more pronounced in HC69 cells that expressed 3X-FLAG-Nurr1 than any of the other receptors, suggesting that overexpression of Nurr1 selectively enhances silencing of active HIV in HC69 cells. We additionally confirmed the western blot data using RNA-Seq (Fig 3E), which permitted us to accurately and simultaneously measure the fluctuations in both HIV and Nurr1 mRNA expression. In the Nurr1 overexpressing cells, in unstimulated conditions, the basal level of HIV proviral expression was 1.8-fold higher than cells expressing empty vector. Following stimulation with either a low dose (20 pg/ml), or high dose (400 pg/ml), of TNF-α, both vector-infected and Nurr1 overexpressing cells showed an increase in proviral expression. While the level of HIV expression was similar between control cells (vector-infected) and Nurr1-overexpressing cells after high dose TNF-α stimulation, Nurr1 overexpressing cells had much lower proviral expression level after low dose TNF-α stimulation (Fig 3E). The level of HIV mRNA after withdrawal of high dose TNF-α was three times lower in Nurr1 overexpressing cells than in vector-infected cells (Fig 3E), strongly suggesting that overexpression of Nurr1 enhanced silencing of active HIV in HC69 cells. As a control for clonal variation, we also generated cell lines from clone#29 stably infected with either the empty 3X-FLAG-lentivirus (vector) or the lentivirus expressing 3X-FLAG-Nurr1 (S2A Fig). Chase experiments performed using these cells confirmed that Nef expression is reduced by more than 50% in cells overexpressing Nurr1 compared to control cells (vector) at the end of the chase experiment (S2A Fig).

Nurr1 knockdown blocks HIV silencing As a complementary approach, we also performed shRNA-mediated knock down (KD) of endogenous Nurr1 in HC69 cells. Cell lines that stably expressed Nurr1-specific, or control (scrambled), shRNA were verified for effective Nurr1 KD by RNA-Seq analyses (Fig 4A). Following the protocol described in Fig 4B, the control and Nurr1 KD cells were activated with a high dose (400 pg/ml) of TNF-α for 24 hr, followed by a 72 hr chase. Western blot analyses also confirmed the Nurr1 knock down efficiency (Fig 4C). The blots also showed that HIV Nef protein, was strongly induced at 24 hr post TNF-α stimulation in both the control and the Nurr1 KD cells. As expected, after the chase, Nef levels decreased significantly in the control cells due to expression of the endogenous Nurr1 but remained high in the Nurr1 KD cells (Fig 4C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Nurr1 knock down (KD) in HC69 cells enhances HIV expression and block proviral silencing during the chase step. A, RNA-Seq confirmation of Nurr1 KD in HC69 cells. Read histograms for the Nurr1 locus is shown for non-targeting shRNA-infected cells, and cells infected with Nurr1 specific shRNA lentiviral constructs. The Nurr1 shRNAs resulted in a 2.6-fold reduction in Nurr1 mRNA level in the Nurr1 KD cells. Annotated genes for the shown locus are indicated on the top, and the position of the locus on chromosome 2 is shown both at the top and the bottom. A read scale for each row is shown on the right, with the values for the knock down studies drawn on a linear scale. B, Schematic depicting the TNF-α stimulation and chase studies. The two shRNA lentiviral transduced cell lines described in A were either untreated or treated with high dose (400 pg/ml) TNF-α for 24 hr. One set of TNF-α induced cells were used in a chase experiment in the absence of TNF-α for an additional 48 hr. The time points at which TNF-α is added or removed are shown by arrows on the top. C, Western blot studies measuring the expression of endogenous Nurr1, Nef, and β-tubulin in cells infected with either a non-targeting control shRNA or Nurr1-specific shRNA lentiviral vectors. The expression patterns from the TNF-α (400 pg/ml) stimulation and the chase step are shown. D, KD of endogenous Nurr1 strongly inhibits