Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Harmine enhances the activity of the HIV-1 latency-reversing agents ingenol A and SAHA

Harmine enhances the activity of the HIV-1 latency-reversing agents ingenol A and SAHA © 2020. Published by The Company of Biologists Ltd | Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 RESEARCH ARTICLE Harmine enhances the activity of the HIV-1 latency-reversing agents ingenol A and SAHA Jared P. Taylor*, Lucas H. Armitage, Daniel L. Aldridge, Melanie N. Cash and Mark A. Wallet ABSTRACT promoter) or active suppression (binding of inhibitory proteins to the LTR or epigenetic silencing of the locus), the outcome is the Infection with human immunodeficiency virus 1 (HIV-1) remains same – HIV is not produced, and the virus remains hidden from host incurable because long-lived, latently-infected cells persist during immunity. In this state, HIV endures for as long as the infected cell prolonged antiretroviral therapy. Attempts to pharmacologically and all progeny from future cell divisions. If replication-competent reactivate and purge the latent reservoir with latency reactivating proviral genomes are harbored, an individual is at risk for viral agents (LRAs) such as protein kinase C (PKC) agonists (e.g. ingenol reactivation. A) or histone deacetylase (HDAC) inhibitors (e.g. SAHA) have shown Even after HIV replication has been pharmacologically promising but incomplete efficacy. Using the J-Lat T cell model of HIV suppressed so that the virus is undetectable in peripheral blood, latency, we found that the plant-derived compound harmine enhanced cessation of treatment leads to the resumption of the HIV/AIDS the efficacy of existing PKC agonist LRAs in reactivating latently- clinical progression (Davey et al., 1999). Elimination of the latent infected cells. Treatment with harmine increased not only the number reservoir or significant reduction of the size of the reservoir are seen of reactivated cells but also increased HIV transcription and protein as the only real hopes for a sterilization or functional cure, expression on a per-cell basis. Importantly, we observed a synergistic respectively. A large body of work is focused on pharmacological effect when harmine was used in combination with ingenol A and the strategies to ‘purge’ the latent reservoir. The most widely studied HDAC inhibitor SAHA. An investigation into the mechanism revealed approach is known as ‘shock and kill’ or ‘kick and kill’ (Deeks, that harmine, when used with LRAs, increased the activity of NFκB, 2012; Hamer, 2004). The goal is to simultaneously reactivate all MAPK p38, and ERK1/2. Harmine treatment also resulted in reduced (or most) replication-competent latent HIV while maintaining expression of HEXIM1, a negative regulator of transcriptional antiretroviral therapy (ART). Ideally, upon re-expression of viral elongation. Thus, harmine enhanced the effects of LRAs by mRNA and proteins, the reservoir cells will die through either increasing the availability of transcription factors needed for HIV cytopathic effects of the virus or through immune mechanisms, such reactivation and promoting transcriptional elongation. Combination as cytotoxic T lymphocytes (CTL) or natural killer (NK) cells. What therapies with harmine and LRAs could benefit patients by achieving has been lacking is a safe and effective pharmacological method to deeper reactivation of the latent pool of HIV provirus. potently reactivate latent HIV in vivo. KEY WORDS: Harmine, Latency, HIV HIV latency is predominantly controlled by two basic mechanisms: chromatin accessibility and transcription factor INTRODUCTION expression/activation/localization. Epigenetic regulation of the Infection with human immunodeficiency virus 1 (HIV-1) remains HIV integration site through modification of histone proteins by incurable because of the ability of this virus to integrate its genetic histone deacetylase (HDAC) enzymes effectively silences HIV material permanently into the host genome of infected cells. The mRNA expression. HDAC inhibitors such as vorinostat (SAHA) single-stranded RNA genome of HIV requires a stable intermediate can elicit HIV replication from latently infected cells in vitro in the form of an integrated proviral DNA to complete its life cycle. (Archin et al., 2009a,b; Contreras et al., 2009; Lehrman et al., 2005). This integrated DNA is subject to the same regulatory mechanisms In vivo, vorinostat can induce some reactivation and increase plasma as host genes including requirements for active transcription factors HIV RNA in subjects receiving ART (Archin et al., 2012). as well as epigenetic control of chromatin structure and accessibility However, thus far HDAC inhibitors have not been able to (Siliciano and Greene, 2011). significantly reduce the pool of latently infected cells. Like host genes, the HIV genome may enter a state of non- Transcription of HIV mRNA relies on the interaction of the viral expression wherein viral mRNA and proteins are not expressed protein Tat and the pTEFb complex. The pTEFb complex is a (Folks et al., 1986). Whether this non-expression is through neglect positive regulator of transcription elongation and the interaction of (lack of stimuli required to drive the HIV long terminal repeat, LTR, Tat with pTEFb is required for efficient elongation of HIV transcripts (Mancebo et al., 1997; Zhu et al., 1997). The activation of host transcription factors such as NFκB and MAPK Department of Pathology, Immunology & Laboratory Medicine, University of Florida, Gainesville, FL 32610, USA. are also required for reactivation of HIV from latency. These transcription factors are normally sequestered in the cytoplasm as *Author for correspondence ( jptaylor@uab.edu) inactive proteins. Numerous early studies of HIV latency have J.P.T., 0000-0001-7700-530X; L.H.A., 0000-0003-3179-4993; M.N.C., 0000- focused on treatments that stimulate T cell activation including 0003-0144-7878; M.A.W., 0000-0003-0802-9548 cytokines (IL-2 and IL-7), ligation of surface proteins (PHA, anti- This is an Open Access article distributed under the terms of the Creative Commons Attribution CD3) or chemical stimulators of signaling pathways such as protein License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, kinase C (PKC) agonists. NFκB, in particular, is a potent inducer of distribution and reproduction in any medium provided that the original work is properly attributed. HIV mRNA expression and several PKC agonists have been validated for their ability to elicit reactivation of latent HIV. These Received 20 April 2020; Accepted 11 November 2020 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 treatments are even more potent when paired with complementary alone, however, harmine and INDY boosted the reactivation effects drugs that target HDACs or BRD4. Other transcription factors such of ingenol A (Fig. 1A). This effect could be seen with doses as low as SP1, STATs, and IRFs play roles in regulating the HIV LTR. The as 5 µM in J-Lat cells and as low as 2.5 µM in primary CD4+ T cells NFAT family of transcription factors may also regulate HIV latency (Fig. S1B,C). Toxicity caused by harmine and INDY was seen with since the HIV LTR contains NFAT transcription factor binding doses above 25 µM (Fig. S1D), therefore, a dose of 20 µM was used sites, and NFAT can enhance HIV mRNA expression. NFAT, like for experiments. NFκB, is sequestered as an inactive protein in the cytoplasm unless When analyzing flow cytometry data, we recognized a unique activated by specific upstream signals. However, unlike NFκB, phenomenon in the J-Lat reactivation experiments. Not only was NFAT is hyperphosphorylated in its inactive state (Macian, 2005). harmine increasing the frequency of cells that became GFP , but it This hyperphosphorylation is driven by a dual-specificity kinase also appeared that the brightness of each GFP cell was increased DYRK1A (Arron et al., 2006; Gwack et al., 2006) and NFAT only when harmine was included (Fig. 1B,C). J-Lat cells were then becomes de-phosphorylated following calcium-dependent activated with increasing doses of ingenol A, PMA, or TNF in the calmodulin signaling (Macian, 2005). presence of DMSO, harmine, or INDY. Both harmine and INDY Harmine is a naturally-occurring tricyclic β-carboline alkaloid boosted the frequency of ingenol A or PMA-activated cells whereas with hallucinogenic properties that is derived from the plant only INDY had a positive effect on reactivation with TNF treatment Banisteriopsis caapi as well as others (Patel et al., 2012). Harmine (Fig. S2). When gating on only the GFP cells, it became clear that has been shown to be an inhibitor of DYRK1A (Adayev et al., 2011; harmine increases the amount of GFP expressed in each reactivated Bain et al., 2007; Göckler et al., 2009). Harmine’s primary target is cell (Fig. S2). This finding suggests that harmine works through a monoamine oxidase A (MAO-A) (Blum et al., 1964; Slotkin and mechanism that is unique from INDY. It also seems to indicate that DiStefano, 1970; TUNG et al., 1965; Yasuhara, 1974), but a well- the boosting effect of harmine is working through at least two reported role for inhibition of DYRK1A leading to enhanced NFAT biological pathways, one that increases the sensitivity of cells to activity has been reported (Adayev et al., 2011; Bain et al., 2007; activating stimuli (reduced dose of stimulus required to activate Egusa et al., 2011; Göckler et al., 2009). The crystal structure of HIV) and a second pathway that enhances the magnitude of HIV harmine complexed with DYRK1A has been solved, confirming its LTR activity in cells where reactivation occurs. binding to the ATP-binding pocket of DYRK1A (Ogawa et al., For a more direct measure of HIV-encoded genes/proteins, J-Lat 2010). Because of its ability to augment NFAT signaling, we cells were activated with ingenol A in the presence or absence of undertook a study of harmine and other DYRK1A inhibitors to harmine, or INDY. Here, like GFP, gag mRNA expression was determine whether these compounds could enhance HIV induced by ingenol A+DMSO and the expression was markedly reactivation alone or in combination with other known latency- increased with the addition of harmine (Fig. 1D). Harmine alone reversing agents (LRAs). We hypothesized that DYRK1A failed to induce gag expression. However, INDY alone did result in inhibitors would augment reactivation with LRAs through increased gag expression (Fig. 1D) even though INDY treatment increased NFAT availability. Here, we report that harmine and alone did not result in increased activation of J-Lat 5A8 cells another DYRK1A inhibitor, INDY (Ogawa et al., 2010), boost HIV (Fig. 1A). Next, western blot analysis was performed to measure reactivation by PKC agonists. Interestingly, the effect was gag protein expression. Gag protein expression was induced by independent of NFAT activity; harmine was effective at boosting ingenol A alone and markedly increased with the addition of LRAs even when no evidence of NFAT activity was detected. In harmine or INDY. The effect was stronger for harmine than with addition, CRISPR knockout of DYRK1A did not mimic the effects INDY (Fig. 1E). To better measure the amount of gag being of harmine or modulate the efficacy of harmine for boosting HIV expressed on a per cell basis, flow cytometry sorting was used to + − reactivation, Instead, we found that harmine enhances MAPK and separate GFP versus GFP cells after ingenol A treatment in the NFκB signalling, leading to increased HIV transcription. Using presence or absence of harmine or INDY. Gag protein expression + − whole-genome microarray we observed that harmine modulates was measured in the GFP and GFP cells by western blot. The expression of key pTEFb components. HEXIM1 expression is GFP cells that were treated with ingenol A+harmine expressed significantly downregulated by harmine whereas the cyclin CCNT2 more gag protein than GFP cells that were treated with ingenol was upregulated. We conclude that harmine regulates pTEFb A+INDY (Fig. 1F&G). Again, these findings indicate that harmine complex heterogeneity and establishes an environment that is more is mediating two distinct effects on HIV latency. One effect conducive to HIV reactivation. Combination treatments with increases the sensitivity of cells to activating LRAs whereas a harmine and LRAs may prove efficacious in vivo. second effect increases the expression level of HIV genes/proteins. This finding leads to two key questions: does harmine enhance HIV RESULTS reactivation through enhanced NFAT activity? And is DYRK1A the DYRK1A inhibitors enhance the efficacy of T-cell activating target of harmine in HIV latency models? PKC agonists PKC agonists have been previously shown to reactivate latent HIV Harmine boosts HIV reactivation independent of NFAT (Brogdon et al., 2016; Díaz et al., 2015; Jiang et al., 2015; Laird DYRK1A negatively regulates NFAT by hyperphosphorylation, et al., 2015; Martínez-Bonet et al., 2015; Perez et al., 2010; Reuse which excludes NFAT from the nucleus. Only through calcium- et al., 2009; Williams et al., 2004). We expected reactivation with dependent calmodulin activation is NFAT de-phosphorylated PKC agonists to be enhanced by combinatorial treatment with resulting in nuclear translocation and activation of NFAT DYRK1A inhibitors, harmine or INDY (Fig. S1A). To test this, we dependent genes such as IL-2. The HIV LTR contains at least two used J-Lat 5A8 reporter cells, which are latently infected with a full- NFAT binding motifs and since harmine is a known inhibitor of length provirus that expresses GFP in place of Nef as an indicator of DYRK1A, we wanted to determine if the anti-latency effects of LTR activity. J-Lat 5A8 cells were pretreated with harmine or INDY harmine are partly due to enhanced NFAT activity. and reactivated with the PKC agonist ingenol A. Both harmine and First, we wanted to determine whether harmine affects NFAT INDY failed to cause any measurable HIV reactivation when used activity in T cells. Jurkat T cells were transduced with lentiviral Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 1. DYRK1A inhibitors enhance the efficacy of PKC agonist ingenol A. (A) The percentage of GFP J-Lat 5A8 cells after reactivation with ingenol A in the presence of inhibitors (n=6), (B) representative flow cytometry plots of the mean fluorescence intensity of GFP cells, and (C) summary mean fluorescence intensity data (n=6). (D) gag mRNA expression determined by RT-qPCR. (E) Gag protein expression determined by western blot with + − representative blot and densitometry (n=3). GFP and GFP cells were sorted after reactivation with ingenol A. (F) Post-sort purity measured by flow − + + cytometry. The percentage of GFP population (black) and the GFP population (green). (G) Gag protein expression of post-sort GFP cells measured by western blot. Error bars represent standard deviation. Statistical analysis was performed by two-way ANOVA and corrected for multiple comparisons by Tukey’s test. **P<0.01; ***P<0.001; ****P<0.0001; ns, not significant. