De novo expression of HIV-1 intron-containing RNA (icRNA) induces caspase-dependent IL-1β activation in macrophages

To identify the potential triggers of diverse innate immune responses in macrophages, we differentiated peripheral blood CD14⁺ monocytes into macrophages (MDMs), and infected these cells with replication-competent, CCR5-tropic HIV-1 (herein referred to Lai/YU-2env) together with SIVmac239 Vpx-incorporating VLPs (Vpx-VLPs) in the absence or presence of RT (EFV), integrase (Ral), transcription (spironolactone) or CRM1 (KPT335, blocks Rev-mediated HIV icRNA export and Gag expression) inhibitors. Establishment of productive HIV infection and secretion of cytokines was measured by ELISA. We previously showed that HIV-1 infection in MDMs resulted in IP-10 secretion [56]. Here, we show that HIV-1 infection of MDMs also leads to secretion of IL-1β, which was abrogated upon pre-treatment of cells with HIV-1 RT, integrase, transcription, or Rev-CRM1 inhibitors (Fig 1A and 1B), suggesting that nucleo-cytoplasmic export of unspliced or partially-spliced HIV-1 RNA is required for both type I IFN responses and IL-1β secretion. Consistent with the requirement for inflammatory caspase activity for processing of pro-IL-1β into mature IL-1β [60], we observed that pre-treatment of cells with either a pan-caspase (zVAD-fmk) or a caspase-1/caspase-4 inhibitor (VX-765) markedly abrogated HIV-1-induced IL-1β secretion in primary MDMs (Fig 1C and 1D).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. De novo expression of HIV-1 icRNA induces inflammasome activation in macrophages.

(A and B) MDMs were infected with HIV-1 Lai/YU-2env (MOI 1 in the presence of Vpx VLPs) and production of (A) p24^Gag or (B) IL-1β in the culture supernatants at 3 days post-infection (dpi) was quantified by ELISA. Infections were done in the absence or presence of HIV-1 inhibitors: EFV: efavirenz (1 µM), Ralt: raltegravir (30 µM), Spiro: spironolactone (100 nM), KPT335: verdinexor (1 µM). Alternatively, MDMs were stimulated with LPS (0.25 µg/ml, 4 h), followed by Nigericin (15 nM, 30 min). (C and D) MDMs were infected with Lai/YU-2env (MOI 1, in the presence of pan-caspase (z-VADfmk, 100 nM) or caspase-1/4 specific inhibitor (VX765, 1µM)), and (C) p24^Gag production or (D) IL-1β secretion was measured by ELISA in the culture supernatants at 3 dpi. (E and F) MDMs infected with Lai/YU-2env (MOI 1 or 2) in the presence of either Vpx VLPs or dNs, and (E) p24^Gag production or (F) IL-1β secretion in the culture supernatants was quantified at 3 dpi by ELISA. (G–I) Infection of MDMs with Lai∆envGFP/G (MOI 1) WT or M10. (G) % infection (GFP⁺) at 3 dpi was quantified by flow cytometry. (H) Cleaved caspase-1 and β-actin expression in the whole cell lysates was analyzed by immunoblotting. (I) IL-1β secretion by MDMs infected with HIV-1 (WT or M10) at MOI 1 in the absence or presence of integrase inhibitor (raltegravir, 30 µM), was measured by ELISA. The means ± SEM are shown, and each symbol represents an independent donor. Statistical significance was determined by one-way ANOVA followed by Dunnett’s post-test, compared to the uninfected (mock) control. P-values: **** < 0.0001; no symbol: not significant (p ≥ 0.05). The data underlying this figure can be found in S1 Data and S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003320.g001

