ABSTRACT
HIV-1-infected people who take drugs that suppress viremia to undetectable levels are protected from developing AIDS. Nonetheless, these individuals have chronic inflammation associated with heightened risk of cardiovascular pathology. HIV-1 establishes proviruses in long-lived CD4+ memory T cells, and perhaps other cell types, that preclude elimination of the virus even after years of continuous antiviral therapy. Though the majority of proviruses that persist during antiviral therapy are defective for production of infectious virions, many are expressed, raising the possibility that the HIV-1provirus or its transcripts contribute to ongoing inflammation. Here we found that the HIV-1 provirus activated innate immune signaling in isolated dendritic cells, macrophages, and CD4+ T cells. Immune activation required transcription from the HIV-1 provirus and expression of CRM1-dependent, Rev-dependent, RRE-containing, unspliced HIV-1 RNA. If rev was provided in trans, all HIV-1 coding sequences were dispensable for activation except those cis-acting sequences required for replication or splicing. These results indicate that the complex, post-transcriptional regulation intrinsic to HIV-1 RNA is detected by the innate immune system as a danger signal, and that drugs which disrupt HIV-1 transcription or HIV-1 RNA metabolism would add qualitative benefit to current antiviral drug regimens.
INTRODUCTION
Drugs that block activity of the three HIV-1 enzymes-reverse transcriptase, integrase, and protease-potently suppress HIV-1 viremia and protect infected people from developing AIDS (Günthard et al., 2016). Despite effective antiviral therapy, many patients have systemic inflammation associated with increased risk of cardiovascular pathology (Brenchley et al., 2006; Freiberg et al., 2013; Sinha et al., 2016). Plausible explanations for this inflammation include ongoing T cell dysfunction, disruption of intestinal epithelium integrity, lymphatic tissue fibrosis, antiviral drug toxicity, and co-morbid infections such as cytomegalovirus (Brenchley et al., 2006; Hunt et al., 2011; Sinha et al., 2016).
Inflammation may also be maintained by HIV-1 itself. HIV-1 is not eliminated from the body after years of continuous suppressive therapy and viremia inevitably rebounds upon drug cessation (Davey et al., 1999). This is because HIV-1 establishes a provirus in long-lived CD4+ memory T cells and perhaps other cell types that include tissue-resident myeloid cells (Jiang et al., 2015; Kandathil et al., 2016; Siliciano et al., 2003). Proviruses are obligate replication intermediates that result from the integration of retroviral cDNA into chromosomal DNA to become permanent, heritable genetic elements in infected cells (Engelman and Singh, 2018). Though the vast majority of proviruses that persist in the presence of antiviral therapy are defective for the production of infectious virus (Bruner et al., 2016), 2-18% express HIV-1 RNA (Wiegand et al., 2017). Here we assessed HIV-1 proviruses and their gene products for the ability to contribute to chronic inflammation.
RESULTS
To determine if HIV-1 proviruses activate innate immune signaling, human blood cells were transduced with single-cycle vectors, either a full-length, single-cycle HIV-1 clone with a frameshift in env and eGFP in place of nef (HIV-1-GFP) (He et al., 1997), or a minimal 3-part lentivector encoding GFP (Fig. 1a and Table 1) (Pertel et al., 2011). Monocyte derived dendritic cells (DCs) were challenged initially since HIV-1 transduction of these specialized antigen-presenting cells activates innate immune signaling (Berg et al., 2012; Gao et al., 2013; Landau, 2014; Manel et al., 2010; Rasaiyaah et al., 2013). To increase the efficiency of provirus establishment, vectors were pseudotyped with the vesicular stomatitis virus glycoprotein (VSV G) and delivered concurrently with virus-like particles (VLPs) bearing SIVMAC251 Vpx (Fig. 1b) (Goujon et al., 2006; Pertel et al., 2011). Transduction efficiency, as determined by flow cytometry for GFP-positive cells, was 30-60% (Fig. 1b), depending on the blood donor.