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 luciferase reporter constructs for NFAT, NFκB or a negative control MFI of GFP cells (Fig. 4C). Although DYRK1A knockout led to a virus with luciferase gene but no promoter. When reporter cells were decrease in the number of GFP cells, treatment with harmine and treated with ionomycin alone, a well-known activator of calmodulin INDY still resulted in an increase in GFP cells (Fig. 4B). This and NFAT, only the NFAT reporter cells expressed luciferase suggests that while DYRK1A expression may be involved in (Fig. 2A). When reporter cells were treated with ingenol A alone, reactivation of cells (Fig. 4B) it is not required and is not the only the NFκB cells expressed luciferase (Fig. 2A). This suggests pathway targeted by harmine and INDY that results in increased that ingenol A-mediated reactivation of J-Lat 5A8 cells is dependent HIV reactivation. on NFκB, not NFAT. Furthermore, treatment of J-Lat 5A8 cells with IKK16, an inhibitor of NFκB signaling, significantly reduced the Harmine downregulates HEXIM1 expression percentage of reactivated cells as well as the MFI of GFP cells in a To better understand how harmine is modulating HIV reactivation, dose-dependent manner (Fig. 2B). To determine if harmine was we used whole-genome microarray to determine which genes were influencing NFAT or NFκB activity, NFAT and NFκB reporter cells differentially expressed in cells treated with harmine. J-Lat cells were treated with ingenol A with harmine or INDY. Harmine had no were treated with DMSO, harmine, PMA, or harmine+PMA. effect on ionomycin-induced luciferase in the NFAT reporter cells. Microarray analysis showed that there were 35 transcripts that were However, harmine significantly boosted luciferase activity in the significantly upregulated or downregulated by at least twofold NFκB reporter cells compared to INDY. INDY had no effect on between PMA and PMA+harmine treatments. A complete list of the luciferase expression in any of the cell lines (Fig. 2C). These data 35 significantly upregulated or downregulated genes are presented suggest that the boosting effects of harmine and INDY are through in Table S1. Fourteen of these transcripts correspond to coding different mechanisms. genes (Fig. 5A and Table 1) and the remaining 21 transcripts were Second, we wanted to determine whether harmine or INDY non-coding transcripts. The gene that was upregulated the most in affected calcium flux upstream of NFAT activation. Jurkat cells PMA+harmine treatment compared to PMA treatment alone was were stained with a calcium-sensitive dye and treated with CCNT2, which codes for the cyclin T2 protein. The gene that was ionomycin or ingenol A. Ionomycin induced a calcium flux as downregulated the most in PMA+harmine treatment compared to expected but ingenol A did not. Neither harmine nor INDY had any PMA treatment alone was HEXIM1. Both genes code for proteins effect on calcium flux regardless of agonist treatment (Fig. 2D). that play a role in transcriptional elongation (Chen et al., 2018). The Harmine is primarily affecting the NFκB pathway with no HIV-encoded transcription factor Tat competes with HEXIM1 for significant involvement of NFAT or the calcium flux. Thus, binding to the pTEFb complex to promote HIV transcription harmine and INDY are working through different mechanisms. (Barboric et al., 2007). Downregulation of HEXIM1 would thus Harmine appears to be acting upon the NFκB pathway. result in less negative regulation of the pTEFb complex resulting in We next wanted to determine whether harmine was only affecting more Tat binding to pTEFb promoting transcript elongation. the NFκB pathway or if it was also affecting other pathways Similarly, upregulation of cyclin T2 would be expected to downstream of PKC. J-Lat cells cultured with PMA +/− harmine increase elongation of transcripts. This combination of effects on were analyzed for phospho-ERK1/2, phospho-MAPKp38, and transcription explains why harmine treatment not only increases the phospho-AKT levels. We found that harmine markedly boosted percentage of reactivated cells but also increases the number of viral phospho-ERK1/2 and phospho-MAPKp38 levels induced by PMA transcripts on a per-cell basis. (Fig. 3A). We found no difference in levels of phospho-AKT To confirm the microarray findings, we treated J-Lat cells with (Fig. 3A). Inclusion of an ERK inhibitor, U0126, resulted in dose- ingenol A in the presence or absence of harmine or INDY. Western blot dependent suppression of ingenol A-induced HIV reactivation analysis showed that harmine treatment alone or with ingenol A leads to (Fig. 3B). Importantly, while the inhibitor markedly reduced the a significant reduction in HEXIM1 protein expression, whereas INDY + + frequency of GFP cells, it also reduced the MFI of GFP cells treatment has no effect (Fig. 5B). The decrease in HEXIM1 protein (Fig. 3C). This reduction in the MFI of the GFP cells was more expression corresponded to an increase in the expression of gag pronounced than the reduction in MFI seen with the inhibition of protein. Thus, harmine treatment alone downregulates HEXIM1 and NFκB. Thus, it appears that harmine boosts the frequency of GFP boosts the efficacy of subsequent LRA stimulation. cells by enhancing sensitivity to PKC agonists as well as increases the magnitude of LTR activity independent of a strong activating Harmine boosts the efficacy of SAHA-induced HIV signal. reactivation We wanted to determine if harmine also enhanced other LRAs Harmine boosts HIV reactivation in the absence of DYRK1A independent of the PKC pathway. We observed that harmine also Since we observed that the boosting effect of harmine was boosts the efficacy of SAHA (vorinostat), an HDAC inhibitor, to independent of NFAT activity and was associated with altered enhance viral reactivation. Harmine only boosts the MFI of SAHA- MAPK signaling, we wanted to confirm that the increase in GFP reactivated cells and not the frequency of reactivated cells (Fig. S3) cells and the increase in the MFI of GFP cells by harmine was because SAHA reactivation is independent of NFκBor NFAT DYRK1A-dependent. To test this, a J-Lat cell line lacking pathways. We next tested the combinatorial effects of ingenol A, DYRK1A expression was derived by CRISPR/Cas9 gene SAHA, and harmine on HIV reactivation. SAHA and harmine both targeting (Fig. 4A). If the phenotype caused by harmine were due boosted the percentage of ingenol A-reactivated cells and that the to its interactions with DYRK1A, then we anticipated that J-Lat cells combination of SAHA and harmine had a more potent effect on lacking DYRK1A would behave similarly to harmine treated cells ingenol A-induced reactivation (Fig. 6A). To determine whether with increased GFP expression in response to activating LRAs. this effect was synergistic, we used the Bliss Independence Model DYRK1A knockout did lead to a significant decrease in the of drug synergy and found that the observed combined effects of percentage of GFP cells. However, this decrease was also seen in SAHA and harmine were significantly greater than those predicted knockout cells not treated with an inhibitor (DMSO control) by the model indicating that the combination of SAHA and harmine (Fig. 4B). Interestingly, DYRK1A knockout had no effect on the is synergistic (Fig. 3C). SAHA did not boost the MFI of ingenol A- Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 2. Harmine’s boosting effect is through NFκB, not NFAT. (A) Ionomycin or ingenol A were titrated on J-Lat 5A8 luciferase reporter cells. The dashed lines indicate the luciferase activity with no ionomycin or ingenol A treatment. (B) NFAT or NFκB luciferase reporter cells reactivated by ionomycin or ingenol A in the presence of inhibitors. Dashed lines represent luciferase activity with no ionomycin or ingenol A treatment. (C) J-Lat 5A8 cells were stained with Calcium Sensor Dye eFluor 514 (2 µM) and treated with DMSO, harmine, or INDY for 30 min. Geometric mean fluorescence was measured by flow cytometry for 240 s. Ionomycin (250 nM) or ingenol A (50 nM) was added after 75 s (arrows). Plots show mean geometric fluorescence intensity over time and the change in area under the curve (n=3). (D) J-Lat 5A8 cells were pretreated with DMSO or an IκB kinase inhibitor, IKK 16, at 1 µM (+) or 10 µM (++) followed by ingenol A treatment (31.25 nM) for 18 h (n=3). Error bars represent standard deviation. Statistical analysis was performed by one-way-ANOVA corrected for multiple comparisons with Dunnett’s test (D) and one-way-ANOVA of the area under the curve corrected for multiple comparisons by Tukey’s test (C). *P<0.05; ***P<0.001; ****P<0.0001; ns, not significant; RLU, relative light units. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 3. Harmine boosts phospho-ERK1/2 and phospho-p38 levels after ingenol A stimulation. J-Lat 5A8 cells were reactivated with ingenol A (31.25 nM) in the presence of inhibitors. (A) Whole cell lysates were analyzed by western blot for phospho-ERK1/2, phospho- AKT, and phospho-p38. (B) J-Lat 5A8 cells were pretreated with DMSO, a MEK inhibitor (U0126), or an IκB kinase inhibitor (IKK 16) for 30 min followed by overnight treatment with ingenol A (100 nM). GFP expression was assessed by flow cytometry. The percentage and (C) mean fluorescence intensity of the GFP cells are shown. Dashed lines represent treatment with ingenol A alone. reactivated GFP cells. However, harmine did boost the MFI of For the shock and kill approach to be effective all replication- ingenol A-reactivated GFP cells and the combination of harmine competent cells must be reactivated to prevent viral rebound after and SAHA had an even greater boosting effect (Fig. 6B). Similarly, removal of ART. One strategy being evaluated for improving the we found that while harmine+ingenol A increases gag mRNA shock and kill method is using combinations of drugs that act (Fig. 6D) and gag protein (Fig. 6E) expression compared to ingenol synergistically to reactivate a greater number of latently infected A alone that harmine+SAHA+ingenol A induced significantly more cells (Laird et al., 2015). In the current study, we demonstrate that gag expression. the plant-derived compound harmine used in combination with ingenol A increases the efficacy of latent HIV-1 reactivation. Our DISCUSSION data show that harmine has two effects: it increases the number of The primary goal of the ‘shock and kill’ strategy is to reactivate cells that are reactivated by ingenol A, and it increases transcription of HIV so that reservoir cells will be killed by the LTR promoter activity in reactivated cells resulting in increased cytopathic effects of the virus or expression of viral proteins will viral transcripts and viral proteins (Fig. 7A,B). enable recognition and killing by the immune system. LRAs In the J-Lat model, GFP acts as a surrogate measure for LTR targeting a diverse range of pathways have been proposed for use activity. Our data demonstrate that harmine treatment results in but most fall short of reactivating all replication-competent cells. increased GFP expression on a per-cell basis suggesting that viral Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 4. Harmine boosts independently of DYRK1A. DYRK1A was knocked out with CRISPR/Cas9 in J-Lat 5A8 cells. (A) Western blot of J-Lat 5A8 cells after treatment with CRISPR compared to Jurkat cells (Ctrl). The CRISPR knockout cells were reactivated with ingenol A (31.25 nM) in the presence of inhibitors. (B) The percentage of GFP cells and the (C) mean fluorescence Fig. 5. Harmine downregulates HEXIM1 expression. J-Lat 5A8 cells were intensity of GFP cells was measured by flow cytometry. Error bars treated pretreated with DMSO, harmine, or INDY for 30 min followed by represent standard deviation. Statistical analysis was performed by PMA stimulation for 2 hours and analyzed by whole transcriptome two-way ANOVA corrected for multiple comparisons with Bonferroni’s test. microarray. (A) Hierarchical clustering of transcripts that were upregulated or **p<0.01; ***p<0.001. downregulated at least twofold between PMA and harmine+PMA treatments. (B) Representative western blot analysis of HEXIM1 expression and densitometry (n=3). Error bars represent standard deviation. gene expression would also occur. Indeed, gag expression was Statistical analysis was performed by two-way ANOVA corrected for increased at both the transcript level and protein level as multiple comparisons with Tukey’s test. **P<0.01; ***P<0.001. demonstrated by RT-qPCR and western blot, respectively (Fig. 1). Increased expression of viral proteins will make the ‘kill’ phase of The combination of these three drugs promoted reactivation of a the ‘shock and kill’ approach more effective. Our data also greater number of cells and increased viral protein production in the demonstrate that the combination of harmine and SAHA has a reactivated cells (Fig. 6D,E). Interestingly, we found that harmine synergistic effect in boosting HIV reactivation by ingenol A (Fig. 6). also boosted the reactivation effects of SAHA alone (Fig. S3), Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Table 1. PMA-Induced Gene Expression with Harmine Martínez-Bonet et al., 2015; Perez et al., 2010; Reuse et al., 2009; Williams et al., 2004). This synergism is due to two different Fold-Change (PMA+harmine mechanisms for latency being targeted: availability of transcription Gene Symbol versus PMA alone) factors and either epigenetic modifications or availability of pTEFb CCNT2 2.2142 complexes. In our current study, we wanted to investigate whether SOX30 2.10846 the addition of the compound harmine would also act synergistically TMEM2 −2.0027 RAB23 −2.03377 with PKC agonists to reactivate latently infected cells. We predicted ANKRD50 −2.12436 harmine would boost reactivation by increasing availability of the IL27RA −2.1761 transcription factor NFAT, since harmine has been shown to inhibit RASA2 −2.3009 DYRK1A (Bain et al., 2007; Göckler et al., 2009), a negative SLC7A11 −2.37159 regulator of NFAT signaling (Adayev et al., 2011; Bain et al., 2007; LRRC8B −2.3876 Egusa et al., 2011; Göckler et al., 2009). However, in the current SKIL −2.4415 USP27X −2.44667 study, we demonstrate that the boosting effect is independent of CHAC1 −2.52968 NFAT (Fig. 2). HEXIM1 −3.29431 Since harmine is a bona fide DYRK1A inhibitor and has been shown to bind to DYRK1A in crystal structures, we hypothesized that harmine may be modifying the function of DYRK1A or which is likely due to harmine and SAHA working on different directing it into a different pathway. We tested if harmine treatment pathways. had any effects on signaling through the MEK/ERK and p38 MAPK Previous studies have demonstrated that PKC agonists such as pathways when used in combination with a PKC agonist. We found prostratin, bryostatin, and ingenol A work synergistically with that harmine treatment resulted in increased signaling through these HDAC inhibitors as well as the BRD4 inhibitor JQ1 (Brogdon et al., pathways in contrast to INDY. While both harmine and INDY 2016; Díaz et al., 2015; Jiang et al., 2015; Laird et al., 2015; boosted the frequency of reactivated cells after treatment with PKC Fig. 6. SAHA and harmine act synergistically to boost ingenol A activation of J-Lat cells. J-Lat 5A8 cells were pretreated with DMSO, harmine, SAHA, or SAHA+harmine for 30 min followed by ingenol A (31.25 nM) stimulation for 18 h. (A) The percentage of GFP cells and (B) the mean fluorescence intensity (n=3). (C) Synergy calculated using the Bliss Independnce Model. Expression of gag mRNA and gag protein were measured by (D) qPCR and (E) western blot, respectively (n=3). Statistical analysis was performed by calculating the area under the curve and performing a two-way ANOVA corrected for multiple comparisons with Tukey’s test (A,B) or by one-way ANOVA analysis corrected for multiple comparisons with Tukey’s test (D,E). Statistical analysis for (C) was performed by a ratio paired t-test comparing the predicted affected fraction fa to the observed affected fraction fa .