Since human monocytes and macrophages have been hypothesized to require priming for inflammasome activation and IL-1β production, we also included TNFα for MDM stimulation, even though HIV-1 icRNA-mediated inflammasome activation in primary MDMs infected with Lai/YU-2env or VSV-G pseudotyped single-cycle HIV-1 did not require priming by exogenous cytokines or other stimuli (Fig 1). Priming with TNFα led to slightly enhanced infection levels and IL-1β secretion in the absence of Vpx-VLPs (S1A and S1B Fig). Supplementation with Vpx-VLPs not only enhanced infection by over 100-fold, but was also sufficient to trigger IL-1β secretion even in the absence of TNFα (S1A and S1B Fig). Since, SAMHD1 inactivation or exogenous Vpx expression in MDMs has been hypothesized to induce innate immune responses [61,62], primary MDMs were infected with Lai/YU-2env in the presence of deoxynucleosides (dNs) to overcome RT restriction in MDMs [63]. Pretreatment with exogenous dNs enhanced HIV-1 infection in MDMs to levels similar to that observed upon pre-treatment with Vpx-VLPs (S1C Fig) and resulted in robust secretion of IL-1β, (S1D Fig), suggesting that HIV-1 infection and viral icRNA expression alone in the absence of priming stimuli can activate inflammasomes in MDMs. To further confirm the requirement of HIV-1 icRNA nuclear export for IL-1β secretion, we infected primary MDMs with VSV-G-pseudotyped single-round HIV-1 expressing GFP (Lai∆envGFP/G, MOI = 1) wildtype (WT) or a rev-deficient mutant (M10) (Fig 1G). As described in our previously published studies [55,56], HIV-1/M10 infection of MDMs selectively attenuates nucleocytoplasmic export of HIV-1 icRNA while preserving multiply spliced viral RNA (msRNA) nuclear export and GFP reporter expression (Fig 1G). In contrast to WT HIV-1 infection of MDMs, caspase-1 activation (Fig 1H) and IL-1β secretion (Fig 1I) were not observed in HIV-1/M10-infected MDMs (Fig 1I). Collectively these findings suggest that cytoplasmic expression of HIV-1 icRNAs is required for inflammasome activation in HIV-1-infected primary macrophages.

HIV-1 icRNA-induced inflammasome activation in macrophages is protease-independent

Recent studies have demonstrated HIV-1 PR-dependent CARD8 inflammasome activation in productively infected cells, which was dependent either on transient overexpression of Gag-Pol polyprotein or non-nucleoside RT inhibitor (NNRTI)-induced Gag-Pol dimerization in infected cells [53,54,64]. Since HIV-1 unspliced RNA encodes for the Gag-Pol polyprotein, including the viral PR, we sought to determine whether Gag-Pol polyprotein expression and viral PR activity were essential for inflammasome activation and IL-1β secretion in HIV-1-infected MDMs. We generated an HIV-1 provirus in which the gag-pol sequence was deleted (HIV-1/ΔGag-Pol, [56]), rendering it null for HIV-1 Gag and PR expression. Both Gag-Pol polyprotein and immature p55^Gag expression were not observed in HEK293T cells transfected with HIV-1/ΔGag-Pol proviral plasmid (Fig 2A). We confirmed similar levels of GFP expression upon infection with HIV-1/ΔGag-Pol to that observed with HIV-1/WT and HIV-1/M10 in primary MDMs at 3 days post-infection (dpi) (Fig 2B). Strikingly, IL-1β secretion was detected in MDMs infected with HIV-1 lacking PR activity or Gag-Pol expression (HIV-1/ΔGag-Pol) in an MOI-dependent manner (S1C and S1D Fig) and IL-1β was significantly higher than that of HIV-1/M10 or uninfected (mock) MDMs, although somewhat attenuated compared to HIV-1/WT (Fig 2C), suggesting that PR activity was dispensable for HIV-1 icRNA-induced inflammasome activation in MDMs.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. HIV-1 icRNA-induced inflammasome activation in macrophages is protease-independent.

(A) Immunoblotting analysis of HEK293T cell lysates transfected with WT, M10, or ∆Gag-pol HIV-1 constructs, compared to mock (no HIV-1 plasmid transfection). (B) % infection (GFP⁺) in MDMs (MOI 1) was measured by flow cytometry at 3 dpi. (C) Secretion of IL-1β in the culture supernatants of HIV-infected MDMs at 3 dpi measured by ELISA. Infection of MDMs with WT (D) or ∆Gag-pol (E) HIV-1 (MOI 1, 3 dpi) in the presence of HIV-1 protease (PR) inhibitors, Darunavir (DAR, 1μM) and Lopinavir (LOP, 1 μM), or RT inhibitor (EFV, 1 μM), compared to untreated (DMSO) control. % infection (GFP⁺) quantified by flow cytometry. IL-1β secreted in the culture supernatants from WT (F) or ∆Gag-pol (G) infected MDMs in the absence (DMSO) or presence of DAR, LOP, or EFV at 3 dpi, quantified by ELISA. The means ± SEM are shown, and each symbol represents an independent donor. Statistical significance was determined by one-way ANOVA followed by Dunnett’s post-test (C, F, G), compared to mock. P-values: **** < 0.0001; *** < 0.001; ** < 0.01; * < 0.05; no symbol: not significant (p ≥ 0.05). The data underlying this figure can be found in S2 Data.