HIV silencing. The percentages of GFP+ cells in the two cell lines, before treatment, at 24 hr post-TNF-α (400 pg/ml) stimulation, and at 72 hr after TNF-α withdrawal (chase) were analyzed by flow cytometry and calculated from three independent experiments. The difference in GFP expression between the two cell lines at 72 hr chase was statistically significant, with a p = 0.0078. E, Expression level of Nurr1 (red rectangles and lines) and the HIV provirus (black bar graph) in transcripts per million cellular transcripts are shown for each of the treatment steps in both non-targeting shRNA infected cells (on the left) and Nurr1-specific shRNA-infected cells (on the right half of the graph). The values shown reflect the average of three replicate RNA-Seq samples from two distinct shRNA constructs per control and Nurr1 knock down groups, with two standard deviations as error bars. The expression values for HIV and Nurr1 are shown on Y axes to the left and right, respectively. https://doi.org/10.1371/journal.ppat.1010110.g004 Similar results were obtained using flow cytometry (Fig 4D). Compared to cells expressing control shRNA with 10.5% GFP+ cells, the Nurr1 KD cells displayed 58.8% GFP+ cells even before TNF-α stimulation, which most likely resulted from failure of silencing spontaneously reactivated HIV in these cells due to Nurr1 depletion (Fig 4D). As expected, after exposure to a high dose of TNF-α for 24 hr, both the control and Nurr1 KD cell lines expressed equally high levels of GFP expression, displaying 86.3% and 91.2% GFP+ cells respectively. However, 72 hrs after TNF-α withdrawal, GFP expression decreased significantly in cells expressing the control shRNA (47.2% GFP+) but remained high (74.6% GFP+) in the Nurr1 KD cells (Fig 4D). Finally, the overall mRNA level of the HIV measured by RNA-Seq was 1.7-fold higher in Nurr1 KD at the end of the chase experiment (Fig 4E). Thus, both the overexpression and the reciprocal KD experiments confirmed an essential role of Nurr1 in the silencing of HIV in microglial cells.

Nurr1 binding to HIV 5’LTR is essential for HIV silencing Like most transcription factors, Nurr1 binds to specific DNA motif for transcriptional regulation of its target genes [66]. The orphan and ligand-mediated nuclear receptors form dimers on their target DNAs via highly cooperative assembly of their DNA-binding domains. We identified a putative Nurr1 binding site overlapping the COUP/AP-1 sites in the U3 region of the HIV-1. The site contained an octanucleotide with a canonical nuclear receptor binding motif (NBREP: AAAGGTCA), and across the dyad axis, the IR5 motif which permits binding of either the Nurr1 homodimer or binding of a heterodimer with other nuclear receptors (Fig 5A) [66]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Nurr1 binds to conserved sites in the HIV LTR. A, Location of the putative Nurr1 binding site in the HIV LTR consists of a 8-nt NGFI-B-responsive element (NBRE) and an inverted 8-nt repeat (IR5), to which Nurr1 monomers, homodimers, and heterodimers bind [66]. B, Lentiviral construct expressing the DNA binding deficient 3X-FLAG-Nurr1 mutant (Nurr1 Mut) was generated from the original lentiviral construct expressing 3X-FLAG-Nurr1 wild type (Nurr1 WT) by converting both residues C280 and E281 to alanine (A) as reported previously [42]. C, ChIP assay measurement of the levels of Nurr1 protein at Nuc-0 in the HIV-1 LTR in HC69 cells expressing empty vector (grey bars), 3X-FLAG-Nurr1 WT (black bars), or 3X-FLAG-Nurr1 Mut (red bars). The mean levels of Nurr1 in 5’LTR, presented as the percentage (%) of ChIP product over input DNA of each sample, was calculated from triplicates qPCR of a single ChIP experiment using primers for the Nuc-0 region. D, Impact of overexpression of Nurr1 carrying DNA-binding defective mutations on GFP expression. HC69 cells expressing empty vector, 3X-FLAG-Nurr1 WT, or 3X-FLAG-Nurr1 Mut were induced with a high dose (400 