*P<0.05; **P<0.01; HS,P HS,O ****P<0.0001; ns, not significant. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 7. Harmine treatment results in increased HIV reactivation by PKC-agonists. (A) Combinatorial treatment of ingenol A and harmine results in + + increased frequency of GFP cells and increased MFI in GFP cells in J-Lat model. (B) Combinatorial treatment of ingenol A and harmine results in increased LTR activity and expression of HIV RNA. (C) Harmine treatment in combination with PKC agonists results in increased availability of transcription factor NFκB, increased MAPK p38 and ERK1/2 activity, and decreased HEXIM1 expression. agonists, only harmine boosted the MFI of the reactivated cells. This Microarray data demonstrated that harmine treatment reduced suggested that harmine may have a secondary effect on DYRK1A. HEXIM1 levels. This was confirmed at the protein level by western However, when we knocked out DYRK1A using CRISPR/Cas9, the blot analysis (Fig. 5). HEXIM1 is a component of the pTEFb boosting effect of harmine was still observed (Fig. 4). This indicates complex. The pTEFb complex consists of multiple proteins that enhanced HIV reactivation seen with harmine treatment is not including CycT1, CDK9, and HEXIM1, and there are dependent on interaction with DYRK1A. heterogeneous combinations of these proteins that determine the Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 specific activity (or inactivity) of pTEFb. A 7SK snRNP complex Thus, the combination of ingenol A, SAHA, and harmine results containing HEXIM1, CDK9, CycT1, MePCE, and LARP7 along in increased reactivation of latently infected J-Lat cells and with a 7SK snRNA holds CDK9 in an inactive complex (Michels increased expression of viral proteins. This is likely due to et al., 2004; Nguyen et al., 2001; Yang et al., 2001; Yik et al., 2003). harmine treatment increasing the availability of NFκB, enhancing A key role for HIV Tat protein is to release CDK9/CycT1 pTEFb the effect of the PKC agonist, while simultaneously boosting the from 7SK snRNA so that it may drive HIV mRNA transcription activation of the MEK/ERK pathway that leads to increased (Fujinaga et al., 2004; Ivanov et al., 2000; Ping and Rana, 2001). transcript elongation. When used in combination with SAHA, Alternatively, CDK9/CycT1 can be bound by BRD4 and this drives which inhibits HDACs and increases the availability of the LTR expression of host genes such as c-Myc (Dey et al., 2009; Jang et al., promoter, the result is a potent drug combination that leads to 2005; Yang et al., 2005, 2008). However, BRD4 competes with Tat enhanced reactivation of latently infected cells. Our findings for pTEFb (Li et al., 2013). It has been shown that the BRD4 suggest that combination therapies that use currently existing inhibitor JQ1 enhances Tat-induced HIV transcription and reversal LRAs would be enhanced by the addition of harmine. Additional of latency (Banerjee et al., 2012; Bartholomeeusen et al., 2012). studies are needed to determine the molecular target of harmine that Thus, the pTEFb transcription complex acts as a three-position results in the phenotype we have reported here before harmine is switch relative to HIV and host gene mRNA expression. HEXIM1- used in vivo. containing complexes are in the off position for all pTEFb- dependent genes. BRD4-containing complexes are in the off MATERIALS AND METHODS position for HIV transcription but the on position for host genes Reagents and resources such as c-Myc. Finally, Tat-containing complexes are in the on A list of key antibodies, inhibitors, commercial kits, and other position for HIV transcription. reagents can be found in Table S2. Our microarray analysis revealed increased expression of CCNT2, the gene that codes for CyclinT2, and decreased Reactivation experiments expression of HEXIM1. While CyclinT1 is the kinase normally J-Lat 5A8 cells are treated with DMSO, harmine (20 µM), INDY associated with the pTEFb complex CyclinT2 has also been (20 µM), or SAHA (5 µM) for 30 min followed by treatment with reported to associate with CDK9 in the PTEF-b complex. Unlike agonists for 18 h unless otherwise indicated in figure legend. CyclinT1, which promotes Tat activity, CyclinT2 has been reported to be inhibitory for HIV-1 transcription (Napolitano et al., 1999). It Cell culture is unclear whether increased expression of CyclinT2 in our J-Lat J-Lat 5A8 cells, a kind gift from Warner Greene (University of model is having a negative effect on HIV transcription. HEXIM1, California, San Francisco, CA, USA), have been previously when part of the pTEFb complex, is also inhibitory for HIV-1 described (Chan et al., 2013). Briefly, J-Lat 5A8 cells are Jurkat replication (Barboric et al., 2007; Michels et al., 2004; Yik et al., cells that are latently infected with a full-length provirus integrated 2003). Therefore, decreased HEXIM1 expression could account for into the MAT2A gene and have the gfp reporter gene in place of nef.A at least some of the boosting effect of harmine. frameshift resulting in defective env production renders the cells non- Signaling through the TCR activates NFκB and NFAT and it has infectious. Any stimulus that activates the LTR will result in been previously reported that TCR signaling enhances transcription of gfp. J-Lat 5A8 cells and Jurkat luciferase reporter transcriptional elongation of latent HIV-1 by activating pTEFb cells were cultured in RPMI 1640 with 2 mM L-glutamine (Corning) through an ERK-dependent pathway (Kim et al., 2011). We found supplemented with 10% heat-inactivated FBS (Sigma-Aldrich, that treatment with ingenol A leads to ERK-1/2 phosphorylation St. Louis, MO, USA) and 100 U/mL penicillin-streptomycin. and that this effect was enhanced by treatment with harmine (Fig. 3), which may be the mechanism by which harmine treatment Primary CD4 T Cells enhances LTR promoter activity through pTEFb. Our data suggest Buffy coats were obtained from Life South Community Blood that harmine is increasing transcriptional elongation through Center (Gainesville, FL, USA) under approval by the Institutional increased availability of pTEFb which may act synergistically with Review Board at the University of Florida. PBMCs were processed the BRD4 inhibitor JQ1. as previously described (Taylor et al., 2018) and CD4 T cells were A recent study by Booiman and colleagues demonstrated that the enriched with the CD4 T Cell Isolation Kit (Miltenyi Biotec) DYRK1A inhibitor, INDY, reactivated latently infected J-Lat cells according to the manufacturer’s instructions. CD4 T cells were without the use of a PKC agonist (Booiman et al., 2015). In our cultured in Advanced RPMI 1640 (Corning) supplemented with hands, however, treatment of J-Lat cells with DYRK1A inhibitors 2 mM GlutaMAX (Gibco), 10% heat-inactivated FBS (Sigma- alone was not sufficient to reactivate latently infected cells as Aldrich), and 100 U/ml penicillin-streptomycin. measured by GFP expression. However, INDY treatment alone did result in increased gag mRNA expression (Fig. 1D). The dose that Flow cytometry and flow sorting we used for our study (20 µM) is much lower than the doses used by Cells were fixed in 2% paraformaldehyde and GFP fluorescence Booiman and colleagues. When higher doses of INDY were used in was measured using the BD Accuri™ C6 flow cytometer (BD our studies, excessive toxicity for J-Lat cells was observed (not Biosciences). Flow sorting was performed on unfixed cells the shown). Treatment with DYRK1A inhibitors only affected J-Lat FACS Aria III (BD Biosciences) at the University of Florida Center cells that were stimulated with PKC agonists, SAHA, or TNF. The for Immunology and Transplantation. study by Booiman and colleagues used J-Lat clones 8.4 and A1, whereas we used clone 5A8, which may account for the differences AlamarBlue assay (Booiman et al., 2015). We did not observe an increase of Gag Jurkat T cells were treated with different doses of harmine, INDY, or expression by western blot in cells treated with INDY alone to an equivalent volume of DMSO overnight. Toxicity of the inhibitors correlate with the increased in gag mRNA levels. This is likely due was measured by adding 10 µl of AlamarBlue Cell Viability Assay to the higher sensitivity of qPCR compared to western blot. Reagent (Thermo Scientific) to 90 µl of cells and waiting for color Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 development. Absorbance was measured at 570 and 600 nm. The CRISPR/Cas9 Knockout of DYRK1A percent reduction was calculated according to the manufacturer’s J-Lat 5A8 cells were co-transfected with two plasmids. One plasmid instructions. on the pCRISPR-CG01 backbone (GeneCopoeia) codes for recombinant Cas9 and a sgRNA (5′-GCCAAACATAAGTGACC- Western blot AAC-3′) that targets exon 2 of DYRK1A. The second plasmid on Western blots were performed as previously described (Taylor the pDONOR-D01 backbone (GeneCopoeia) has a mCherry-T2A- et al., 2018). Puro reporter cassette flanked by homology regions adjacent to the sgRNA target site in the genome. After co-transfection, the J-Lat RT-qPCR cells were selected with puromycin (1 µg/ml). Total RNA was isolated with the RNeasy Plus Mini Kit (Qiagen). cDNA synthesis was carried out with High-Capacity cDNA Reverse Drug synergy calculations Transcription Kit (Applied Biosystems, Foster City, CA, USA) Synergy between harmine and SAHA was calculated using the Bliss according to the manufacturer’s instructions. RT-qPCR reactions Independence Model as previously described (Laird et al., 2015). were carried out in SYBR™ Select Master Mix (Thermo Fisher The predicted fraction fa can be calculated using the equation HS,P Scientific, Waltham, MA, USA) according to the manufacturer’s fa =fa +fa −( fa fa ) fa =fraction of GFP cells reactivated HS,P H S H S H instructions. RT-qPCR reactions were performed on the by harmine with ingenol A fa =fraction of GFP cells reactivated by StepOnePlus™ Real-Time PCR System (Applied Biosystems). SAHA with ingenol A fa =fraction of GFP cells reactivated HS,O Gag primers used were forward: 5′- GAGCTAGAACGATTCGC- by harmine, SAHA, and ingenol A. To determine if the combination AGTTA-3′; reverse: 5′- CTGTCTGAAGGGATGGTTGTAG-3′. of harmine and SAHA is synergistic, the experimentally observed fraction of cells reactivated by harmine and SAHA ( fa ) can be HS,O Luciferase assays compared to the predicted fraction of reactivated GFP cells fa HS,P Reporter cell lines were created by transducing Jurkat (E6.1) T cells using the following equation: with Cignal Lenti NFAT reporter (Catalogue number CLS-015 L), Cignal Lenti NFκB reporter (Catalogue number CLS-013L), or D fa ¼ fa  fa : HS HS;O HS;P Cignal Lenti Negative Control (CLS-NCL) lentiviral particles purchased from Qiagen. An equal volume of Bright-Glo™ If ΔFa is greater than 1 then the combination is synergistic. If HS Luciferase Assay System (Promega) was added to an equal ΔFa is less than 1 then the interaction is antagonistic. If Δfa is HS HS volume of cells. After 5 min, the cells were lysed, and the lysate equal to 0 than the mechanisms are independent of each other. was transferred to a black Costar EIA/RIA polystyrene half area 96-well plate (Corning). Luminescence was measured with the Statistical analysis VICTOR™ X4 Multi-Plate Reader (PerkinElmer). All statistical analyses were performed using GraphPad Prism for Windows (GraphPad Software, La Jolla, CA, USA). A P-value of Microarrays less than 0.05 was considered statistically significant. J-Lat 5A8 cells were treated with DMSO or harmine (20 µM) for Acknowledgements 30 min. After pretreatment, the cells were treated with either The authors would like to thank Warner Greene (University of California, medium or PMA (10 nM) for 2 h. Total RNA was collected with the San Francisco CA, USA), for generously providing the J-Lat 5A8 cells used in the RNeasy Plus Mini Kit (Qiagen). Gene expression was assessed with study. GeneChip™ Human Transcriptome Array 2.0 arrays (Affymetrix) by the Interdisciplinary Center for Biotechnology Research at the Competing interests The authors declare no competing or financial interests. University of Florida. The analysis was performed with Partek Genomics Suite v. 6.6 (Partek Inc., St. Louis, MO, USA). CEL files Author contributions were imported and the raw data were subjected to multi-array average Conceptualization: J.P.T., M.A.W.; Methodology: J.P.T., M.A.W.; Validation: J.P.T., (RMA) background correction and quantile normalization. Probesets M.A.W.; Formal analysis: J.P.T., M.A.W.; Investigation: J.P.T., L.H.A., D.L.A., were summarized by the median polish method and the summarized M.N.C., M.A.W.; Resources: M.A.W.; Data curation: J.P.T.; Writing - original draft: J.P.T., M.A.W.; Writing - review & editing: J.P.T., M.A.W.; Visualization: J.P.T., signals were transformed to log base 2. A one-way ANOVA with M.A.W.; Supervision: M.A.W.; Project administration: M.A.W.; Funding acquisition: contrast was performed to determine fold changes between PMA and M.A.W. PMA+harmine-treated groups. Transcripts that were significantly (step-up FDR <0.05) upregulated or downregulated by at least twofold Funding are listed in Table S1 (coding and noncoding) and Table 1 (coding). This work was supported by funding from the National Institutes of Health grants R56AI108434 and R56AI122813. J.P.T. was supported by the Ruth L. Kirschstein National Research Service Award Institutional Research Training Grant Calcium flux assays T32AI007110. J-Lat 5A8 cells were incubated for 30 min with the Calcium Sensor Dye eFluor 514 (eBioscience) at a concentration of 2 µM with Data availability DMSO, harmine (20 µM), or INDY (20 µM). The cells were washed Microarray data are available in Gene Expression Omnibus (GEO) under accession GSE136172. and resuspended in sample buffer (1X PBS, 0.1% BSA, 2 mM EDTA) containing DMSO, harmine (20 µM), or INDY (20 µM). Supplementary information Flow cytometry data were collected for 1 min and then ionomycin Supplementary information available online at (250 nM) or ingenol A (50 nM) was added and flow cytometry data https://bio.biologists.org/lookup/doi/10.1242/bio.052969.supplemental were collected for an additional 3 min. The change in the area under References the curve (ΔAUC) was calculated by dividing the area under the Adayev, T., Wegiel, J. and Hwang, Y.