DCs matured in response to HIV-1-GFP transduction, as indicated by increased mean fluorescence intensity of co-stimulatory or activation molecules, including HLA-DR, CD80, CD86, CD40, CD83, CCR7, CD141, ISG15, MX1, and IFIT (Sousa, 2006) (Fig. 1b, c). Maturation was evident among both GFP positive and negative cells (Fig. 1b), the latter resulting from activation in trans by type 1 IFN as several others have shown (Manel et al., 2010; Rasaiyaah et al., 2013). Identical results were obtained with full-length, single-cycle vectors generated from primary, transmitted/founder clones that were derived by single genome sequencing, HIV-1AD17 (Parrish et al., 2013), HIV-1Z331M- TF (Deymier et al., 2015), and HIV-1ZM249M (Salazar-Gonzalez et al., 2009), the first virus being clade B, the other two clade C (Fig. 1d). A single cycle HIV-2 vector also induced maturation, indicating that this innate response was not unique to HIV-1 (Fig. 1e).
DCs matured when HIV-1-GFP transduction efficiency was augmented with nucleosides (Reinhard et al., 2014), rather than with SIV VLPs, indicating that Vpx was not required for maturation (Fig. 1f). DCs were then challenged with replication-competent HIV-1 bearing CCR5-tropic Env, either T cell-tropic or macrophage-tropic (Granelli-Piperno et al., 1998), with or without Vpx-VLPs (Fig. 1g). The percent of cells transduced by vector bearing either Env increased with Vpx, though DC maturation was observed under all conditions, even among the very few DCs transduced by T cell-tropic env (see inset of Fig. 1g). These results indicate that neither VSV G, nor Vpx, nor high-titer infection, was required for DC maturation.
In response to transduction with HIV-1-GFP, steady-state CXCL10, IFNB1, and IL15 mRNAs reached maximum levels at 48 hrs, increasing 31,000-, 92-, and 140-fold relative to mock-treated cells, respectively (Fig. 1h, i). Correspondingly, IFNα2, CCL7, IL-6, CXCL10, and TNFα proteins accumulated in the supernatant (Fig. 1j). In contrast to the results with HIV-1-GFP, there were no signs of maturation after transduction with the 3-part minimal lentivector (Fig. 1b, h, j).
To determine if early stages in the HIV-1 replication cycle were necessary for maturation, reverse transcription was inhibited by nevirapine (NVP) or the HIV-1 RT mutant D185K/D186L, and integration was inhibited with raltegravir or the HIV-1 IN mutant D116A (Tables 1 and 2), as previously described (De Iaco and Luban, 2011). Each of these four conditions abrogated maturation, as indicated by cell surface CD86 (Fig. 2a) and steady-state CXCL10 mRNA (Fig. 2b). When integration was inhibited, CXCL10 mRNA increased in response to challenge with HIV-1-GFP, but levels were nearly 1,000 times lower than when integration was not blocked (Fig. 2b).
HIV-1 virion RNA and newly synthesized viral cDNA are reported to be detected by RIG-I and by cGAS, respectively (Berg et al., 2012; Gao et al., 2013). Signal transduction downstream of both sensors requires TBK1 and IRF3. The TBK1 inhibitor BX795 (Table 2) blocked DC maturation in response to cGAMP but had no effect on maturation after HIV-1-GFP transduction (Fig. 2c). Moreover, IRF3 knockdown (Table 1) (Pertel et al., 2011) suppressed activation of CD86 or ISG15 in response to cGAMP, but not in response to HIV-1 transduction (Fig. 2c). Similarly, no effect on HIV-1-induced DC maturation was observed with knockdown of IRF1, 5, 7, or 9, or of STAT1 or 2, or TAK1 (Table 1) (Pertel et al., 2011), or of pharmacologic inhibition of CypA, PKR, c-Raf, IkBa, NF-kB, MEK1+2, p38, JNK, Caspase 1, pan-Caspases, ASK1, eIF2a, TBK1, IKKe, TAK1, or NLRP3 (Table 2). Under the conditions used here, then, DC maturation required reverse transcription and integration but was independent of most well-characterized innate immune signaling pathways.
Completion of the HIV-1 integration reaction requires cellular DNA repair enzymes (Craigie and Bushman, 2012). That DCs did not mature in response to transduction with minimal lentivectors (Fig. 1b, i, j) indicates that activation of the DNA repair process is not sufficient, and that transcription from the HIV-1-GFP provirus must be necessary for maturation. Indeed, gag expression from an integrated vector has been reported to be necessary for DC maturation (Manel et al., 2010). To determine if any individual HIV-1 proteins were sufficient to mature DCs, a minimal lentivector was used to express codon optimized versions of each of the open reading frames possessed by HIV-1-GFP (Fig. 2d, Table 1). Among these vectors was a gag-expression vector that produced as much p24 protein as did HIV-1-GFP (Fig. 2d). None of these vectors matured DCs (Fig. 2d).