https://doi.org/10.1371/journal.pbio.3003320.g002

To confirm this HIV-1 PR-independent mechanism of inflammasome activation, we infected MDMs with either WT or PR-deficient HIV-1 (∆Gag-pol, (MOI = 1) for 3 days in the absence (DMSO) or presence of HIV-1 PR inhibitors (Darunavir (DAR) and Lopinavir (LOP)), or an HIV-1 RT inhibitor (efavirenz (EFV)) (Fig 2D). While HIV-1 infection establishment was unaffected upon treatment with PR inhibitors (Fig 2D and 2E), both DAR and LOP pretreatment reduced, but did not completely abrogate IL-1β secretion in HIV-1/WT-infected MDMs (Fig 2F), while having a negligible impact on IL-1β secretion in HIV-1/ΔGag-Pol-infected MDMs (Fig 2G, respectively). The lack of complete attenuation of IL-1β secretion in HIV-1/WT-infected MDMs in the presence of PR inhibitors suggests that cytoplasmic HIV-1 icRNA expression alone triggers inflammasome activation and IL-1β secretion in MDMs, independent of HIV-1 PR activity or Gag-Pol dimerization.

Cytoplasmic expression of de novo HIV-1 icRNA in iMGs leads to inflammasome activation

Since we recently reported that human iMGs) are highly susceptible to HIV-1 infection [55], we next sought to look at the ability of HIV-1 icRNA to induce IL-1β release in iMGs. We generated iMGs from hematopoietic progenitors derived from two iPSC lines (Fig 3A), and obtained a highly pure microglia population (>98% CD45⁺CD11b⁺), expressing microglia-specific markers TREM2 and P2RY12 (Fig 3B). While HIV-1 infection of myeloid cells is antagonized by SAMHD1 [61,65–69], iMGs expressed lower levels of antiviral (dephosphorylated) SAMHD1 compared to primary MDMs (S2A Fig) and were permissive to both Lai/YU-2env (Figs 3C and S2B) and Lai∆envGFP/G (S2C Fig), and did not require co-infection with Vpx-VLPs or supplementation with excess dNs to boost infection. Importantly, infection of iMGs from both iPSC lines (iMG #1 and iMG #2) with Lai/YU-2env resulted in an MOI-dependent increase in p24^Gag secretion (Fig 3C; left panel) and IL-1β secretion (Fig 3C; right panel) at day 3 pi, which was suppressed upon EFV treatment. Infection of iMGs with single-cycle HIV-1 (Lai∆envGFP/G) also resulted in robust infection (S2D Fig, left panel) and led to significant IL-1β production which was abrogated upon EFV pretreatment (S2D Fig, right panel), suggesting that viral spread was not required for IL-1β secretion. These results suggest that similar to the findings in MDMs, HIV-1 infection in primary microglia leads to inflammasome activation and IL-1β release.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Cytoplasmic expression of HIV-1 icRNA in iMGs induces inflammasome activation.