-W. (2011). Harmine is an ATP-competitive curve after the addition of ionomycin or ingenol A divided by the inhibitor for dual-specificity tyrosine phosphorylation-regulated kinase 1A (Dyrk1A). Arch. Biochem. Biophys. 507, 212-218. doi:10.1016/j.abb.2010.12.024 AUC before the addition of ionomycin or ingenol A. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Archin, N. M., Espeseth, A., Parker, D., Cheema, M., Hazuda, D. and Margolis, Gö ckler, N., Jofre, G., Papadopoulos, C., Soppa, U., Tejedor, F. J. and Becker, W. D. M. (2009a). Expression of latent HIV induced by the potent HDAC inhibitor (2009). Harmine specifically inhibits protein kinase DYRK1A and interferes with suberoylanilide hydroxamic acid. AIDS Res. Hum. Retroviruses 25, 207-212. neurite formation. FEBS J. 276, 6324-6337. doi:10.1111/j.1742-4658.2009.07346.x doi:10.1089/aid.2008.0191 Gwack, Y., Sharma, S., Nardone, J., Tanasa, B., Iuga, A., Srikanth, S., Okamura, Archin, N. M., Keedy, K. S., Espeseth, A., Dang, H., Hazuda, D. J. and Margolis, H., Bolton, D., Feske, S., Hogan, P. G. et al. (2006). A genome-wide Drosophila D. M. (2009b). Expression of latent human immunodeficiency type 1 is induced by RNAi screen identifies DYRK-family kinases as regulators of NFAT. Nature 441, novel and selective histone deacetylase inhibitors. AIDS 23, 1799-1806. doi:10. 646-650. doi:10.1038/nature04631 1097/QAD.0b013e32832ec1dc Hamer, D. H. (2004). Can HIV be Cured? Mechanisms of HIV persistence and Archin, N. M., Liberty, A. L., Kashuba, A. D., Choudhary, S. K., Kuruc, J. D., strategies to combat it. Curr. HIV Res. 2, 99-111. doi:10.2174/1570162043484915 Crooks, A. M., Parker, D. C., Anderson, E. M., Kearney, M. F., Strain, M. C. Ivanov, D., Kwak, Y. T., Guo, J. and Gaynor, R. B. (2000). Domains in the SPT5 et al. (2012). Administration of vorinostat disrupts HIV-1 latency in patients on protein that modulate its transcriptional regulatory properties. Mol. Cell. Biol. 20, antiretroviral therapy. Nature 487, 482-485. doi:10.1038/nature11286 2970-2983. doi:10.1128/MCB.20.9.2970-2983.2000 Arron, J. R., Winslow, M. M., Polleri, A., Chang, C. P., Wu, H., Gao, X., Neilson, Jang, M. K., Mochizuki, K., Zhou, M., Jeong, H.-S., Brady, J. N. and Ozato, K. J. R., Chen, L., Heit, J. J., Kim, S. K. et al. (2006). NFAT dysregulation by (2005). The bromodomain protein Brd4 is a positive regulatory component of P- increased dosage of DSCR1 and DYRK1A on chromosome 21. Nature 441, TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19, 595-600. doi:10.1038/nature04678 523-534. doi:10.1016/j.molcel.2005.06.027 Bain, J., Plater, L., Elliott, M., Shpiro, N., Hastie, C. J., McLauchlan, H., Jiang, G., Mendes, E. A., Kaiser, P., Wong, D. P., Tang, Y., Cai, I., Fenton, A., Klevernic, I., Arthur, J. S. C., Alessi, D. R. and Cohen, P. (2007). The selectivity Melcher, G. P., Hildreth, J. E. K., Thompson, G. R. et al. (2015). Synergistic of protein kinase inhibitors: a further update. Biochem. J. 408, 297-315. doi:10. Reactivation of Latent HIV Expression by Ingenol-3-Angelate, PEP005, Targeted 1042/BJ20070797 NF-kB Signaling in Combination with JQ1 Induced p-TEFb Activation. PLoS Banerjee, C., Archin, N., Michaels, D., Belkina, A. C., Denis, G. V., Bradner, J., Pathog. 11, e1005066. doi:10.1371/journal.ppat.1005066 Sebastiani, P., Margolis, D. M. and Montano, M. (2012). BET bromodomain Kim, Y. K., Mbonye, U., Hokello, J. and Karn, J. (2011). T-Cell receptor signaling inhibition as a novel strategy for reactivation of HIV-1. J. Leukoc. Biol. 92, enhances transcriptional elongation from latent HIV proviruses by activating 1147-1154. doi:10.1189/jlb.0312165 P-TEFb through an ERK-dependent pathway. J. Mol. Biol. 410, 896-916. doi:10. Barboric, M., Yik, J. H. N., Czudnochowski, N., Yang, Z., Chen, R., Contreras, X., 1016/j.jmb.2011.03.054 Geyer, M., Matija Peterlin, B. and Zhou, Q. (2007). Tat competes with HEXIM1 Laird, G. M., Bullen, C. K., Rosenbloom, D. I. S., Martin, A. R., Hill, A. L., Durand, to increase the active pool of P-TEFb for HIV-1 transcription. Nucleic Acids Res. C. M., Siliciano, J. D. and Siliciano, R. F. (2015). Ex vivo analysis identifies 35, 2003-2012. doi:10.1093/nar/gkm063 effective HIV-1 latency-reversing drug combinations. J. Clin. Invest. 125, Bartholomeeusen, K., Xiang, Y., Fujinaga, K. and Peterlin, B. M. (2012). 1901-1912. doi:10.1172/JCI80142 Bromodomain and extra-terminal (BET) bromodomain inhibition activate Lehrman, G., Hogue, I. B., Palmer, S., Jennings, C., Spina, C. A., Wiegand, A., transcription via transient release of positive transcription elongation factor b Landay, A. L., Coombs, R. W., Richman, D. D., Mellors, J. W. et al. (2005). (P-TEFb) from 7SK small nuclear ribonucleoprotein. J. Biol. Chem. 287, Depletion of latent HIV-1 infection in vivo: a proof-of-concept study. Lancet 36609-36616. doi:10.1074/jbc.M112.410746 (London, England) 366, 549-555. doi:10.1016/S0140-6736(05)67098-5 Blum, B., Weizmann, H., Simpson, G. M., Krasilowsky, D., Kulcsar, I. S. and Li, Z., Guo, J., Wu, Y. and Zhou, Q. (2013). The BET bromodomain inhibitor JQ1 Merskey, H. (1964). Harmine antagonism of drug-induced extra-pyramidal activates HIV latency through antagonizing Brd4 inhibition of Tat-transactivation. disturbances. Psychopharmacologia 6, 307-310. doi:10.1007/BF00413162 Nucleic Acids Res. 41, 277-287. doi:10.1093/nar/gks976 Booiman, T., Loukachov, V. V., van Dort, K. A., van ‘t Wout, A. B. and Kootstra, Macian, F. (2005). NFAT proteins: key regulators of T-cell development and N. A. (2015). DYRK1A Controls HIV-1 Replication at a Transcriptional Level in an function. Nat. Rev. Immunol. 5, 472-484. doi:10.1038/nri1632 NFAT Dependent Manner. PLoS ONE 10, e0144229. doi:10.1371/journal.pone. Mancebo, H. S. Y., Lee, G., Flygare, J., Tomassini, J., Luu, P., Zhu, Y., Peng, J., Blau, C., Hazuda, D., Price, D. et al. (1997). P-TEFb kinase is required for HIV Brogdon, J., Ziani, W., Wang, X., Veazey, R. S. and Xu, H. (2016). In vitro effects of Tat transcriptional activation in vivo and in vitro. Genes Dev. 11, 2633-2644. the small-molecule protein kinase C agonists on HIV latency reactivation. Sci. doi:10.1101/gad.11.20.2633 Rep. 6, 39032. doi:10.1038/srep39032 Martınez-Bonet, ́ M., Clemente, M. I., Serramıa,́ M. J., Munoz, E., Moreno, S. and Chan, J. K., Bhattacharyya, D., Lassen, K. G., Ruelas, D. and Greene, W. C. Munoz-Fernández, M. Á. (2015). Synergistic Activation of Latent HIV-1 (2013). Calcium/calcineurin synergizes with prostratin to promote NF-κB Expression by Novel Histone Deacetylase Inhibitors and Bryostatin-1. Sci. Rep. dependent activation of latent HIV. PLoS ONE 8, e77749. doi:10.1371/journal. 5, 16445. doi:10.1038/srep16445 pone.0077749 Michels, A. A., Fraldi, A., Li, Q., Adamson, T. E., Bonnet, F., Nguyen, V. T., Chen, F. X., Smith, E. R. and Shilatifard, A. (2018). Born to run: Control of Sedore, S. C., Price, J. P., Price, D. H., Lania, L. et al. (2004). Binding of the 7SK transcription elongation by RNA polymerase II. Nat. Rev. Mol. Cell Biol. 19, snRNA turns the HEXIM1 protein into a P-TEFb (CDK9/cyclin T) inhibitor. EMBO 464-478. doi:10.1038/s41580-018-0010-5 J. 23, 2608-2619. doi:10.1038/sj.emboj.7600275 Contreras, X., Schweneker, M., Chen, C.-S., McCune, J. M., Deeks, S. G., Martin, Napolitano, G., Licciardo, P., Gallo, P., Majello, B., Giordano, A. and Lania, L. J. and Peterlin, B. M. (2009). Suberoylanilide hydroxamic acid reactivates HIV (1999). The CDK9-associated cyclins T1 and T2 exert opposite effects on HIV-1 from latently infected cells. J. Biol. Chem. 284, 6782-6789. doi:10.1074/jbc. Tat activity. AIDS 13, 1453-1459. doi:10.1097/00002030-199908200-00003 M807898200 Nguyen, V. T., Kiss, T., Michels, A. A. and Bensaude, O. (2001). 7SK small Davey, R. T., Bhat, N., Yoder, C., Chun, T. W., Metcalf, J. A., Dewar, R., Natarajan, nuclear RNA binds to and inhibits the activity of CDK9/cyclin T complexes. Nature V., Lempicki, R. A., Adelsberger, J. W., Miller, K. D. et al. (1999). HIV-1 and T 414, 322-325. doi:10.1038/35104581 cell dynamics after interruption of highly active antiretroviral therapy (HAART) in Ogawa, Y., Nonaka, Y., Goto, T., Ohnishi, E., Hiramatsu, T., Kii, I., Yoshida, M., patients with a history of sustained viral suppression. Proc. Natl. Acad. Sci. USA Ikura, T., Onogi, H., Shibuya, H. et al. (2010). Development of a novel selective 96, 15109-15114. doi:10.1073/pnas.96.26.15109 inhibitor of the Down syndrome-related kinase Dyrk1A. Nat. Commun. 1, 86. Deeks, S. G. (2012). HIV: Shock and kill. Nature 487, 439-440. doi:10.1038/ doi:10.1038/ncomms1090 487439a Patel, K., Gadewar, M., Tripathi, R., Prasad, S. K. and Patel, D. K. (2012). A review Dey, A., Nishiyama, A., Karpova, T., McNally, J. and Ozato, K. (2009). Brd4 marks on medicinal importance, pharmacological activity and bioanalytical aspects of select genes on mitotic chromatin and directs postmitotic transcription. Mol. Biol. beta-carboline alkaloid “‘Harmine’”. Asian Pac. J. Trop. Biomed. 2, 660-664. Cell 20, 4899-4909. doi:10.1091/mbc.e09-05-0380 doi:10.1016/S2221-1691(12)60116-6 Dıaz, ́ L., Martınez-Bonet, ́ M., Sánchez, J., Fernández-Pineda, A., Jiménez, J. L., Perez, M., de Vinuesa, A., Sanchez-Duffhues, G., Marquez, N., Bellido, M., Munoz, E., Moreno, S., Álvarez, S. and Munoz-Fernández, M. Á. (2015). ̃ ̃ Munoz-Fernandez, M., Moreno, S., Castor, T., Calzado, M. and Munoz, E. Bryostatin activates HIV-1 latent expression in human astrocytes through a PKC (2010). Bryostatin-1 synergizes with histone deacetylase inhibitors to reactivate and NF-κB-dependent mechanism. Sci. Rep. 5, 12442. doi:10.1038/srep12442 HIV-1 from latency. Curr. HIV Res. 8, 418-429. doi:10.2174/157016210793499312 Egusa, H., Doi, M., Saeki, M., Fukuyasu, S., Akashi, Y., Yokota, Y., Yatani, H. and Ping, Y. H. and Rana, T. M. (2001). DSIF and NELF interact with RNA polymerase II Kamisaki, Y. (2011). The small molecule harmine regulates NFATc1 and Id2 elongation complex and HIV-1 Tat stimulates P-TEFb-mediated phosphorylation expression in osteoclast progenitor cells. Bone 49, 264-274. doi:10.1016/j.bone. 2011.04.003 of RNA polymerase II and DSIF during transcription elongation. J. Biol. Chem. 276, 12951-12958. doi:10.1074/jbc.M006130200 Folks, T., Powell, D. M., Lightfoote, M. M., Benn, S., Martin, M. A. and Fauci, A. S. Reuse, S., Calao, M., Kabeya, K., Guiguen, A., Gatot, J.-S., Quivy, V., Vanhulle, (1986). Induction of HTLV-III/LAV from a nonvirus-producing T-cell line: implications for latency. Science 231, 600-602. doi:10.1126/science.3003906 C., Lamine, A., Vaira, D., Demonte, D. et al. (2009). Synergistic activation of HIV- Fujinaga, K., Irwin, D., Huang, Y., Taube, R., Kurosu, T. and Peterlin, B. M. (2004). 1 expression by deacetylase inhibitors and prostratin: implications for treatment of Dynamics of human immunodeficiency virus transcription: P-TEFb phosphorylates latent infection. PLoS ONE 4, e6093. doi:10.1371/journal.pone.0006093 RD and dissociates negative effectors from the transactivation response element. Siliciano, R. F. and Greene, W. C. (2011). HIV latency. Cold Spring Harb. Perspect. Mol. Cell. Biol. 24, 787-795. doi:10.1128/MCB.24.2.787-795.2004 Med. 1, a007096. doi:10.1101/cshperspect.a007096 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Slotkin, T. and DiStefano, V. (1970). Urinary metabolities of harmine in the rat and Yang, Z., Yik, J. H. N., Chen, R., He, N., Jang, M. K., Ozato, K. and Zhou, Q. (2005). their inhibition of monoamine oxidase. Biochem. Pharmacol. 19, 125-131. doi:10. Recruitment of P-TEFb for stimulation of transcriptional elongation by the 1016/0006-2952(70)90333-3 bromodomain protein Brd4. Mol. Cell 19, 535-545. doi:10.1016/j.molcel.2005.06.029 Taylor, J. P., Cash, M. N., Santostefano, K. E., Nakanishi, M., Terada, N. and Yang, Z., He, N. and Zhou, Q. (2008). Brd4 recruits P-TEFb to chromosomes at late Wallet, M. A. (2018). CRISPR/Cas9 knockout of USP18 enhances type I IFN mitosis to promote G1 gene expression and cell cycle progression. Mol. Cell. Biol. responsiveness and restricts HIV-1 infection in macrophages. J. Leukoc. Biol. 28, 967-976. doi:10.1128/MCB.01020-07 103, 1225-1240. doi:10.1002/JLB.3MIA0917-352R Yasuhara, H. (1974). Studies on monoamine oxidase (report XXIV). Effect of TUNG, Y. C., HWANG, Y. S. and WU, C. Y. (1965). Studies on beta- harmine on monoamine oxidase. Jpn. J. Pharmacol. 24, 523-533. doi:10.1254/jjp. carbolines. 3. Effect of harman hydrochloride on monoamine oxidase and the 24.523 uterine contractility induced by n-monomethyltryptamine hydrochloride. Tsa. Yik, J. H. N., Chen, R., Nishimura, R., Jennings, J. L., Link, A. J. and Zhou, Q. Chih. Gaoxiong Yi Xue Yuan. Tong Xue Hui 64, 44-50. (2003). Inhibition of P-TEFb (CDK9/Cyclin T) kinase and RNA polymerase II Williams, S. A., Chen, L.-F., Kwon, H., Fenard, D., Bisgrove, D., Verdin, E. and transcription by the coordinated actions of HEXIM1 and 7SK snRNA. Mol. Cell 12, Greene, W. C. (2004). Prostratin antagonizes HIV latency by activating NF- 971-982. doi:10.1016/S1097-2765(03)00388-5 kappaB. J. Biol. Chem. 279, 42008-42017. doi:10.1074/jbc.M402124200 Zhu, Y., Pe’ery, T., Peng, J., Ramanathan, Y., Marshall, N., Marshall, T., Amendt, Yang, Z., Zhu, Q., Luo, K. and Zhou, Q. (2001). The 7SK small nuclear RNA B., Mathews, M. B. and Price, D. H. (1997). Transcription elongation factor P- inhibits the CDK9/cyclin T1 kinase to control transcription. Nature 414, 317-322. TEFb is required for HIV-1 tat transactivation in vitro. Genes Dev. 11, 2622-2632. doi:10.1038/35104575 doi:10.1101/gad.11.20.2622 Biology Open http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biology Open The Company of Biologists

Harmine enhances the activity of the HIV-1 latency-reversing agents ingenol A and SAHA

Download PDF
13 pages

Loading next page...