pg/ml) TNF-α for 24 hr and then chased in the absence of TNF-α for 48 hr and the fraction of GFP+ cells was measured by flow cytometry. The p-values of pair-sample, Student’s t-tests for differences in the numbers of GFP+ cells between the control cells (Vector) and cells expressing wild-type Nurr1 (Nurr1 WT) or between the control cells (Vector) and cells expressing mutant Nurr1 (Nurr1 Mut) at the end of chase were calculated from three technical repeats. E, Impact of overexpression of Nurr1 carrying DNA-binding defective mutations on HIV Nef expression. Western blots were performed on HC69 cells expressing empty vector, 3X-FLAG-Nurr1 WT, or 3X-FLAG-Nurr1 Mut using antibodies directed against FLAG, Nef or β-Tubulin, which served as a loading control. https://doi.org/10.1371/journal.ppat.1010110.g005 Data from ChIP-seq experiments provided direct evidence for Nurr1 recruitment to the proviral LTR in untreated HC69 cells (S2B Fig). As expected, Chip assays also showed that there were increased levels of Nurr1 at the HIV 5’LTR in GFP- cells compared to GFP+ cells at 48 hr after TNF-α withdrawal in a chase experiment (S2B Fig). To confirm that Nurr1 binding to HIV is due to direct DNA recognition we generated a DNA binding defective mutant by converting C280 and E281 in the Nurr1 DNA binding domain to alanine (A) (Fig 5B) [42]. As shown in Fig 5C, ChIP assays showed that the level of Nurr1 at the LTR was increased approximately 2-fold in cells that over expressed the wild type Nurr1 compared to cells expressing either the empty vector or the mutant Nurr1. The behavior of HC69 cells stably expressing the empty vector, 3X-FLAG-Nurr1 WT, and 3X-FLAG-Nurr1 mutant was then compared in chase experiments following the protocol described in Fig 3C. As shown in Fig 5D and 5E, the reduction in GFP and Nef expression after the chase was much more pronounced in cells overexpressing the wild type Nurr1, compared to the control vector. In contrast, in cells overexpressing the mutant Nurr1, expression of both GFP and Nef remained high after TNF-α withdrawal, demonstrating that Nurr1 DNA binding is essential for HIV silencing. It is important to note that the overall cellular levels of Nurr1 protein do not correlate with the levels of Nurr1 at the HIV LTR. In fact, cellular Nurr1 protein levels increase slightly after TNF-α stimulation and were higher in the GFP+ cells than in the GFP- cells at the end of chase (S2C Fig). Instead, the decrease in Nurr1 at the LTR appears to be caused by gene-specific degradation due to sumoylation of Nurr1 upon TNF-α stimulation (S2D Fig), consistent with previous reports [42].

Nurr1 promotes the recruitment of the CoREST/HDAC1/G9a/EZH2 repressor complex to the HIV promoter Nurr1 binding to HIV provirus may inhibit HIV transcription either through direct repression or by engaging a trans-repression mechanism involving the recruitment of specific transcription repressors. A Nurr1-mediated trans-repression mechanism involving the recruitment of the CoREST transcription repressor complex regulates inflammatory cytokine synthesis, such as TNF-α and IL-1β in microglial cells, following LPS stimulation [42,67]. CoREST can interact in a dynamic manner with multiple epigenetic silencing machinery components including histone deacetylases 1/2 (HDAC1/2), euchromatic histone lysine N-methyltransferase 2 (G9a; EHMT2), lysine (K)-specific demethylase 1A (KDM1A), and enhancer of zeste 2 polycomb repressive complex 2 subunit (EZH2) [68,69] (Fig 6A). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. Nurr1 promotes recruitment of the CoREST repressor complex to HIV promoter. A, Schematic illustration of Nurr1-mediated epigenetic silencing of active HIV in microglial cells by recruiting the CoREST/HDAC1/G9a/EZH2 repression complex to HIV promoter. B, Nurr1 associates with CoREST, HDAC1, G91, and EZH2 to form a transcription repression complex in microglial cells (HC69). HC69-3X-FLAG-vector and HC69-3X-FLAG-Nurr1 cells were cultured in the absence (untreated) or presence of high dose (400 pg/ml) TNF-α for 4 hr and 24 hr respectively. A portion of these cells were also used in a chase experiment by culturing the cells for an additional 24 hr (chase) after stimulation with high dose TNF-α for 24 hr and subsequent washing with PBS (TNF-α 24h+24h). Total protein lysates from the differently treated cells were isolated and used for co-immunoprecipitation (Co-IP) with a mouse anti-FLAG monoclonal antibody. The original protein lysates (Input) and the Co-IP products were analyzed by Western blot analysis with antibodies to FLAG, CoREST, HDAC1, G9a, EZH2, and β-tubulin respectively. C, levels of CoREST (Top) and G9a (Bottom) at HIV Nuc1 (+30 to +134) in HC69-control shRNA (Control) and HC69-Nurr1 shRNA (Nurr1 KD) cell lines. Cell were activated with TNF-α and chased as described in B. The levels of CoREST and G9a in HIV 5’LTR were measured by qPCR and calculated as percentages of the amounts of ChIP products over input DNA from three technical replicates. https://doi.org/10.1371/journal.ppat.1010110.g006 To evaluate the role of CoREST in HIV silencing in microglial cells, we first conducted co-immunoprecipitation (Co-IP) assays to confirm the association of Nurr1 with components of the CoREST repressor complex in HC69 cells (Fig 6B). HC69-3X-FLAG-vector and HC69-3X-FLAG-Nurr1 cells were treated with and without a high dose (400 pg/ml) of TNF-α for either 4 hr or 24 hr. After 24 hr of TNF-α treatment, the cells were chased in the absence of TNF-α for a further 24 hr. Total protein lysates from the differently treated cells were immunoprecipitated using a mouse monoclonal anti-FLAG antibody conjugated to magnetic beads. The anti-FLAG beads pulled down not only FLAG-tagged Nurr1 but also CoREST, HDAC1, G9a, and EZH2 from the HC69-3X-FLAG-Nurr1 cell lysates, demonstrating that in the microglial cells Nurr1 bound directly to the CoREST repressor complex. Notably, the amount of CoREST associated with Nurr1 increased after the cells were stimulated with TNF-α. In contrast, the amounts of G9a and EZH2 proteins associated with Nurr1 decreased at 4 hr post-TNF-α stimulation but rebounded at 24 hr post-TNF-α stimulation. Together, these results suggested that the Nurr1/CoREST/HDAC1/G9a/EZH2 complex were most likely dissociated from each other during early time points of TNF-α stimulation but were reassembled at later time points. To provide direct evidence that Nurr1 mediates the recruitment of the CoREST/HDAC1/G9a/EZH2 complex to HIV promoter, we treated HC69 cells expressing control shRNA and Nurr1 shRNA with high dose TNF-α, followed by a 24 hr chase. We then conducted additional ChIP experiments and measured the ChIP products by quantitative PCR (qPCR). As shown in Fig 6C, CoREST was strongly recruited to HIV promoter at 4 hr post TNF-α stimulation in HC69 cells expressing control shRNA, however, its recruitment was substantially inhibited in Nurr1 KD cells. Similarly, G9a and HDAC1 levels in HIV promoter peaked at 24 hr post TNF-α stimulation in HC69 cells expressing control shRNA but their recruitment was also reduced in Nurr1 KD cells (Fig 6D). A higher resolution analysis of the recruitment of Nurr1 and the CoREST complex to the HIV provirus was obtained by ChIP-Seq experiments (Fig 7). Nurr1, CoREST, EZH2 and G9a were present near the promoter region of the HIV provirus in untreated cells, but there were only low levels of HDAC1. Each factor showed unique kinetics following TNF-α activation and the subsequent chase. For example, following TNF-α activation for 24 hrs there was a precipitous loss of Nurr1 and an increase of HDAC1, consistent with maximal HIV transcription. After the chase, when latency was restored, Nurr1 levels increased and HDAC1 levels decreased. The levels of CoREST at the HIV promoter peaked at 4 hr post-TNF-α stimulation and then declined to basal levels by 24 hr and after the chase. By contrast the levels of G9a and EZH2 at the HIV promoter peaked at 24 hr post-stimulation and remained high after the chase. Thus, there is a dynamic exchange between the epigenetic silencing factors and CoREST. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. ChIP-Seq analysis of the recruitment of Nurr1 and histone modifiers to the HIV provirus. Histograms show numbers of sequence reads on the Y axis along the length of the reporter HIV-1 pro-viral genome on the X axis. A, Aligned reporter genome. B, Nurr1 (Santa Cruz Biotech, Cat #sc-81345). C, CoREST (Cell Signaling, Cat #14567. D, EZH2 (Cell Signaling, Cat #5246S). E, G9a (Cell Signaling, Cat #3306S). F, HDAC1 (Santa Cruz Biotechnology, Cat #sc-7872). G, IgG control. For each antibody, chromatin was prepared from HC69 cells that were untreated, induced with TNF-α (400 pg/ml) for 4 hr and 24 hr respectively, or used in a chase experiment by continuously culturing HC69 cells in the absence of TNF-α for 24 hr after stimulating the cells with TNF-α (400 pg/ml) for 24 hr. Construction of ChIP-Seq DNA libraries with the ChIP products, enrichment for HIV-1 specific sequences, and data analysis following Ion Torrent sequencing were described in Materials & Methods. The sequence reads on the Y axis were set at the same scales for different time points of ChIP-seq using the same antibody but varied for ChIP-seq with different antibodies, which were from 0 to 2800 reads for ChIP-Seq/Nurr1, 0 to 200 reads for ChIP-Seq/EZH2, 0 to 400 reads for ChIP-Seq/CoREST, 0 to 300 reads for ChIP-seq/G9a, 0 to 1300 reads for ChIP-Seq/HDAC1, and 0 to 200 reads for ChIP-Seq/IgG, respectively. https://doi.org/10.1371/journal.ppat.1010110.g007

Epigenetic silencing of HIV in microglial cells To confirm the functional role of the histone modifying machinery recruited by Nurr1 and CoREST to the HIV promoter, we conducted additional ChIP assays to monitor the expected histone modifications. As shown in Fig 8, higher levels of repressive histone methylation marks H3K27me3 and H3K9me2 but lower levels of acetylated histone mark H3K27Ac and RNA polymerase II (RNAP II) were detected in HC69-3X-FLAG-Nurr1 cells than in HC69-3X-FLAG-vector cells at the end of 48 hr chase after TNF-α stimulation for 24 hr, which is fully consistent with the rapid HIV silencing in the Nurr1 overexpression cells shown earlier (Fig 3D). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 8. Nurr1 overexpression promotes repressive histone methylation and deacetylation. A, Location of Nurr1 binding site and positions of the Nuc0, promoter, Nuc1, and Nuc2 regions in the HIV-1 LTR. B, ChIP assays with antibodies for RNA polymerase II (RNAP II). C, H3K27-Ac. D, control IgG. E, H3K27me3. F, H3K9me2. HC69 cells infected with either a vector control (grey bars) or a 3X-FLAG-Nurr1 overexpression vector (red bars) were cultured for 48 hr after stimulation with high dose (400 pg/ml) TNF-α for 24 hr. Data is from three technical replicates. Dotted line indicates the IgG background signal. https://doi.org/10.1371/journal.ppat.1010110.g008 To further investigate how the CoREST/HDAC1/G9a/EZH2 complex contributes to HIV silencing, we treated HC69 cells with high dose (400 pg/ml) of TNF-α for 24 hr followed by a chase in the absence or presence of epigenetic inhibitors that target the CoREST complex, specifically: HDAC inhibitor suberoylanilide hydroxamic acid (SAHA), G9a inhibitor UNC0638, and EZH2 inhibitor GSK343 (Fig 9A). The activity of each of these inhibitors in inducing relevant histone modifications and HIV expression in T cells has been extensively examined [70], and they had minimal toxicity in HC69 cells (S1A Fig). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 9. The CoREST repressor complex plays a pivotal role in silencing active HIV in microglial cells. A, Inhibition of HDAC1, G9a, and EZH2 blocked silencing of activated HIV in HC69 cells. HC69 cells were stimulated with high dose (400 pg/ml) TNF-α for 24 hr. After washing with PBS, the cells were cultured in the presence of DMSO (placebo, Control), HDAC inhibitor SAHA (1 μM), G9a inhibitor UNC0638 (1.25 μM), or EZH2 inhibitor GSK343 (1.25 μM), for 48 hr. The levels of GFP expression for each treatment were measured by flow cytometry and calculated from three independent experiments, with p values between the control and treatment with each inhibitor indicated. B, Verification of CoREST KD by Western blot detection of CoREST protein expression in HC69 cell lines stably expressing control shRNA or CoREST-specific shRNA. C, Verification of EZH2 and G9a KO by Western blot detection of G9a and EZH2 protein expression in HC69 cells stably expressing CRISPR/Cas9 and G9a or EZH2 specific gRNA, which were compared to the control HC69 cells stably expressing CRISPR/Cas9 without gRNA. β-tubulin was used as a loading control for all Western blot analysis. D, CoREST KD prevents HIV silencing. The HC69-control shRNA and HC69-CoREST-shRNA cells were untreated, induced with high dose (400 pg/ml) TNF-α for 24 hr, or used in a chase experiment by continuous culturing the cells for 48 hr after TNF-α stimulation for 24 hr and washes with PBS. GFP expression levels of all cells were measured by flow cytometry and the mean values were calculated from three independent experiments. Significant differences were observed between the HC69-control shRNA and HC69-CoREST shRNA cell lines. E, G9a and EZH2 KO prevents HIV silencing. Evaluation of the HC69 cell lines expressing G9a or EZH2 specific gRNA or empty vector by flow cytometry following the same protocol as in panel D. There was a significant difference between HC69-vector and HC69 EZH2 or G9a KO cell lines at 48 hr after TNF-α withdrawal, with p < 0.01. https://doi.org/10.1371/journal.ppat.1010110.g009 As described previously the levels of GFP+ cells dropped from 88.4% to 67.03% during the chase when cells were cultured in the absence of the inhibitors (Fig 9A). However, in the presence of SAHA, UNC0638, or GSK343, the numbers of GFP+ cells remained higher (i.e., 77.8%, 85.5%, and 84.7% respectively), indicating that functional inhibition of these epigenetic silencers prevented active HIV from reverting to latency. All three inhibitors also induced HIV reactivation in latently infected HC69 cells (S1B Fig), thus underscoring the importance of these epigenetic mechanisms for HIV silencing in microglial cells. To confirm the role of these epigenetic silencers, we generated HC69 cell lines stably expressing CoREST-specific shRNA or CRISPR/Cas9/guide RNA (gRNA) for G9a or EZH2. We confirmed successful KD or knock out (KO) of these proteins in these cell lines by Western blot analysis (Fig 9B and 9C). The genetically modified cells were activated with a high dose (400 pg/ml) of TNF-α for 24 hr, followed by culturing the cells in the absence of TNF-α for 48 hr and measurement of GFP expression. CoREST KD substantially increased GFP expression (80.1% GFP+ vs. 25.8% in control cell) even without TNF-α stimulation (Fig 9D). Stimulation with high-dose TNF-α for 24 hr resulted in 94.1% and 84.9% GFP+ cells in CoREST KD and control cells respectively. However, after TNF-α withdrawal and subsequent culture for 48 hr, the numbers of GFP+ cells decreased significantly in cells expressing control shRNA (67.7%) but remained high in CoREST KD cells (91.3%), confirming that CoREST was crucial for the silencing of active HIV in microglial cells. Similar results were seen with the G9a and EZH2 KO cell lines (Fig 9E). Therefore, both the ChIP experiments and gene knockout results demonstrate a pivotal role for the CoREST/HDAC1/G9a/EZH2 transcription repressor complex in silencing active HIV in microglial cells. Taken together, these results clearly demonstrated a pivotal role for Nurr1 in mediating recruitment of the CoREST/HDAC1/G9a/EZH2 machinery to the promoter of active HIV for epigenetic silencing.