HIV-1-GFP was then mutated to determine if any protein coding sequences were necessary for DC maturation. For these and any subsequent experiments in which an essential viral component was disrupted within HIV-1-GFP, the factor in question was provided in trans, either during assembly in transfected HEK293 cells, or within transduced DCs, as appropriate (see Methods). Mutations that disrupted both gag and pol, either a double frameshift in gag, or a mutant in which the first 14 ATGs in gag were mutated, abolished synthesis of CA (p24) yet retained full maturation activity (Fig. 2e, Table 1). Deletion mutations encompassing gag/pol, vif/vpr, vpu/env, or nef/U3-LTR, each designed so as to leave cis-acting RNA elements intact, all matured DCs (Fig. 2f and Table 1). These results indicate that these HIV-1-GFP RNA sequences, as well as the proteins that they encode, were not required for DC maturation.
Tat and Rev coding sequences were individually disrupted by combining start codon point mutations with nonsense codons that were silent with respect to overlapping reading frames (Table 1). Neither Δtat nor Δrev matured DCs upon transduction (Fig. 2g). However, DCs matured upon co-transduction of Δtat and Δrev, or when minimal lentivectors expressing codon-optimized Tat and Rev were co-transduced in trans (Fig. 2g). These results indicate that the maturation defect with the individual vectors was due to disruption of Tat and Rev function, and not due to a cis-acting defect of the mutant RNA.
The minimal 3-part lentivector expressed GFP from a heterologous promoter and had a deletion mutation encompassing the essential, cis-acting TATA box and enhancer elements (Zufferey et al., 1998), as well as in the trans-acting tat and rev, that inactivated the promoter in the proviral 5’ LTR (Fig. 1a). To test the importance of LTR-driven transcription for DC maturation by the HIV-1 provirus, the HIV-1 LTR was restored in the minimal vector (Fig. 2h and Table 1); in addition, GFP was inserted in place of gag as a marker for LTR expression, and the heterologous promoter was used to drive tat, rev, or both genes separated by P2A coding sequence (Fig. 2h). None of the LTR-driven, minimal vectors matured DCs (Fig. 2h).
To determine if tat was necessary for DC maturation, tat and TAR were mutated in HIV-1-GFP and the LTR promoter was modified to be tetracycline-inducible, as previously described (Das and Berkhout, 2016) (Tet-HIV-1 in Fig. 3a and Table 1). The doxycycline-dependent reverse transactivator (rtTA) was delivered in trans by lentivector. In the presence of doxycycline (Table 2), Tet-HIV-1 and rtTA matured DCs when given in combination, but neither vector matured DCs when given in isolation (Fig. 3a). Additionally, the magnitude of cell surface CD86 was dependent on the doxycycline concentration, indicating that maturation was dependent on the level of HIV-1 transcription (Fig. 3b). These results demonstrated that tat was not required for maturation, so long as the provirus was expressed.
To ascertain whether rev was necessary for DC maturation, the RTE from a murine intracisternal A-particle retroelement (IAP), and the CTE from SRV-1, were inserted in place of nef (HIV-RTE/CTE in Fig. 3c and Table 1) (Smulevitch et al., 2006). Each of these elements utilizes the NXF1 nuclear RNA export pathway, thereby bypassing the need for CRM1 and rev (Fornerod et al., 1997). p24 levels with this construct were similar to those of HIV-1-GFP, indicating that unspliced RNA was exported from the nucleus at least as well as with Rev (Fig. 3c). Nonetheless, the HIV-RTE/CTE vector did not mature DCs (Fig. 3c), indicating that maturation was dependent upon rev and CRM1-mediated RNA export. Consistent with this conclusion, the CRM1 inhibitor leptomycin B (Table 2) abrogated DC maturation by HIV-1-GFP (Fig. 3d). In contrast, leptomycin B had no effect on DC maturation in response to Sendai virus infection (Fig. 3e). ISG15 was used to monitor maturation in these experiments since, as previously reported for DCs, leptomycin B altered background levels of CD86 (Chemnitz et al., 2010).