 
/lp/the-company-of-biologists/harmine-enhances-the-activity-of-the-hiv-1-latency-reversing-agents-dv244URn9V

References (128)

Publisher
The Company of Biologists
Copyright
© 2021 The Company of Biologists. All rights reserved.
eISSN
2046-6390
DOI
10.1242/bio.052969
Publisher site
See Article on Publisher Site

Abstract

© 2020. Published by The Company of Biologists Ltd | Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 RESEARCH ARTICLE Harmine enhances the activity of the HIV-1 latency-reversing agents ingenol A and SAHA Jared P. Taylor*, Lucas H. Armitage, Daniel L. Aldridge, Melanie N. Cash and Mark A. Wallet ABSTRACT promoter) or active suppression (binding of inhibitory proteins to the LTR or epigenetic silencing of the locus), the outcome is the Infection with human immunodeficiency virus 1 (HIV-1) remains same – HIV is not produced, and the virus remains hidden from host incurable because long-lived, latently-infected cells persist during immunity. In this state, HIV endures for as long as the infected cell prolonged antiretroviral therapy. Attempts to pharmacologically and all progeny from future cell divisions. If replication-competent reactivate and purge the latent reservoir with latency reactivating proviral genomes are harbored, an individual is at risk for viral agents (LRAs) such as protein kinase C (PKC) agonists (e.g. ingenol reactivation. A) or histone deacetylase (HDAC) inhibitors (e.g. SAHA) have shown Even after HIV replication has been pharmacologically promising but incomplete efficacy. Using the J-Lat T cell model of HIV suppressed so that the virus is undetectable in peripheral blood, latency, we found that the plant-derived compound harmine enhanced cessation of treatment leads to the resumption of the HIV/AIDS the efficacy of existing PKC agonist LRAs in reactivating latently- clinical progression (Davey et al., 1999). Elimination of the latent infected cells. Treatment with harmine increased not only the number reservoir or significant reduction of the size of the reservoir are seen of reactivated cells but also increased HIV transcription and protein as the only real hopes for a sterilization or functional cure, expression on a per-cell basis. Importantly, we observed a synergistic respectively. A large body of work is focused on pharmacological effect when harmine was used in combination with ingenol A and the strategies to ‘purge’ the latent reservoir. The most widely studied HDAC inhibitor SAHA. An investigation into the mechanism revealed approach is known as ‘shock and kill’ or ‘kick and kill’ (Deeks, that harmine, when used with LRAs, increased the activity of NFκB, 2012; Hamer, 2004). The goal is to simultaneously reactivate all MAPK p38, and ERK1/2. Harmine treatment also resulted in reduced (or most) replication-competent latent HIV while maintaining expression of HEXIM1, a negative regulator of transcriptional antiretroviral therapy (ART). Ideally, upon re-expression of viral elongation. Thus, harmine enhanced the effects of LRAs by mRNA and proteins, the reservoir cells will die through either increasing the availability of transcription factors needed for HIV cytopathic effects of the virus or through immune mechanisms, such reactivation and promoting transcriptional elongation. Combination as cytotoxic T lymphocytes (CTL) or natural killer (NK) cells. What therapies with harmine and LRAs could benefit patients by achieving has been lacking is a safe and effective pharmacological method to deeper reactivation of the latent pool of HIV provirus. potently reactivate latent HIV in vivo. KEY WORDS: Harmine, Latency, HIV HIV latency is predominantly controlled by two basic mechanisms: chromatin accessibility and transcription factor INTRODUCTION expression/activation/localization. Epigenetic regulation of the Infection with human immunodeficiency virus 1 (HIV-1) remains HIV integration site through modification of histone proteins by incurable because of the ability of this virus to integrate its genetic histone deacetylase (HDAC) enzymes effectively silences HIV material permanently into the host genome of infected cells. The mRNA expression. HDAC inhibitors such as vorinostat (SAHA) single-stranded RNA genome of HIV requires a stable intermediate can elicit HIV replication from latently infected cells in vitro in the form of an integrated proviral DNA to complete its life cycle. (Archin et al., 2009a,b; Contreras et al., 2009; Lehrman et al., 2005). This integrated DNA is subject to the same regulatory mechanisms In vivo, vorinostat can induce some reactivation and increase plasma as host genes including requirements for active transcription factors HIV RNA in subjects receiving ART (Archin et al., 2012). as well as epigenetic control of chromatin structure and accessibility However, thus far HDAC inhibitors have not been able to (Siliciano and Greene, 2011). significantly reduce the pool of latently infected cells. Like host genes, the HIV genome may enter a state of non- Transcription of HIV mRNA relies on the interaction of the viral expression wherein viral mRNA and proteins are not expressed protein Tat and the pTEFb complex. The pTEFb complex is a (Folks et al., 1986). Whether this non-expression is through neglect positive regulator of transcription elongation and the interaction of (lack of stimuli required to drive the HIV long terminal repeat, LTR, Tat with pTEFb is required for efficient elongation of HIV transcripts (Mancebo et al., 1997; Zhu et al., 1997). The activation of host transcription factors such as NFκB and MAPK Department of Pathology, Immunology & Laboratory Medicine, University of Florida, Gainesville, FL 32610, USA. are also required for reactivation of HIV from latency. These transcription factors are normally sequestered in the cytoplasm as *Author for correspondence ( jptaylor@uab.edu) inactive proteins. Numerous early studies of HIV latency have J.P.T., 0000-0001-7700-530X; L.H.A., 0000-0003-3179-4993; M.N.C., 0000- focused on treatments that stimulate T cell activation including 0003-0144-7878; M.A.W., 0000-0003-0802-9548 cytokines (IL-2 and IL-7), ligation of surface proteins (PHA, anti- This is an Open Access article distributed under the terms of the Creative Commons Attribution CD3) or chemical stimulators of signaling pathways such as protein License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, kinase C (PKC) agonists. NFκB, in particular, is a potent inducer of distribution and reproduction in any medium provided that the original work is properly attributed. HIV mRNA expression and several PKC agonists have been validated for their ability to elicit reactivation of latent HIV. These Received 20 April 2020; Accepted 11 November 2020 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 treatments are even more potent when paired with complementary alone, however, harmine and INDY boosted the reactivation effects drugs that target HDACs or BRD4. Other transcription factors such of ingenol A (Fig. 1A). This effect could be seen with doses as low as SP1, STATs, and IRFs play roles in regulating the HIV LTR. The as 5 µM in J-Lat cells and as low as 2.5 µM in primary CD4+ T cells NFAT family of transcription factors may also regulate HIV latency (Fig. S1B,C). Toxicity caused by harmine and INDY was seen with since the HIV LTR contains NFAT transcription factor binding doses above 25 µM (Fig. S1D), therefore, a dose of 20 µM was used sites, and NFAT can enhance HIV mRNA expression. NFAT, like for experiments. NFκB, is sequestered as an inactive protein in the cytoplasm unless When analyzing flow cytometry data, we recognized a unique activated by specific upstream signals. However, unlike NFκB, phenomenon in the J-Lat reactivation experiments. Not only was NFAT is hyperphosphorylated in its inactive state (Macian, 2005). harmine increasing the frequency of cells that became GFP , but it This hyperphosphorylation is driven by a dual-specificity kinase also appeared that the brightness of each GFP cell was increased DYRK1A (Arron et al., 2006; Gwack et al., 2006) and NFAT only when harmine was included (Fig. 1B,C). J-Lat cells were then becomes de-phosphorylated following calcium-dependent activated with increasing doses of ingenol A, PMA, or TNF in the calmodulin signaling (Macian, 2005). presence of DMSO, harmine, or INDY. Both harmine and INDY Harmine is a naturally-occurring tricyclic β-carboline alkaloid boosted the frequency of ingenol A or PMA-activated cells whereas with hallucinogenic properties that is derived from the plant only INDY had a positive effect on reactivation with TNF treatment Banisteriopsis caapi as well as others (Patel et al., 2012). Harmine (Fig. S2). When gating on only the GFP cells, it became clear that has been shown to be an inhibitor of DYRK1A (Adayev et al., 2011; harmine increases the amount of GFP expressed in each reactivated Bain et al., 2007; Göckler et al., 2009). Harmine’s primary target is cell (Fig. S2). This finding suggests that harmine works through a monoamine oxidase A (MAO-A) (Blum et al., 1964; Slotkin and mechanism that is unique from INDY. It also seems to indicate that DiStefano, 1970; TUNG et al., 1965; Yasuhara, 1974), but a well- the boosting effect of harmine is working through at least two reported role for inhibition of DYRK1A leading to enhanced NFAT biological pathways, one that increases the sensitivity of cells to activity has been reported (Adayev et al., 2011; Bain et al., 2007; activating stimuli (reduced dose of stimulus required to activate Egusa et al., 2011; Göckler et al., 2009). The crystal structure of HIV) and a second pathway that enhances the magnitude of HIV harmine complexed with DYRK1A has been solved, confirming its LTR activity in cells where reactivation occurs. binding to the ATP-binding pocket of DYRK1A (Ogawa et al., For a more direct measure of HIV-encoded genes/proteins, J-Lat 2010). Because of its ability to augment NFAT signaling, we cells were activated with ingenol A in the presence or absence of undertook a study of harmine and other DYRK1A inhibitors to harmine, or INDY. Here, like GFP, gag mRNA expression was determine whether these compounds could enhance HIV induced by ingenol A+DMSO and the expression was markedly reactivation alone or in combination with other known latency- increased with the addition of harmine (Fig. 1D). Harmine alone reversing agents (LRAs). We hypothesized that DYRK1A failed to induce gag expression. However, INDY alone did result in inhibitors would augment reactivation with LRAs through increased gag expression (Fig. 1D) even though INDY treatment increased NFAT availability. Here, we report that harmine and alone did not result in increased activation of J-Lat 5A8 cells another DYRK1A inhibitor, INDY (Ogawa et al., 2010), boost HIV (Fig. 1A). Next, western blot analysis was performed to measure reactivation by PKC agonists. Interestingly, the effect was gag protein expression. Gag protein expression was induced by independent of NFAT activity; harmine was effective at boosting ingenol A alone and markedly increased with the addition of LRAs even when no evidence of NFAT activity was detected. In harmine or INDY. The effect was stronger for harmine than with addition, CRISPR knockout of DYRK1A did not mimic the effects INDY (Fig. 1E). To better measure the amount of gag being of harmine or modulate the efficacy of harmine for boosting HIV expressed on a per cell basis, flow cytometry sorting was used to + − reactivation, Instead, we found that harmine enhances MAPK and separate GFP versus GFP cells after ingenol A treatment in the NFκB signalling, leading to increased HIV transcription. Using presence or absence of harmine or INDY. Gag protein expression + − whole-genome microarray we observed that harmine modulates was measured in the GFP and GFP cells by western blot. The expression of key pTEFb components. HEXIM1 expression is GFP cells that were treated with ingenol A+harmine expressed significantly downregulated by harmine whereas the cyclin CCNT2 more gag protein than GFP cells that were treated with ingenol was upregulated. We conclude that harmine regulates pTEFb A+INDY (Fig. 1F&G). Again, these findings indicate that harmine complex heterogeneity and establishes an environment that is more is mediating two distinct effects on HIV latency. One effect conducive to HIV reactivation. Combination treatments with increases the sensitivity of cells to activating LRAs whereas a harmine and LRAs may prove efficacious in vivo. second effect increases the expression level of HIV genes/proteins. This finding leads to two key questions: does harmine enhance HIV RESULTS reactivation through enhanced NFAT activity? And is DYRK1A the DYRK1A inhibitors enhance the efficacy of T-cell activating target of harmine in HIV latency models? PKC agonists PKC agonists have been previously shown to reactivate latent HIV Harmine boosts HIV reactivation independent of NFAT (Brogdon et al., 2016; Díaz et al., 2015; Jiang et al., 2015; Laird DYRK1A negatively regulates NFAT by hyperphosphorylation, et al., 2015; Martínez-Bonet et al., 2015; Perez et al., 2010; Reuse which excludes NFAT from the nucleus. Only through calcium- et al., 2009; Williams et al., 2004). We expected reactivation with dependent calmodulin activation is NFAT de-phosphorylated PKC agonists to be enhanced by combinatorial treatment with resulting in nuclear translocation and activation of NFAT DYRK1A inhibitors, harmine or INDY (Fig. S1A). To test this, we dependent genes such as IL-2. The HIV LTR contains at least two used J-Lat 5A8 reporter cells, which are latently infected with a full- NFAT binding motifs and since harmine is a known inhibitor of length provirus that expresses GFP in place of Nef as an indicator of DYRK1A, we wanted to determine if the anti-latency effects of LTR activity. J-Lat 5A8 cells were pretreated with harmine or INDY harmine are partly due to enhanced NFAT activity. and reactivated with the PKC agonist ingenol A. Both harmine and First, we wanted to determine whether harmine affects NFAT INDY failed to cause any measurable HIV reactivation when used activity in T cells. Jurkat T cells were transduced with lentiviral Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 1. DYRK1A inhibitors enhance the efficacy of PKC agonist ingenol A. (A) The percentage of GFP J-Lat 5A8 cells after reactivation with ingenol A in the presence of inhibitors (n=6), (B) representative flow cytometry plots of the mean fluorescence intensity of GFP cells, and (C) summary mean fluorescence intensity data (n=6). (D) gag mRNA expression determined by RT-qPCR. (E) Gag protein expression determined by western blot with + − representative blot and densitometry (n=3). GFP and GFP cells were sorted after reactivation with ingenol A. (F) Post-sort purity measured by flow − + + cytometry. The percentage of GFP population (black) and the GFP population (green). (G) Gag protein expression of post-sort GFP cells measured by western blot. Error bars represent standard deviation. Statistical analysis was performed by two-way ANOVA and corrected for multiple comparisons by Tukey’s test. **P<0.01; ***P<0.001; ****P<0.0001; ns, not significant. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 luciferase reporter constructs for NFAT, NFκB or a negative control MFI of GFP cells (Fig. 4C). Although DYRK1A knockout led to a virus with luciferase gene but no promoter. When reporter cells were decrease in the number of GFP cells, treatment with harmine and treated with ionomycin alone, a well-known activator of calmodulin INDY still resulted in an increase in GFP cells (Fig. 4B). This and NFAT, only the NFAT reporter cells expressed luciferase suggests that while DYRK1A expression may be involved in (Fig. 2A). When reporter cells were treated with ingenol A alone, reactivation of cells (Fig. 4B) it is not required and is not the only the NFκB cells expressed luciferase (Fig. 2A). This suggests pathway targeted by harmine and INDY that results in increased that ingenol A-mediated reactivation of J-Lat 5A8 cells is dependent HIV reactivation. on NFκB, not NFAT. Furthermore, treatment of J-Lat 5A8 cells with IKK16, an inhibitor of NFκB signaling, significantly reduced the Harmine downregulates HEXIM1 expression percentage of reactivated cells as well as the MFI of GFP cells in a To better understand how harmine is modulating HIV reactivation, dose-dependent manner (Fig. 2B). To determine if harmine was we used whole-genome microarray to determine which genes were influencing NFAT or NFκB activity, NFAT and NFκB reporter cells differentially expressed in cells treated with harmine. J-Lat cells were treated with ingenol A with harmine or INDY. Harmine had no were treated with DMSO, harmine, PMA, or harmine+PMA. effect on ionomycin-induced luciferase in the NFAT reporter cells. Microarray analysis showed that there were 35 transcripts that were However, harmine significantly boosted luciferase activity in the significantly upregulated or downregulated by at least twofold NFκB reporter cells compared to INDY. INDY had no effect on between PMA and PMA+harmine treatments. A complete list of the luciferase expression in any of the cell lines (Fig. 2C). These data 35 significantly upregulated or downregulated genes are presented suggest that the boosting effects of harmine and INDY are through in Table S1. Fourteen of these transcripts correspond to coding different mechanisms. genes (Fig. 5A and Table 1) and the remaining 21 transcripts were Second, we wanted to determine whether harmine or INDY non-coding transcripts. The gene that was upregulated the most in affected calcium flux upstream of NFAT activation. Jurkat cells PMA+harmine treatment compared to PMA treatment alone was were stained with a calcium-sensitive dye and treated with CCNT2, which codes for the cyclin T2 protein. The gene that was ionomycin or ingenol A. Ionomycin induced a calcium flux as downregulated the most in PMA+harmine treatment compared to expected but ingenol A did not. Neither harmine nor INDY had any PMA treatment alone was HEXIM1. Both genes code for proteins effect on calcium flux regardless of agonist treatment (Fig. 2D). that play a role in transcriptional elongation (Chen et al., 2018). The Harmine is primarily affecting the NFκB pathway with no HIV-encoded transcription factor Tat competes with HEXIM1 for significant involvement of NFAT or the calcium flux. Thus, binding to the pTEFb complex to promote HIV transcription harmine and INDY are working through different mechanisms. (Barboric et al., 2007). Downregulation of HEXIM1 would thus Harmine appears to be acting upon the NFκB pathway. result in less negative regulation of the pTEFb complex resulting in We next wanted to determine whether harmine was only affecting more Tat binding to pTEFb promoting transcript elongation. the NFκB pathway or if it was also affecting other pathways Similarly, upregulation of cyclin T2 would be expected to downstream of PKC. J-Lat cells cultured with PMA +/− harmine increase elongation of transcripts. This combination of effects on were analyzed for phospho-ERK1/2, phospho-MAPKp38, and transcription explains why harmine treatment not only increases the phospho-AKT levels. We found that harmine markedly boosted percentage of reactivated cells but also increases the number of viral phospho-ERK1/2 and phospho-MAPKp38 levels induced by PMA transcripts on a per-cell basis. (Fig. 3A). We found no difference in levels of phospho-AKT To confirm the microarray findings, we treated J-Lat cells with (Fig. 3A). Inclusion of an ERK inhibitor, U0126, resulted in dose- ingenol A in the presence or absence of harmine or INDY. Western blot dependent suppression of ingenol A-induced HIV reactivation analysis showed that harmine treatment alone or with ingenol A leads to (Fig. 3B). Importantly, while the inhibitor markedly reduced the a significant reduction in HEXIM1 protein expression, whereas INDY + + frequency of GFP cells, it also reduced the MFI of GFP cells treatment has no effect (Fig. 5B). The decrease in HEXIM1 protein (Fig. 3C). This reduction in the MFI of the GFP cells was more expression corresponded to an increase in the expression of gag pronounced than the reduction in MFI seen with the inhibition of protein. Thus, harmine treatment alone downregulates HEXIM1 and NFκB. Thus, it appears that harmine boosts the frequency of GFP boosts the efficacy of subsequent LRA stimulation. cells by enhancing sensitivity to PKC agonists as well as increases the magnitude of LTR activity independent of a strong activating Harmine boosts the efficacy of SAHA-induced HIV signal. reactivation We wanted to determine if harmine also enhanced other LRAs Harmine boosts HIV reactivation in the absence of DYRK1A independent of the PKC pathway. We observed that harmine also Since we observed that the boosting effect of harmine was boosts the efficacy of SAHA (vorinostat), an HDAC inhibitor, to independent of NFAT activity and was associated with altered enhance viral reactivation. Harmine only boosts the MFI of SAHA- MAPK signaling, we wanted to confirm that the increase in GFP reactivated cells and not the frequency of reactivated cells (Fig. S3) cells and the increase in the MFI of GFP cells by harmine was because SAHA reactivation is independent of NFκBor NFAT DYRK1A-dependent. To test this, a J-Lat cell line lacking pathways. We next tested the combinatorial effects of ingenol A, DYRK1A expression was derived by CRISPR/Cas9 gene SAHA, and harmine on HIV reactivation. SAHA and harmine both targeting (Fig. 4A). If the phenotype caused by harmine were due boosted the percentage of ingenol A-reactivated cells and that the to its interactions with DYRK1A, then we anticipated that J-Lat cells combination of SAHA and harmine had a more potent effect on lacking DYRK1A would behave similarly to harmine treated cells ingenol A-induced reactivation (Fig. 6A). To determine whether with increased GFP expression in response to activating LRAs. this effect was synergistic, we used the Bliss Independence Model DYRK1A knockout did lead to a significant decrease in the of drug synergy and found that the observed combined effects of percentage of GFP cells. However, this decrease was also seen in SAHA and harmine were significantly greater than those predicted knockout cells not treated with an inhibitor (DMSO control) by the model indicating that the combination of SAHA and harmine (Fig. 4B). Interestingly, DYRK1A knockout had no effect on the is synergistic (Fig. 3C). SAHA did not boost the MFI of ingenol A- Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 2. Harmine’s boosting effect is through NFκB, not NFAT. (A) Ionomycin or ingenol A were titrated on J-Lat 5A8 luciferase reporter cells. The dashed lines indicate the luciferase activity with no ionomycin or ingenol A treatment. (B) NFAT or NFκB luciferase reporter cells reactivated by ionomycin or ingenol A in the presence of inhibitors. Dashed lines represent luciferase activity with no ionomycin or ingenol A treatment. (C) J-Lat 5A8 cells were stained with Calcium Sensor Dye eFluor 514 (2 µM) and treated with DMSO, harmine, or INDY for 30 min. Geometric mean fluorescence was measured by flow cytometry for 240 s. Ionomycin (250 nM) or ingenol A (50 nM) was added after 75 s (arrows). Plots show mean geometric fluorescence intensity over time and the change in area under the curve (n=3). (D) J-Lat 5A8 cells were pretreated with DMSO or an IκB kinase inhibitor, IKK 16, at 1 µM (+) or 10 µM (++) followed by ingenol A treatment (31.25 nM) for 18 h (n=3). Error bars represent standard deviation. Statistical analysis was performed by one-way-ANOVA corrected for multiple comparisons with Dunnett’s test (D) and one-way-ANOVA of the area under the curve corrected for multiple comparisons by Tukey’s test (C). *P<0.05; ***P<0.001; ****P<0.0001; ns, not significant; RLU, relative light units. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 3. Harmine boosts phospho-ERK1/2 and phospho-p38 levels after ingenol A stimulation. J-Lat 5A8 cells were reactivated with ingenol A (31.25 nM) in the presence of inhibitors. (A) Whole cell lysates were analyzed by western blot for phospho-ERK1/2, phospho- AKT, and phospho-p38. (B) J-Lat 5A8 cells were pretreated with DMSO, a MEK inhibitor (U0126), or an IκB kinase inhibitor (IKK 16) for 30 min followed by overnight treatment with ingenol A (100 nM). GFP expression was assessed by flow cytometry. The percentage and (C) mean fluorescence intensity of the GFP cells are shown. Dashed lines represent treatment with ingenol A alone. reactivated GFP cells. However, harmine did boost the MFI of For the shock and kill approach to be effective all replication- ingenol A-reactivated GFP cells and the combination of harmine competent cells must be reactivated to prevent viral rebound after and SAHA had an even greater boosting effect (Fig. 6B). Similarly, removal of ART. One strategy being evaluated for improving the we found that while harmine+ingenol A increases gag mRNA shock and kill method is using combinations of drugs that act (Fig. 6D) and gag protein (Fig. 6E) expression compared to ingenol synergistically to reactivate a greater number of latently infected A alone that harmine+SAHA+ingenol A induced significantly more cells (Laird et al., 2015). In the current study, we demonstrate that gag expression. the plant-derived compound harmine used in combination with ingenol A increases the efficacy of latent HIV-1 reactivation. Our DISCUSSION data show that harmine has two effects: it increases the number of The primary goal of the ‘shock and kill’ strategy is to reactivate cells that are reactivated by ingenol A, and it increases transcription of HIV so that reservoir cells will be killed by the LTR promoter activity in reactivated cells resulting in increased cytopathic effects of the virus or expression of viral proteins will viral transcripts and viral proteins (Fig. 7A,B). enable recognition and killing by the immune system. LRAs In the J-Lat model, GFP acts as a surrogate measure for LTR targeting a diverse range of pathways have been proposed for use activity. Our data demonstrate that harmine treatment results in but most fall short of reactivating all replication-competent cells. increased GFP expression on a per-cell basis suggesting that viral Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 4. Harmine boosts independently of DYRK1A. DYRK1A was knocked out with CRISPR/Cas9 in J-Lat 5A8 cells. (A) Western blot of J-Lat 5A8 cells after treatment with CRISPR compared to Jurkat cells (Ctrl). The CRISPR knockout cells were reactivated with ingenol A (31.25 nM) in the presence of inhibitors. (B) The percentage of GFP cells and the (C) mean fluorescence Fig. 5. Harmine downregulates HEXIM1 expression. J-Lat 5A8 cells were intensity of GFP cells was measured by flow cytometry. Error bars treated pretreated with DMSO, harmine, or INDY for 30 min followed by represent standard deviation. Statistical analysis was performed by PMA stimulation for 2 hours and analyzed by whole transcriptome two-way ANOVA corrected for multiple comparisons with Bonferroni’s test. microarray. (A) Hierarchical clustering of transcripts that were upregulated or **p<0.01; ***p<0.001. downregulated at least twofold between PMA and harmine+PMA treatments. (B) Representative western blot analysis of HEXIM1 expression and densitometry (n=3). Error bars represent standard deviation. gene expression would also occur. Indeed, gag expression was Statistical analysis was performed by two-way ANOVA corrected for increased at both the transcript level and protein level as multiple comparisons with Tukey’s test. **P<0.01; ***P<0.001. demonstrated by RT-qPCR and western blot, respectively (Fig. 1). Increased expression of viral proteins will make the ‘kill’ phase of The combination of these three drugs promoted reactivation of a the ‘shock and kill’ approach more effective. Our data also greater number of cells and increased viral protein production in the demonstrate that the combination of harmine and SAHA has a reactivated cells (Fig. 6D,E). Interestingly, we found that harmine synergistic effect in boosting HIV reactivation by ingenol A (Fig. 6). also boosted the reactivation effects of SAHA alone (Fig. S3), Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Table 1. PMA-Induced Gene Expression with Harmine Martínez-Bonet et al., 2015; Perez et al., 2010; Reuse et al., 2009; Williams et al., 2004). This synergism is due to two different Fold-Change (PMA+harmine mechanisms for latency being targeted: availability of transcription Gene Symbol versus PMA alone) factors and either epigenetic modifications or availability of pTEFb CCNT2 2.2142 complexes. In our current study, we wanted to investigate whether SOX30 2.10846 the addition of the compound harmine would also act synergistically TMEM2 −2.0027 RAB23 −2.03377 with PKC agonists to reactivate latently infected cells. We predicted ANKRD50 −2.12436 harmine would boost reactivation by increasing availability of the IL27RA −2.1761 transcription factor NFAT, since harmine has been shown to inhibit RASA2 −2.3009 DYRK1A (Bain et al., 2007; Göckler et al., 2009), a negative SLC7A11 −2.37159 regulator of NFAT signaling (Adayev et al., 2011; Bain et al., 2007; LRRC8B −2.3876 Egusa et al., 2011; Göckler et al., 2009). However, in the current SKIL −2.4415 USP27X −2.44667 study, we demonstrate that the boosting effect is independent of CHAC1 −2.52968 NFAT (Fig. 2). HEXIM1 −3.29431 Since harmine is a bona fide DYRK1A inhibitor and has been shown to bind to DYRK1A in crystal structures, we hypothesized that harmine may be modifying the function of DYRK1A or which is likely due to harmine and SAHA working on different directing it into a different pathway. We tested if harmine treatment pathways. had any effects on signaling through the MEK/ERK and p38 MAPK Previous studies have demonstrated that PKC agonists such as pathways when used in combination with a PKC agonist. We found prostratin, bryostatin, and ingenol A work synergistically with that harmine treatment resulted in increased signaling through these HDAC inhibitors as well as the BRD4 inhibitor JQ1 (Brogdon et al., pathways in contrast to INDY. While both harmine and INDY 2016; Díaz et al., 2015; Jiang et al., 2015; Laird et al., 2015; boosted the frequency of reactivated cells after treatment with PKC Fig. 6. SAHA and harmine act synergistically to boost ingenol A activation of J-Lat cells. J-Lat 5A8 cells were pretreated with DMSO, harmine, SAHA, or SAHA+harmine for 30 min followed by ingenol A (31.25 nM) stimulation for 18 h. (A) The percentage of GFP cells and (B) the mean fluorescence intensity (n=3). (C) Synergy calculated using the Bliss Independnce Model. Expression of gag mRNA and gag protein were measured by (D) qPCR and (E) western blot, respectively (n=3). Statistical analysis was performed by calculating the area under the curve and performing a two-way ANOVA corrected for multiple comparisons with Tukey’s test (A,B) or by one-way ANOVA analysis corrected for multiple comparisons with Tukey’s test (D,E). Statistical analysis for (C) was performed by a ratio paired t-test comparing the predicted affected fraction fa to the observed affected fraction fa .