Regulation of Nurr1 trans-repression of HIV-1 transcription by phosphorylation The NF-κB inducible kinase (NIK) played a crucial role in inducing Nurr1 serine phosphorylation following LPS stimulation, which triggered the association of Nurr1 with the CoREST repressor complex and subsequent gene silencing [42]. To test whether this mechanism also applied to HIV (Fig 10A), we knocked down NIK in HC69 cells by using lentiviruses expressing NIK specific shRNAs (Fig 10B). The NIK knockdown cells were then induced with a high dose (400 pg/ml) of TNF-α for different time points and a co-IP experiment with a mouse anti-Nurr1 monoclonal antibody and the control IgG was performed (Fig 10C). Western blot analysis of the co-IP products using an anti-phospho-serine antibody and an anti-CoREST antibody showed that in cells expressing the control shRNA, the levels of phosphorylated Nurr1 and CoREST pulled down with the anti-Nurr1 antibody increased substantially at 4 hr post-TNF-α stimulation (Fig 10C). However, both Nurr1 phosphorylation and its association with CoREST after TNF-α stimulation were largely inhibited in the NIK KD cells. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 10. TNF-α stimulation induces Nurr1 phosphorylation by NIK, which triggers its association with CoREST repressor complex for HIV silencing. A, schematic presentation of TNF-α induced Nurr1 phosphorylation and its association with CoREST for HIV silencing. B Knock down (KD) of NIK in HC69 cells with NIK specific shRNAs (Santa Cruz Biotech, Cat# sc-36079V) and control shRNA respectively. Left: The relative levels of NIK mRNA from HC69 cells expressing control shRNA (grey bars) and NIK shRNA (green bars) was measured by qRT-PCR using the primers 5’ TGCGGAAAGTGGGAGATCCTGAAT 3’ (forward) and 5’ TGTACTGTTTGGACCCAGCGATGA 3’ (reverse). Expression levels were then normalized to the β-actin gene. Right: Western blot analysis of NIK protein levels using a rabbit monoclonal antibody (abcam Cat # ab203568). C. Co-immunoprecipitation (co-IP) of CoREST, pSer-Nurr1 and Nurr1. Total protein extracts from HC69 cells expressing control shRNA or NIK shRNAs that were untreated, induced with a high dose (400 pg/ml) of TNF-α for different time points, or stimulated for 24 hr followed by a 24 hr chase. co-IP experiments were performed using a mouse monoclonal anti-Nurr1 antibody (Santa Cruz Biotech) or control mouse IgG. Right: The co-IP products were analyzed by Western blot to detect the pulled down Nurr1 protein (using the same anti-Nurr1 antibody), serine-phosphorylated Nurr1 by using a mouse monoclonal anti-phospho-serine antibody from Millipore-Sigma (Cat# P5747-25), and CoREST by using a rabbit polyclonal anti-CoREST antibody (EMD Millipore, Cat# 07–579). Left: Ratios of pSer-Nurr1/Nurr1 and CoRES/Nurr1 in Co-IP samples measured by densitometry of the Western blots. D, GFP expression in HC69 cells expressing either control shRNA (grey bars) and NIK shRNAs (green bars). Cells were untreated, induced with a high dose (400 pg/ml) of TNF-α for 24 hr, or continuously cultured in the absence of TNF-α for 48 hr (chase), in presence or absence of 1 μM 6-MP. The p-values of pair-sample, Student’s t-tests for differences in the numbers of GFP+ cells were calculated from three technical replicates. E, Western blot detection of Nef protein in cells treated as described in D. β-Tubulin was used as loading control. https://doi.org/10.1371/journal.ppat.1010110.g010 To examine how NIK KD affects HIV silencing, we performed chase experiments. As shown in Fig 10D, GFP expression persisted at 48 hr after TNF-α withdrawal (chase) compared to the control cells. Notably, GFP expression during the chase phase was further decreased in the control cells when they were cultured in the presence of Nurr1 agonist 6-MP. However, the NIK KD cells evidently lost their responsiveness to 6-MP, suggesting that Nurr1 phosphorylation is required for 6-MP mediated HIV silencing as well. Consistent with the GFP expression results, NIK KD also prevented the “shut-down” of Nef expression during the chase phase (Fig 10E). These results not only highlight a critical role of NIK in inducing Nurr1 phosphorylation and its association with CoREST for HIV silencing (Fig 10A) but also strongly support the trans-repression mechanism for Nurr1-mediated HIV silencing.