To determine if innate immune detection of HIV-1 was unique to DCs, monocyte-derived macrophages and CD4+ T cells were examined. In response to transduction with HIV-1-GFP, macrophages upregulated CD86, ISG15, and HLA-DR, and CD4+ T cells upregulated MX1, IFIT1, and HLA-DR (Fig. 4a). DCs, macrophages, and CD4+ T cells were then transduced side-by-side with mutant constructs to determine if the mechanism of innate immune activation was similar to that in DCs. As with DCs, HIV-1-GFP bearing the Δgag/pol deletion activated macrophages and CD4+ T cells (Fig. 4b). Also in agreement with the DC results, neither the minimal lentivector, nor HIV-1-GFP bearing mutations in integrase, tat, or rev, matured any of the three cell types (Fig. 4b). CD4+ T cells were infected with either macrophage-tropic or T cell-tropic HIV-1 to determine whether replication-competent HIV-1 was similarly capable of innate immune activation in these cells, in the absence of VSV G. As with DCs, innate immune activation, as detected by MX1 and ISG15 upregulation, was observed in cells productively infected with HIV-1, but not with minimal lentivector (Fig. 4c). Finally, to test the effect of HIV-1 proviral RNA on non-activated T cells, CD4+ T cells were co-transduced with Tet-HIV-1 and the rtTA3 vector, and cultured for 9 days in the absence of stimulation. Upon doxycycline treatment, T cells expressed GFP and MX1 (Fig. 4d). As in DCs, dose-dependent activation was observed with doxycycline (Fig. 4d). These data indicate that innate immune activation by HIV-1, in all three cell types, requires integration, transcription, and Rev-dependent, HIV-1 intron-containing RNA.
DISCUSSION
The HIV-1 LTR generates a single primary transcript that gives rise to over 100 alternatively spliced RNAs (Ocwieja et al., 2012). The full-length, unspliced, intron-bearing transcript acts as viral genomic RNA in the virion and as mRNA for essential gag- and pol-encoded proteins. Expression of the unspliced transcript requires specialized viral and cellular machinery, HIV-1 Rev and CRM1 (Fornerod et al., 1997), in order to escape from the spliceosome. Results here indicate that unspliced or partially spliced HIV-1 RNA is detected by human cells as a danger signal, as has been reported for inefficiently spliced mRNAs from transposable elements in distantly related eukaryotes (Dumesic et al., 2013). Transposable elements are mutagenic to the host genome and it stands to reason that molecular features such as transcripts bearing multiple, inefficient splice signals characteristic of retrotransposons, would activate innate immune signaling pathways.
HIV-1 genomic RNA contains extensive secondary and higher order structures that could be detected by innate immune sensors. Our knockdown of IRF3 and inhibition of TBK1, both required for signal transduction of the RNA sensors RIG-I, MDA5, and TLR3, did not impede HIV-1 maturation of DCs (Figure 2c). Furthermore, we suppressed an extensive list of innate signaling pathways and sensors including knockdowns of IRF’s 1, 5, 7, and 9, STAT’s 1 and 2, or of TAK1, as well as pharmacologic inhibition of CypA, CRM1, PKR, c-Raf, IkBa, NF-kB, MEK1+2, p38, JNK, Caspase 1, pan-Caspases, ASK1, eIF2a, IKKe, TAK1, or NLRP3 (Table 2). None of these perturbations had any effect on limiting innate immune activation by HIV-1, suggesting requirement for an alternative detection mechanism. Such mechanisms might include uncharacterized RNA sensors, direct detection of stalled splicing machinery, or overload of the CRM1 nuclear export pathway itself.
The replication competent HIV-1 reservoir in memory CD4+ T cells has a 44 wk half-life and thus patients must take antiviral medication for life (Crooks et al., 2015). Long-lived, replication competent HIV-1 reservoirs in other cell types have not been clearly demonstrated, but these may also contribute to the HIV-1 reservoir (Kandathil et al., 2016). The common genetic determinants in HIV-1 for maturation of CD4+ T cells, macrophages, and DCs suggests that HIV-1 is detected by a mechanism that is conserved across cell types, and that this mechanism would be active in any cell type that possesses a transcriptionally active provirus. Data here show that proviruses need not be replication competent to contribute to inflammation. Rather, HIV-1 transcription and export of unspliced RNA, regardless of replication competence, is sufficient to induce immune activation. Consistent with our findings, T cell activation correlates directly with the level of cell-associated HIV-1 RNA in patients receiving antiretroviral therapy (El-Diwany et al., 2017). Furthermore, our data suggests that new drugs that block HIV-1 transcription, Tat-mediated transcriptional elongation, or Rev-mediated preservation of unspliced transcripts (Mousseau et al., 2015), would limit inflammation, and offer an important addition to the current anti-HIV-1 drug armamentarium.