*P<0.05; **P<0.01; HS,P HS,O ****P<0.0001; ns, not significant. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Fig. 7. Harmine treatment results in increased HIV reactivation by PKC-agonists. (A) Combinatorial treatment of ingenol A and harmine results in + + increased frequency of GFP cells and increased MFI in GFP cells in J-Lat model. (B) Combinatorial treatment of ingenol A and harmine results in increased LTR activity and expression of HIV RNA. (C) Harmine treatment in combination with PKC agonists results in increased availability of transcription factor NFκB, increased MAPK p38 and ERK1/2 activity, and decreased HEXIM1 expression. agonists, only harmine boosted the MFI of the reactivated cells. This Microarray data demonstrated that harmine treatment reduced suggested that harmine may have a secondary effect on DYRK1A. HEXIM1 levels. This was confirmed at the protein level by western However, when we knocked out DYRK1A using CRISPR/Cas9, the blot analysis (Fig. 5). HEXIM1 is a component of the pTEFb boosting effect of harmine was still observed (Fig. 4). This indicates complex. The pTEFb complex consists of multiple proteins that enhanced HIV reactivation seen with harmine treatment is not including CycT1, CDK9, and HEXIM1, and there are dependent on interaction with DYRK1A. heterogeneous combinations of these proteins that determine the Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 specific activity (or inactivity) of pTEFb. A 7SK snRNP complex Thus, the combination of ingenol A, SAHA, and harmine results containing HEXIM1, CDK9, CycT1, MePCE, and LARP7 along in increased reactivation of latently infected J-Lat cells and with a 7SK snRNA holds CDK9 in an inactive complex (Michels increased expression of viral proteins. This is likely due to et al., 2004; Nguyen et al., 2001; Yang et al., 2001; Yik et al., 2003). harmine treatment increasing the availability of NFκB, enhancing A key role for HIV Tat protein is to release CDK9/CycT1 pTEFb the effect of the PKC agonist, while simultaneously boosting the from 7SK snRNA so that it may drive HIV mRNA transcription activation of the MEK/ERK pathway that leads to increased (Fujinaga et al., 2004; Ivanov et al., 2000; Ping and Rana, 2001). transcript elongation. When used in combination with SAHA, Alternatively, CDK9/CycT1 can be bound by BRD4 and this drives which inhibits HDACs and increases the availability of the LTR expression of host genes such as c-Myc (Dey et al., 2009; Jang et al., promoter, the result is a potent drug combination that leads to 2005; Yang et al., 2005, 2008). However, BRD4 competes with Tat enhanced reactivation of latently infected cells. Our findings for pTEFb (Li et al., 2013). It has been shown that the BRD4 suggest that combination therapies that use currently existing inhibitor JQ1 enhances Tat-induced HIV transcription and reversal LRAs would be enhanced by the addition of harmine. Additional of latency (Banerjee et al., 2012; Bartholomeeusen et al., 2012). studies are needed to determine the molecular target of harmine that Thus, the pTEFb transcription complex acts as a three-position results in the phenotype we have reported here before harmine is switch relative to HIV and host gene mRNA expression. HEXIM1- used in vivo. containing complexes are in the off position for all pTEFb- dependent genes. BRD4-containing complexes are in the off MATERIALS AND METHODS position for HIV transcription but the on position for host genes Reagents and resources such as c-Myc. Finally, Tat-containing complexes are in the on A list of key antibodies, inhibitors, commercial kits, and other position for HIV transcription. reagents can be found in Table S2. Our microarray analysis revealed increased expression of CCNT2, the gene that codes for CyclinT2, and decreased Reactivation experiments expression of HEXIM1. While CyclinT1 is the kinase normally J-Lat 5A8 cells are treated with DMSO, harmine (20 µM), INDY associated with the pTEFb complex CyclinT2 has also been (20 µM), or SAHA (5 µM) for 30 min followed by treatment with reported to associate with CDK9 in the PTEF-b complex. Unlike agonists for 18 h unless otherwise indicated in figure legend. CyclinT1, which promotes Tat activity, CyclinT2 has been reported to be inhibitory for HIV-1 transcription (Napolitano et al., 1999). It Cell culture is unclear whether increased expression of CyclinT2 in our J-Lat J-Lat 5A8 cells, a kind gift from Warner Greene (University of model is having a negative effect on HIV transcription. HEXIM1, California, San Francisco, CA, USA), have been previously when part of the pTEFb complex, is also inhibitory for HIV-1 described (Chan et al., 2013). Briefly, J-Lat 5A8 cells are Jurkat replication (Barboric et al., 2007; Michels et al., 2004; Yik et al., cells that are latently infected with a full-length provirus integrated 2003). Therefore, decreased HEXIM1 expression could account for into the MAT2A gene and have the gfp reporter gene in place of nef.A at least some of the boosting effect of harmine. frameshift resulting in defective env production renders the cells non- Signaling through the TCR activates NFκB and NFAT and it has infectious. Any stimulus that activates the LTR will result in been previously reported that TCR signaling enhances transcription of gfp. J-Lat 5A8 cells and Jurkat luciferase reporter transcriptional elongation of latent HIV-1 by activating pTEFb cells were cultured in RPMI 1640 with 2 mM L-glutamine (Corning) through an ERK-dependent pathway (Kim et al., 2011). We found supplemented with 10% heat-inactivated FBS (Sigma-Aldrich, that treatment with ingenol A leads to ERK-1/2 phosphorylation St. Louis, MO, USA) and 100 U/mL penicillin-streptomycin. and that this effect was enhanced by treatment with harmine (Fig. 3), which may be the mechanism by which harmine treatment Primary CD4 T Cells enhances LTR promoter activity through pTEFb. Our data suggest Buffy coats were obtained from Life South Community Blood that harmine is increasing transcriptional elongation through Center (Gainesville, FL, USA) under approval by the Institutional increased availability of pTEFb which may act synergistically with Review Board at the University of Florida. PBMCs were processed the BRD4 inhibitor JQ1. as previously described (Taylor et al., 2018) and CD4 T cells were A recent study by Booiman and colleagues demonstrated that the enriched with the CD4 T Cell Isolation Kit (Miltenyi Biotec) DYRK1A inhibitor, INDY, reactivated latently infected J-Lat cells according to the manufacturer’s instructions. CD4 T cells were without the use of a PKC agonist (Booiman et al., 2015). In our cultured in Advanced RPMI 1640 (Corning) supplemented with hands, however, treatment of J-Lat cells with DYRK1A inhibitors 2 mM GlutaMAX (Gibco), 10% heat-inactivated FBS (Sigma- alone was not sufficient to reactivate latently infected cells as Aldrich), and 100 U/ml penicillin-streptomycin. measured by GFP expression. However, INDY treatment alone did result in increased gag mRNA expression (Fig. 1D). The dose that Flow cytometry and flow sorting we used for our study (20 µM) is much lower than the doses used by Cells were fixed in 2% paraformaldehyde and GFP fluorescence Booiman and colleagues. When higher doses of INDY were used in was measured using the BD Accuri™ C6 flow cytometer (BD our studies, excessive toxicity for J-Lat cells was observed (not Biosciences). Flow sorting was performed on unfixed cells the shown). Treatment with DYRK1A inhibitors only affected J-Lat FACS Aria III (BD Biosciences) at the University of Florida Center cells that were stimulated with PKC agonists, SAHA, or TNF. The for Immunology and Transplantation. study by Booiman and colleagues used J-Lat clones 8.4 and A1, whereas we used clone 5A8, which may account for the differences AlamarBlue assay (Booiman et al., 2015). We did not observe an increase of Gag Jurkat T cells were treated with different doses of harmine, INDY, or expression by western blot in cells treated with INDY alone to an equivalent volume of DMSO overnight. Toxicity of the inhibitors correlate with the increased in gag mRNA levels. This is likely due was measured by adding 10 µl of AlamarBlue Cell Viability Assay to the higher sensitivity of qPCR compared to western blot. Reagent (Thermo Scientific) to 90 µl of cells and waiting for color Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 development. Absorbance was measured at 570 and 600 nm. The CRISPR/Cas9 Knockout of DYRK1A percent reduction was calculated according to the manufacturer’s J-Lat 5A8 cells were co-transfected with two plasmids. One plasmid instructions. on the pCRISPR-CG01 backbone (GeneCopoeia) codes for recombinant Cas9 and a sgRNA (5′-GCCAAACATAAGTGACC- Western blot AAC-3′) that targets exon 2 of DYRK1A. The second plasmid on Western blots were performed as previously described (Taylor the pDONOR-D01 backbone (GeneCopoeia) has a mCherry-T2A- et al., 2018). Puro reporter cassette flanked by homology regions adjacent to the sgRNA target site in the genome. After co-transfection, the J-Lat RT-qPCR cells were selected with puromycin (1 µg/ml). Total RNA was isolated with the RNeasy Plus Mini Kit (Qiagen). cDNA synthesis was carried out with High-Capacity cDNA Reverse Drug synergy calculations Transcription Kit (Applied Biosystems, Foster City, CA, USA) Synergy between harmine and SAHA was calculated using the Bliss according to the manufacturer’s instructions. RT-qPCR reactions Independence Model as previously described (Laird et al., 2015). were carried out in SYBR™ Select Master Mix (Thermo Fisher The predicted fraction fa can be calculated using the equation HS,P Scientific, Waltham, MA, USA) according to the manufacturer’s fa =fa +fa −( fa fa ) fa =fraction of GFP cells reactivated HS,P H S H S H instructions. RT-qPCR reactions were performed on the by harmine with ingenol A fa =fraction of GFP cells reactivated by StepOnePlus™ Real-Time PCR System (Applied Biosystems). SAHA with ingenol A fa =fraction of GFP cells reactivated HS,O Gag primers used were forward: 5′- GAGCTAGAACGATTCGC- by harmine, SAHA, and ingenol A. To determine if the combination AGTTA-3′; reverse: 5′- CTGTCTGAAGGGATGGTTGTAG-3′. of harmine and SAHA is synergistic, the experimentally observed fraction of cells reactivated by harmine and SAHA ( fa ) can be HS,O Luciferase assays compared to the predicted fraction of reactivated GFP cells fa HS,P Reporter cell lines were created by transducing Jurkat (E6.1) T cells using the following equation: with Cignal Lenti NFAT reporter (Catalogue number CLS-015 L), Cignal Lenti NFκB reporter (Catalogue number CLS-013L), or D fa ¼ fa  fa : HS HS;O HS;P Cignal Lenti Negative Control (CLS-NCL) lentiviral particles purchased from Qiagen. An equal volume of Bright-Glo™ If ΔFa is greater than 1 then the combination is synergistic. If HS Luciferase Assay System (Promega) was added to an equal ΔFa is less than 1 then the interaction is antagonistic. If Δfa is HS HS volume of cells. After 5 min, the cells were lysed, and the lysate equal to 0 than the mechanisms are independent of each other. was transferred to a black Costar EIA/RIA polystyrene half area 96-well plate (Corning). Luminescence was measured with the Statistical analysis VICTOR™ X4 Multi-Plate Reader (PerkinElmer). All statistical analyses were performed using GraphPad Prism for Windows (GraphPad Software, La Jolla, CA, USA). A P-value of Microarrays less than 0.05 was considered statistically significant. J-Lat 5A8 cells were treated with DMSO or harmine (20 µM) for Acknowledgements 30 min. After pretreatment, the cells were treated with either The authors would like to thank Warner Greene (University of California, medium or PMA (10 nM) for 2 h. Total RNA was collected with the San Francisco CA, USA), for generously providing the J-Lat 5A8 cells used in the RNeasy Plus Mini Kit (Qiagen). Gene expression was assessed with study. GeneChip™ Human Transcriptome Array 2.0 arrays (Affymetrix) by the Interdisciplinary Center for Biotechnology Research at the Competing interests The authors declare no competing or financial interests. University of Florida. The analysis was performed with Partek Genomics Suite v. 6.6 (Partek Inc., St. Louis, MO, USA). CEL files Author contributions were imported and the raw data were subjected to multi-array average Conceptualization: J.P.T., M.A.W.; Methodology: J.P.T., M.A.W.; Validation: J.P.T., (RMA) background correction and quantile normalization. Probesets M.A.W.; Formal analysis: J.P.T., M.A.W.; Investigation: J.P.T., L.H.A., D.L.A., were summarized by the median polish method and the summarized M.N.C., M.A.W.; Resources: M.A.W.; Data curation: J.P.T.; Writing - original draft: J.P.T., M.A.W.; Writing - review & editing: J.P.T., M.A.W.; Visualization: J.P.T., signals were transformed to log base 2. A one-way ANOVA with M.A.W.; Supervision: M.A.W.; Project administration: M.A.W.; Funding acquisition: contrast was performed to determine fold changes between PMA and M.A.W. PMA+harmine-treated groups. Transcripts that were significantly (step-up FDR <0.05) upregulated or downregulated by at least twofold Funding are listed in Table S1 (coding and noncoding) and Table 1 (coding). This work was supported by funding from the National Institutes of Health grants R56AI108434 and R56AI122813. J.P.T. was supported by the Ruth L. Kirschstein National Research Service Award Institutional Research Training Grant Calcium flux assays T32AI007110. J-Lat 5A8 cells were incubated for 30 min with the Calcium Sensor Dye eFluor 514 (eBioscience) at a concentration of 2 µM with Data availability DMSO, harmine (20 µM), or INDY (20 µM). The cells were washed Microarray data are available in Gene Expression Omnibus (GEO) under accession GSE136172. and resuspended in sample buffer (1X PBS, 0.1% BSA, 2 mM EDTA) containing DMSO, harmine (20 µM), or INDY (20 µM). Supplementary information Flow cytometry data were collected for 1 min and then ionomycin Supplementary information available online at (250 nM) or ingenol A (50 nM) was added and flow cytometry data https://bio.biologists.org/lookup/doi/10.1242/bio.052969.supplemental were collected for an additional 3 min. The change in the area under References the curve (ΔAUC) was calculated by dividing the area under the Adayev, T., Wegiel, J. and Hwang, Y.