Activation of Nurr1-mediated trans-repression by 6-MP As a final evidence that Nurr1 mediates HIV silencing by trans-repression, we demonstrated that the Nurr1 agonist, 6-MP, also engaged the CoREST repressor complex. HC69 cells expressing either control shRNA or Nurr1 specific shRNAs were treated with 6-MP for 48 hr and GFP expression was monitored to evaluate HIV transcription during a chase experiment (S3A Fig). As expected, Nurr1 knockdown increased basal expression approximately 3-fold. In unstimulated control cells, 6-MP treatment reduced the number of GFP+ cells from 30.1% to 15.8% but did not significantly reduce GFP expression in the Nurr1 KD cells. Importantly, at the end of a chase experiment, the numbers of GFP+ cells in the control cells were reduced from 95% to 46.8% in the absence of 6-MP and were further decreased to 28.8% when cultured in the presence of 6-MP. In contrast, the numbers of GFP+ cells in Nurr1 KD cells were only slightly decreased (from 96.3% to 91.6%) in the absence of 6-MP and remained at 86.6% when cultured in the presence of 6-MP. To further confirm that 6-MP silences HIV via this mechanism, we stimulated G9a KO HC69 cells and the control cells (S3B Fig) with high dose TNF-α for 24 hr, followed by a chase. At the end of the chase period, the number of GFP+ cells in the vector infected control cells went down from 91.7% to 54% in the absence of 6-MP and further decreased to 31.7% in the presence of 6-MP. In contrast, the numbers of GFP+ cells in G9a KO cells only slightly decreased from 91.9% to 90.4% in the absence of 6-MP and remained at 86.6% in the presence of 6-MP. We next conducted ChIP assays to determine if 6-MP treatment enhances recruitment of the CoREST repressor complex to HIV 5’LTR. As expected, 6-MP treatment dramatically increased the levels of CoREST and G9a at the HIV LTR but only had a modest effect on HDAC1 levels (S3C Fig). Finally, to confirm that Nurr1 is also critical for the silencing of HIV in primary microglial cells, we infected iPSC-derived human microglial cells (iMG) with the same HIV reporter virus described earlier (Fig 1A). About 50% of the iMG became GFP+ two days after HIV infection (Fig 11A). We then treated the infected iMG with 6-MP and another Nurr1 agonist, amodiaquine (AQ) [48,55], for four days. Both 6-MP and AQ decreased the number of GFP+ cells in a dose-dependent manner (Fig 11B and 11C) and lowered the levels of HIV un-spliced transcripts (Fig 11D). Both agonists also dose-dependently reduced MMP2 mRNA in iMG (Fig 11E). Collectively, results from both hμglia and iMG strongly suggested an important role for Nurr1 in HIV silencing in microglial cells. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 11. Nurr1 Mediates HIV silencing in iPSC-derived microglial cells (iMG). A, Representative phase contrast, GFP, and overlapped images of iMG that were un-infected or infected with the reporter HIV-1 shown in Fig 1A, at 48 hr post-infection (hpi). HIV-infected iMG were treated with different doses of Nurr1 agonist 6-MP or AQ for four days, followed by flow cytometry analysis of GFP expression. B, The average levels of GFP expression in iMG treated with various doses of 6-MP were calculated from three replicates. C, The average levels of GFP expression in iMG treated with various doses of AQ were calculated from three replicates. D, The average levels of HIV RNA (un-spliced) in the cells described in panels A and B, were measured by RT-qPCR and calculated from three replicates of qRT-PCR. The mean value of HIV transcript from HIV infected but untreated cells was referred to as level “1”. E, The mRNA level of Nurr1 target gene MMP2 in the same cells described in D was measured by qRT-PCR. The mean value of MMP2 mRNA from three replicates of un-infected cells was referred to as level “1”. The average levels of HIV transcript and MMP2 mRNA in each sample were calculated from three technical replicates. Differences in HIV and MMP2 mRNA levels between untreated cells and cells treated with different doses of 6-MP or AQ were statistically significant (** p-values <0.001). HIV transcripts were only detected in infected iMG cells (panel D). MMP2 mRNA was significantly elevated in HIV infected iMG (panel E). https://doi.org/10.1371/journal.ppat.1010110.g011 Thus, our results clearly demonstrate that 6-MP silences HIV through activation of Nurr1 and fortifies our hypothesis that this Nurr1 agonist silences HIV by engaging the CoREST repressor complex.

[END]
---
[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010110

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/