METHODS
Data reporting
No statistical methods were used to predetermine sample size. The experiments were not randomized. The investigators were not blinded to allocation during experiments and outcome assessment.
Plasmids
The plasmids used here were either previously described or generated using standard cloning methods (Pertel et al., 2011). The full list of plasmids used here, along with their purpose and characteristics, is provided in Table 1. All plasmid DNAs with complete nucleotide sequence files are available at www.addgene.com.
Cell culture
Cells were cultured at 37°C in 5% CO2 humidified incubators and monitored for mycoplasma contamination using the Lonza Mycoplasma Detection kit by Lonza (LT07-318). HEK293 cells (ATCC) were used for viral production and were maintained in DMEM supplemented with 10% FBS, 20 mM L-glutamine (ThermoFisher), 25 mM HEPES pH 7.2 (SigmaAldrich), 1 mM sodium pyruvate (ThermoFisher), and 1x MEM non-essential amino acids (ThermoFisher). Cytokine conditioned media was produced from HEK293 cells stably transduced with pAIP-hGMCSF-co (Addgene #74168), pAIP-hIL4-co (Addgene #74169), or pAIP-hIL2 (Addgene #90513), as previously described (Pertel et al., 2011).
Leukopaks were obtained from anonymous, healthy, blood bank donors (New York Biologics). As per NIH guidelines (http://grants.nih.gov/grants/policy/hs/faqs_aps_definitions.htm), experiments with these cells were declared non-human subjects research by the UMMS IRB. PBMCs were isolated from leukopaks by gradient centrifugation on Histopaque 1077 (Sigma-Aldrich). CD14+ mononuclear cells were isolated via positive selection using anti-CD14 antibody microbeads (Miltenyi). Enrichment for CD14+ cells was routinely>X98%.
To generate DCs or macrophages, CD14+ cells were plated at a density of 1 to 2 × 106 cells/ml in RPMI-1640 supplemented with 5% heat inactivated human AB+ serum (Omega Scientific, Tarzana, CA), 20 mM L-glutamine, 25 mM HEPES pH 7.2, 1 mM sodium pyruvate, and 1x MEM non-essential amino acids (RPMI-HS complete) in the presence of cytokines that promote differentiation. DCs were generated by culturing monocytes for 6 days in the presence of 1:100 cytokine-conditioned media containing human GM-CSF and human IL-4. DC preparations were consistently >99% DC-SIGNhigh, CD11chigh, and CD14low by flow cytometry. Macrophages were generated by culturing for 7 days with GM-CSF conditioned media in the absence of IL-4, and were routinely >99% CD11b. CD4+ T cells were isolated from PBMCs that had been depleted of CD14+ cells, as above, using anti-CD4 microbeads (Miltenyi), and were >99% CD4+. CD4+ T cells were then cultured in RPMI-1640 supplemented with 10% heat inactivated FBS, 20 mM L-glutamine, 25 mM HEPES pH 7.2, 1 mM sodium pyruvate, 1x MEM non-essential amino acids (RPMI-FBS complete), and 1:2000 IL-2 conditioned media. Cells from particular donors were excluded from experiments if percent enrichment deviated more than 5% from the numbers mentioned above, or if there was no increase in activation markers in response to control stimuli (LPS, Sendai virus, and wild-type HIV-1-GFP).