-W. (2011). Harmine is an ATP-competitive curve after the addition of ionomycin or ingenol A divided by the inhibitor for dual-specificity tyrosine phosphorylation-regulated kinase 1A (Dyrk1A). Arch. Biochem. Biophys. 507, 212-218. doi:10.1016/j.abb.2010.12.024 AUC before the addition of ionomycin or ingenol A. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Archin, N. M., Espeseth, A., Parker, D., Cheema, M., Hazuda, D. and Margolis, Gö ckler, N., Jofre, G., Papadopoulos, C., Soppa, U., Tejedor, F. J. and Becker, W. D. M. (2009a). Expression of latent HIV induced by the potent HDAC inhibitor (2009). Harmine specifically inhibits protein kinase DYRK1A and interferes with suberoylanilide hydroxamic acid. AIDS Res. Hum. Retroviruses 25, 207-212. neurite formation. FEBS J. 276, 6324-6337. doi:10.1111/j.1742-4658.2009.07346.x doi:10.1089/aid.2008.0191 Gwack, Y., Sharma, S., Nardone, J., Tanasa, B., Iuga, A., Srikanth, S., Okamura, Archin, N. M., Keedy, K. S., Espeseth, A., Dang, H., Hazuda, D. J. and Margolis, H., Bolton, D., Feske, S., Hogan, P. G. et al. (2006). A genome-wide Drosophila D. M. (2009b). Expression of latent human immunodeficiency type 1 is induced by RNAi screen identifies DYRK-family kinases as regulators of NFAT. Nature 441, novel and selective histone deacetylase inhibitors. AIDS 23, 1799-1806. doi:10. 646-650. doi:10.1038/nature04631 1097/QAD.0b013e32832ec1dc Hamer, D. H. (2004). Can HIV be Cured? Mechanisms of HIV persistence and Archin, N. M., Liberty, A. L., Kashuba, A. D., Choudhary, S. K., Kuruc, J. D., strategies to combat it. Curr. HIV Res. 2, 99-111. doi:10.2174/1570162043484915 Crooks, A. M., Parker, D. C., Anderson, E. M., Kearney, M. F., Strain, M. C. Ivanov, D., Kwak, Y. T., Guo, J. and Gaynor, R. B. (2000). Domains in the SPT5 et al. (2012). Administration of vorinostat disrupts HIV-1 latency in patients on protein that modulate its transcriptional regulatory properties. Mol. Cell. Biol. 20, antiretroviral therapy. Nature 487, 482-485. doi:10.1038/nature11286 2970-2983. doi:10.1128/MCB.20.9.2970-2983.2000 Arron, J. R., Winslow, M. M., Polleri, A., Chang, C. P., Wu, H., Gao, X., Neilson, Jang, M. K., Mochizuki, K., Zhou, M., Jeong, H.-S., Brady, J. N. and Ozato, K. J. R., Chen, L., Heit, J. J., Kim, S. K. et al. (2006). NFAT dysregulation by (2005). The bromodomain protein Brd4 is a positive regulatory component of P- increased dosage of DSCR1 and DYRK1A on chromosome 21. Nature 441, TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19, 595-600. doi:10.1038/nature04678 523-534. doi:10.1016/j.molcel.2005.06.027 Bain, J., Plater, L., Elliott, M., Shpiro, N., Hastie, C. J., McLauchlan, H., Jiang, G., Mendes, E. A., Kaiser, P., Wong, D. P., Tang, Y., Cai, I., Fenton, A., Klevernic, I., Arthur, J. S. C., Alessi, D. R. and Cohen, P. (2007). The selectivity Melcher, G. P., Hildreth, J. E. K., Thompson, G. R. et al. (2015). Synergistic of protein kinase inhibitors: a further update. Biochem. J. 408, 297-315. doi:10. Reactivation of Latent HIV Expression by Ingenol-3-Angelate, PEP005, Targeted 1042/BJ20070797 NF-kB Signaling in Combination with JQ1 Induced p-TEFb Activation. PLoS Banerjee, C., Archin, N., Michaels, D., Belkina, A. C., Denis, G. V., Bradner, J., Pathog. 11, e1005066. doi:10.1371/journal.ppat.1005066 Sebastiani, P., Margolis, D. M. and Montano, M. (2012). BET bromodomain Kim, Y. K., Mbonye, U., Hokello, J. and Karn, J. (2011). T-Cell receptor signaling inhibition as a novel strategy for reactivation of HIV-1. J. Leukoc. Biol. 92, enhances transcriptional elongation from latent HIV proviruses by activating 1147-1154. doi:10.1189/jlb.0312165 P-TEFb through an ERK-dependent pathway. J. Mol. Biol. 410, 896-916. doi:10. Barboric, M., Yik, J. H. N., Czudnochowski, N., Yang, Z., Chen, R., Contreras, X., 1016/j.jmb.2011.03.054 Geyer, M., Matija Peterlin, B. and Zhou, Q. (2007). Tat competes with HEXIM1 Laird, G. M., Bullen, C. K., Rosenbloom, D. I. S., Martin, A. R., Hill, A. L., Durand, to increase the active pool of P-TEFb for HIV-1 transcription. Nucleic Acids Res. C. M., Siliciano, J. D. and Siliciano, R. F. (2015). Ex vivo analysis identifies 35, 2003-2012. doi:10.1093/nar/gkm063 effective HIV-1 latency-reversing drug combinations. J. Clin. Invest. 125, Bartholomeeusen, K., Xiang, Y., Fujinaga, K. and Peterlin, B. M. (2012). 1901-1912. doi:10.1172/JCI80142 Bromodomain and extra-terminal (BET) bromodomain inhibition activate Lehrman, G., Hogue, I. B., Palmer, S., Jennings, C., Spina, C. A., Wiegand, A., transcription via transient release of positive transcription elongation factor b Landay, A. L., Coombs, R. W., Richman, D. D., Mellors, J. W. et al. (2005). (P-TEFb) from 7SK small nuclear ribonucleoprotein. J. Biol. Chem. 287, Depletion of latent HIV-1 infection in vivo: a proof-of-concept study. Lancet 36609-36616. doi:10.1074/jbc.M112.410746 (London, England) 366, 549-555. doi:10.1016/S0140-6736(05)67098-5 Blum, B., Weizmann, H., Simpson, G. M., Krasilowsky, D., Kulcsar, I. S. and Li, Z., Guo, J., Wu, Y. and Zhou, Q. (2013). The BET bromodomain inhibitor JQ1 Merskey, H. (1964). Harmine antagonism of drug-induced extra-pyramidal activates HIV latency through antagonizing Brd4 inhibition of Tat-transactivation. disturbances. Psychopharmacologia 6, 307-310. doi:10.1007/BF00413162 Nucleic Acids Res. 41, 277-287. doi:10.1093/nar/gks976 Booiman, T., Loukachov, V. V., van Dort, K. A., van ‘t Wout, A. B. and Kootstra, Macian, F. (2005). NFAT proteins: key regulators of T-cell development and N. A. (2015). DYRK1A Controls HIV-1 Replication at a Transcriptional Level in an function. Nat. Rev. Immunol. 5, 472-484. doi:10.1038/nri1632 NFAT Dependent Manner. PLoS ONE 10, e0144229. doi:10.1371/journal.pone. Mancebo, H. S. Y., Lee, G., Flygare, J., Tomassini, J., Luu, P., Zhu, Y., Peng, J., Blau, C., Hazuda, D., Price, D. et al. (1997). P-TEFb kinase is required for HIV Brogdon, J., Ziani, W., Wang, X., Veazey, R. S. and Xu, H. (2016). In vitro effects of Tat transcriptional activation in vivo and in vitro. Genes Dev. 11, 2633-2644. the small-molecule protein kinase C agonists on HIV latency reactivation. Sci. doi:10.1101/gad.11.20.2633 Rep. 6, 39032. doi:10.1038/srep39032 Martınez-Bonet, ́ M., Clemente, M. I., Serramıa,́ M. J., Munoz, E., Moreno, S. and Chan, J. K., Bhattacharyya, D., Lassen, K. G., Ruelas, D. and Greene, W. C. Munoz-Fernández, M. Á. (2015). Synergistic Activation of Latent HIV-1 (2013). Calcium/calcineurin synergizes with prostratin to promote NF-κB Expression by Novel Histone Deacetylase Inhibitors and Bryostatin-1. Sci. Rep. dependent activation of latent HIV. PLoS ONE 8, e77749. doi:10.1371/journal. 5, 16445. doi:10.1038/srep16445 pone.0077749 Michels, A. A., Fraldi, A., Li, Q., Adamson, T. E., Bonnet, F., Nguyen, V. T., Chen, F. X., Smith, E. R. and Shilatifard, A. (2018). Born to run: Control of Sedore, S. C., Price, J. P., Price, D. H., Lania, L. et al. (2004). Binding of the 7SK transcription elongation by RNA polymerase II. Nat. Rev. Mol. Cell Biol. 19, snRNA turns the HEXIM1 protein into a P-TEFb (CDK9/cyclin T) inhibitor. EMBO 464-478. doi:10.1038/s41580-018-0010-5 J. 23, 2608-2619. doi:10.1038/sj.emboj.7600275 Contreras, X., Schweneker, M., Chen, C.-S., McCune, J. M., Deeks, S. G., Martin, Napolitano, G., Licciardo, P., Gallo, P., Majello, B., Giordano, A. and Lania, L. J. and Peterlin, B. M. (2009). Suberoylanilide hydroxamic acid reactivates HIV (1999). The CDK9-associated cyclins T1 and T2 exert opposite effects on HIV-1 from latently infected cells. J. Biol. Chem. 284, 6782-6789. doi:10.1074/jbc. Tat activity. AIDS 13, 1453-1459. doi:10.1097/00002030-199908200-00003 M807898200 Nguyen, V. T., Kiss, T., Michels, A. A. and Bensaude, O. (2001). 7SK small Davey, R. T., Bhat, N., Yoder, C., Chun, T. W., Metcalf, J. A., Dewar, R., Natarajan, nuclear RNA binds to and inhibits the activity of CDK9/cyclin T complexes. Nature V., Lempicki, R. A., Adelsberger, J. W., Miller, K. D. et al. (1999). HIV-1 and T 414, 322-325. doi:10.1038/35104581 cell dynamics after interruption of highly active antiretroviral therapy (HAART) in Ogawa, Y., Nonaka, Y., Goto, T., Ohnishi, E., Hiramatsu, T., Kii, I., Yoshida, M., patients with a history of sustained viral suppression. Proc. Natl. Acad. Sci. USA Ikura, T., Onogi, H., Shibuya, H. et al. (2010). Development of a novel selective 96, 15109-15114. doi:10.1073/pnas.96.26.15109 inhibitor of the Down syndrome-related kinase Dyrk1A. Nat. Commun. 1, 86. Deeks, S. G. (2012). HIV: Shock and kill. Nature 487, 439-440. doi:10.1038/ doi:10.1038/ncomms1090 487439a Patel, K., Gadewar, M., Tripathi, R., Prasad, S. K. and Patel, D. K. (2012). A review Dey, A., Nishiyama, A., Karpova, T., McNally, J. and Ozato, K. (2009). Brd4 marks on medicinal importance, pharmacological activity and bioanalytical aspects of select genes on mitotic chromatin and directs postmitotic transcription. Mol. Biol. beta-carboline alkaloid “‘Harmine’”. Asian Pac. J. Trop. Biomed. 2, 660-664. Cell 20, 4899-4909. doi:10.1091/mbc.e09-05-0380 doi:10.1016/S2221-1691(12)60116-6 Dıaz, ́ L., Martınez-Bonet, ́ M., Sánchez, J., Fernández-Pineda, A., Jiménez, J. L., Perez, M., de Vinuesa, A., Sanchez-Duffhues, G., Marquez, N., Bellido, M., Munoz, E., Moreno, S., Álvarez, S. and Munoz-Fernández, M. Á. (2015). ̃ ̃ Munoz-Fernandez, M., Moreno, S., Castor, T., Calzado, M. and Munoz, E. Bryostatin activates HIV-1 latent expression in human astrocytes through a PKC (2010). Bryostatin-1 synergizes with histone deacetylase inhibitors to reactivate and NF-κB-dependent mechanism. Sci. Rep. 5, 12442. doi:10.1038/srep12442 HIV-1 from latency. Curr. HIV Res. 8, 418-429. doi:10.2174/157016210793499312 Egusa, H., Doi, M., Saeki, M., Fukuyasu, S., Akashi, Y., Yokota, Y., Yatani, H. and Ping, Y. H. and Rana, T. M. (2001). DSIF and NELF interact with RNA polymerase II Kamisaki, Y. (2011). The small molecule harmine regulates NFATc1 and Id2 elongation complex and HIV-1 Tat stimulates P-TEFb-mediated phosphorylation expression in osteoclast progenitor cells. Bone 49, 264-274. doi:10.1016/j.bone. 2011.04.003 of RNA polymerase II and DSIF during transcription elongation. J. Biol. Chem. 276, 12951-12958. doi:10.1074/jbc.M006130200 Folks, T., Powell, D. M., Lightfoote, M. M., Benn, S., Martin, M. A. and Fauci, A. S. Reuse, S., Calao, M., Kabeya, K., Guiguen, A., Gatot, J.-S., Quivy, V., Vanhulle, (1986). Induction of HTLV-III/LAV from a nonvirus-producing T-cell line: implications for latency. Science 231, 600-602. doi:10.1126/science.3003906 C., Lamine, A., Vaira, D., Demonte, D. et al. (2009). Synergistic activation of HIV- Fujinaga, K., Irwin, D., Huang, Y., Taube, R., Kurosu, T. and Peterlin, B. M. (2004). 1 expression by deacetylase inhibitors and prostratin: implications for treatment of Dynamics of human immunodeficiency virus transcription: P-TEFb phosphorylates latent infection. PLoS ONE 4, e6093. doi:10.1371/journal.pone.0006093 RD and dissociates negative effectors from the transactivation response element. Siliciano, R. F. and Greene, W. C. (2011). HIV latency. Cold Spring Harb. Perspect. Mol. Cell. Biol. 24, 787-795. doi:10.1128/MCB.24.2.787-795.2004 Med. 1, a007096. doi:10.1101/cshperspect.a007096 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio052969. doi:10.1242/bio.052969 Slotkin, T. and DiStefano, V. (1970). Urinary metabolities of harmine in the rat and Yang, Z., Yik, J. H. N., Chen, R., He, N., Jang, M. K., Ozato, K. and Zhou, Q. (2005). their inhibition of monoamine oxidase. Biochem. Pharmacol. 19, 125-131. doi:10. Recruitment of P-TEFb for stimulation of transcriptional elongation by the 1016/0006-2952(70)90333-3 bromodomain protein Brd4. Mol. Cell 19, 535-545. doi:10.1016/j.molcel.2005.06.029 Taylor, J. P., Cash, M. N., Santostefano, K. E., Nakanishi, M., Terada, N. and Yang, Z., He, N. and Zhou, Q. (2008). Brd4 recruits P-TEFb to chromosomes at late Wallet, M. A. (2018). CRISPR/Cas9 knockout of USP18 enhances type I IFN mitosis to promote G1 gene expression and cell cycle progression. Mol. Cell. Biol. responsiveness and restricts HIV-1 infection in macrophages. J. Leukoc. Biol. 28, 967-976. doi:10.1128/MCB.01020-07 103, 1225-1240. doi:10.1002/JLB.3MIA0917-352R Yasuhara, H. (1974). Studies on monoamine oxidase (report XXIV). Effect of TUNG, Y. C., HWANG, Y. S. and WU, C. Y. (1965). Studies on beta- harmine on monoamine oxidase. Jpn. J. Pharmacol. 24, 523-533. doi:10.1254/jjp. carbolines. 3. Effect of harman hydrochloride on monoamine oxidase and the 24.523 uterine contractility induced by n-monomethyltryptamine hydrochloride. Tsa. Yik, J. H. N., Chen, R., Nishimura, R., Jennings, J. L., Link, A. J. and Zhou, Q. Chih. Gaoxiong Yi Xue Yuan. Tong Xue Hui 64, 44-50. (2003). Inhibition of P-TEFb (CDK9/Cyclin T) kinase and RNA polymerase II Williams, S. A., Chen, L.-F., Kwon, H., Fenard, D., Bisgrove, D., Verdin, E. and transcription by the coordinated actions of HEXIM1 and 7SK snRNA. Mol. Cell 12, Greene, W. C. (2004). Prostratin antagonizes HIV latency by activating NF- 971-982. doi:10.1016/S1097-2765(03)00388-5 kappaB. J. Biol. Chem. 279, 42008-42017. doi:10.1074/jbc.M402124200 Zhu, Y., Pe’ery, T., Peng, J., Ramanathan, Y., Marshall, N., Marshall, T., Amendt, Yang, Z., Zhu, Q., Luo, K. and Zhou, Q. (2001). The 7SK small nuclear RNA B., Mathews, M. B. and Price, D. H. (1997). Transcription elongation factor P- inhibits the CDK9/cyclin T1 kinase to control transcription. Nature 414, 317-322. TEFb is required for HIV-1 tat transactivation in vitro. Genes Dev. 11, 2622-2632. doi:10.1038/35104575 doi:10.1101/gad.11.20.2622 Biology Open

Journal

Biology Open – The Company of Biologists

Published: Dec 15, 2020

There are no references for this article.