HIV-1 vector production
HEK293E cells were seeded at 75% confluency in 6-well plates and transfected with 6.25 uL Transit LT1 lipid reagent (Mirus) in 250 μL Opti-MEM (Gibco) with 2.25 µg total plasmid DNA. 2-part HIV-1 vectors based on HIV-1-GFP (He et al., 1997) and described in detail in Table 1 were transfected at a 7:1 ratio in terms of µgs of HIV-1 plasmid DNA to pMD2.G VSV G expression plasmid DNA (Pertel et al., 2011). 3-part lentivectors were produced by transfection of the lentivector genome, psPAX2 GagPol vector, and pMD2.G, at a DNA ratio of 4:3:1. These also include 2-part HIV-1-GFP constructs that are mutated in such a way as to prevent GagPol, Tat, or Rev production. As these would be defective for viral production, psPAX2 was included in the transfections at the same 4:3:1 ratio. VPX-bearing SIV-VLPs were produced by transfection at a 7:1 plasmid ratio of SIV3+ to pMD2.G (Pertel et al., 2011). 12 hrs after transfection, media was changed to the specific media for the cells that were to be transduced. Viral supernatant was harvested 2 days later, filtered through a 0.45 µm filter, and stored at 4°C.
Virions in the transfection supernatant were quantified by a PCR-based assay for reverse transcriptase activity (Pertel et al., 2011). 5 μl transfection supernatant were lysed in 5 μL 0.25% Triton X-100, 50 mM KCl, 100 mM Tris-HCl pH 7.4, and 0.4 U/μl RNase inhibitor (RiboLock, ThermoFisher). Viral lysate was then diluted 1:100 in a buffer of 5 mM (NH4)2SO4, 20 mM KCl, and 20 mM Tris–HCl pH 8.3. 10 μL was then added to a single-step, RT PCR assay with 35 nM MS2 RNA (IDT) as template, 500 nM of each primer (5’-TCCTGCTCAACTTCCTGTCGAG-3’ and 5’-CACAGGTCAAACCTCCTAGGAATG-3’), and hot-start Taq (Promega) in a buffer of 20 mM Tris-Cl pH 8.3, 5 mM (NH4)2SO4, 20 mM KCl, 5 mM MgCl2, 0.1 mg/ml BSA,
1/20,000 SYBR Green I (Invitrogen), and 200 μM dNTPs. The RT-PCR reaction was carried out in a Biorad CFX96 cycler with the following parameters: 42°C 20 min, 95°C 2 min, and 40 cycles [95°C for 5 s, 60°C 5 s, 72°C for 15 s and acquisition at 80°C for 5 s]. 2 part vectors typically yielded 107 RT units/µL, and 3 part vector transfections yielded 106 RT units/µL.
Transductions
106 DCs/mL, or 5 × 105 macrophages/ml, were plated into RPMI-HS complete with Vpx+ SIV-VLP transfection supernatant added at a dilution of 1:6. After 2 hrs, 108 RT units of viral vector was added. In some cases, drugs were added to the culture media as specified in Table 2. In most cases, transduced DC were harvested for analysis 3 days following challenge. For gene knockdown or for expression of factors in trans, 2 × 106 CD14+ monocytes/mL were transduced directly following magnetic bead isolation with 1:6 volume of SIV-VLPs and 1:6 volume of vector. When drug selection was required, 4 µg/mL puromycin was added 3 days after monocyte transduction and cells were selected for 3 days. SIV-VLPs were re-administered in all cases with HIV-1 or lentivector challenge. For DCs in Tet-HIV-1 experiments, fresh monocytes were SIV-VLP treated and co-transduced with rtTA3 and Tet-HIV-1. DCs were harvested 6 days later and treated with indicated doxycycline concentrations.
For deoxynucleoside-assisted transductions, DCs were plated at 106 DCs/mL and treated with 2mM of combined deoxynucleosides for 2 hrs before transduction with HIV-1. Deoxynucleosides were purchased from Sigma-Aldrich (2′deoxyguanosine monohydrate, cat# D0901; thymidine, cat# T1895; 2′deoxyadenosine monohydrate, cat# D8668; 2′deoxycytidine hydrochloride, cat# D0776). A 100 mM stock solution was prepared by dissolving each of the four nucleotides at 100 mM in RPMI 1640 by heating the medium at 80°C for 15 min.
CD4+ T cells were stimulated in RPMI-FBS complete with 1:2000 hIL-2 conditioned media and 5 μg/mL PHA-P. After 3 days, T cells were replated at 106 cells/mL in RPMI-FBS complete with hIL-2. Cells were transduced with 108 RT units of viral vector per 106 cells and assayed 3 days later. T cells were co-transduced with rtTA3 and Tet-HIV-1 every day for 3 days after PHA stimulation. Cells were then replated in RPMI-FBS complete with hIL-2. Transduced T cells were cultured for 9 days with fresh media added at day 5. After 9 days, doxycycline was added at the indicated concentrations and assayed 3 days later.
Non-HIV-1 Challenge Viruses
Sendai Virus Cantell Strain was purchased from Charles River Laboratories. Infections were performed with 200 HA units/ml on DCs for 3 days before assay by flow cytometry.
Spreading Infections
DCs were plated at 106 DCs/mL, in RPMI-HS complete media, with or without Vpx+ SIV-VLP transfection supernatant added at a dilution of 1:6. After 2 hrs, 108 RT units of HEK-293 transfection supernatant of either NL4-3-GFP with JRCSF env (T tropic) or NL4-3-GFP with JRFL env (mac tropic) was added. Every 3 days (for a total of 12 days) samples were harvested for detection of viral RT activity in supernatant and flow cytometry assessment.
CD4+ T cells were stimulated in RPMI-FBS complete with 1:2000 hIL-2 conditioned media and 5 μg/mL PHA-P. After 3 days, T cells were replated at 106 cells/mL in RPMI-FBS complete with hIL-2 and transduced with 108 RT units of NL4-3-GFP with JRFL (mac tropic) or JRCSF (T tropic) env. Cells were harvested every 3 days (for a total of 12 days) and assayed for infectivity and activation via flow cytometry.
Cytokine analysis
Supernatants from DCs were collected 3 days following transduction with HIV-1-GFP or minimal lentivector. Supernatant was spun at 500 × g for 5 mins and filtered through a 0.45 μm filter. Multiplex soluble protein analysis was carried out by Eve Technologies (Calgary, AB, Canada).
qRT-PCR
Total RNA was isolated from 5 × 105 DCs using RNeasy Plus Mini (Qiagen) with Turbo DNase (ThermoFisher) treatment between washes. First-strand synthesis used Superscript III Vilo Master mix (Invitrogen) with random hexamers. qPCR was performed in 20 μL using 1× TaqMan Gene Expression Master Mix (Applied Biosystems), 1 μL cDNA, and 1 μL TaqMan Gene Expression Assays (ThermoFisher) specified in Table 3. Amplification was on a CFX96 Real Time Thermal Cycler (Bio-Rad) using the following program: 95°C for 10 min [45 cycles of 95°C for 15 s and 60°C for 60 s]. Housekeeping gene OAZ1 was used as control (Pertel et al., 2011).
Flow cytometry
105 cells were surface stained in FACS buffer (PBS, 2% FBS, 0.1% Sodium Azide), using the antibodies in Table 4. Cells were then fixed in a 1:4 dilution of BD Fixation Buffer and assayed on a BD C6 Accuri. BD Biosciences Fixation and Permeabilization buffers were utilized for intracellular staining. Data was analyzed in FlowJo.
Sampling
All individual experiments were performed with biological duplicates, using cells isolated from two different blood donors. Flow cytometry plots in the figures show representative data taken from experiments performed with cells from the number of donors indicated in the figure legends.
Statistical Analysis
Experimental n values and information regarding specific statistical tests can be found in the figure legends. The mean fluorescence intensity for all live cells analyzed under a given condition was calculated as fold-change to negative control/mock. The exception to this methodology was in Figures 1g and 4c where the percent infected cells was too low to use MFI for the bulk population; in these cases MFI was determined for the subset of cells within the GFP+ gate. Significance of flow cytometry data was determined via one-way ANOVA. A Dunnett’s post-test for multiple comparisons was applied, where MFI fold change was compared to either mock treatment or positive treatment depending on the experimental question. qRT-PCR and luminex data was analyzed via two-way Anova, with Dunnett’s post-test comparing all samples to mock. All ANOVAs were performed using PRISM 7.02 software (GraphPad Software, La Jolla, CA).
Data availability
The plasmids described in Table 1, along with their complete nucleotide sequences, are available at www.addgene.com.
ACKNOWLEDGEMENTS
We thank the 256 anonymous blood donors who contributed leukocytes to this project, and Ben Berkhout, Abraham Brass, Barbara Felber, Massimo Pizzato, and Didier Trono for advice and reagents. This work was supported by NIH Grants 5R01AI111809, 5DP1DA034990, and 1R01AI117839, to J.L. The plasmids described in Table 1, along with their complete nucleotide sequences, are available at www.addgene.com.