(A) Schematic of iMGs differentiation protocol. (B) Flow cytometry profiles of iMGs showing expression of myeloid cell markers (CD45 and CD11b) and microglia-specific markers (TREM2 and P2RY12) at day 28 of differentiation. Top panels represent isotype IgG control staining, and bottom panels represent specific antibodies against indicated cell-specific markers. (C) iMGs were infected with HIV-1 Lai/YU-2env (MOI 0.1, 0.3, or 1) ± EFV (1 μM). Alternatively, iMGs were stimulated with LPS (0.25 µg/ml, 4 h) followed by Nigericin (15 nm, 30 min). Cell culture supernatants were harvested at 3 dpi for quantifying p24^Gag (left panel) and IL-1β (right panel) release by ELISA. RT-qPCR analysis of multiply-spliced (tat/rev) viral transcripts (D) and unspliced (usRNA) transcripts (E) in iMGs infected with HIV-1 WT, M10, or ΔGag-pol at 3 dpi. The CT values were normalized first to that of GAPDH, and then to uninfected (mock) control. (F) Quantitation of IL-1β secretion in the culture supernatants of HIV-infected iMGs at 3 dpi, compared to uninfected (mock) control. The means ± SD are shown, and each symbol represents an independent experiment. Statistical significance was determined by two-way ANOVA followed by the Dunnett’s post-test comparing to EFV-treated (C, left panel) or mock (C, right panel); or one-way ANOVA followed by the Dunnett’s post-test comparing to HIV-1 WT (D–F). P-values: **** < 0.0001; *** < 0.001; ** < 0.01; ns: not significant (p ≥ 0.05). The data underlying this figure can be found in S3 Data.

https://doi.org/10.1371/journal.pbio.3003320.g003

We next sought to determine whether cytoplasmic HIV-1 icRNA expression alone, independent of Gag-Pol expression or viral PR activity, is a trigger of inflammasome activation and IL-1β secretion in iMGs, similar to the findings in MDMs (Fig 2). To this end, we infected iMGs with single-cycle HIV-1 WT or mutant (M10 and ∆Gag-pol) viruses (MOI = 1) and confirmed infection establishment by RT-qPCR analysis for viral RNA expression from cytoplasmic fractions of infected cells at 3 dpi (Fig 3D and 3E). While similar levels of multiply-spliced viral transcripts (tat/rev) were observed in microglia infected with either WT, M10, or ∆Gag-Pol viruses (Fig 3D), unspliced viral RNA (icRNA transcripts) were only detected in cytoplasmic fractions of WT or ∆Gag-Pol, but not in M10-infected microglia (Fig 3E). In correlation to cytoplasmic viral icRNA expression, infection of iMGs with either WT or ΔGag-Pol/HIV-1, but not M10, resulted in IL-1β secretion (Fig 3F). These findings suggest that inflammasome activation in microglia requires cytoplasmic viral icRNA expression and that, similar to primary MDMs (Figs 1 and 2), Gag-Pol expression or PR activity is not essential for IL-1β secretion.

Cytoplasmic expression of HIV-1 icRNA induces inflammasome activation in macrophages in a MAVS-independent manner

Previously published findings by us and others have demonstrated the selective requirement of the Rev/CRM1-mediated nuclear export of HIV-1 icRNA for inducing MDA5-dependent type I responses [57,59]. In contrast, HIV-1 icRNA nuclear export mediated by the constitutive transport element (CTE) from Mason–Pfizer monkey virus (MPMV) failed to induce type-I interferon responses [56]. Hence, we sought to confirm the requirement of Rev-CRM1 nuclear export pathway in triggering icRNA-mediated inflammasome activation and IL1-β secretion. MDMs were infected with single-cycle VSV-G pseudotyped WT Lai∆envGFP/G (Rev/CRM1-dependent nuclear export of HIV-1 icRNA), M10 (Rev-deficient), or M10-4xCTE virus (MPMV CTE-dependent nuclear export of HIV-1 icRNA) in the absence (DMSO) or presence of EFV (1 μM). Analysis for infection efficiency (%GFP positive) by flow cytometry revealed that GFP expression in productively infected MDMs was equivalent amongst all three viruses (Fig 4A). Notably, the insertion of four copies of CTE from MPMV in the pol open reading frame of the M10 mutant virus led to rescue of cytoplasmic p55^Gag protein expression (Fig 4B), and unspliced (gRNA) HIV-1 RNA expression (Fig 4C) and spliced (Tat/Rev) transcripts (Fig 4D) in macrophages. In alignment with previously published studies, type-I IFN production was only observed when cytoplasmic expression of HIV-1 icRNA was facilitated by Rev-CRM1 nuclear export pathway (Fig 4E, [56,57]. In contrast, nuclear export and cytoplasmic expression of HIV-1 icRNA by either Rev-CRM1 or MPMV CTE induced IL-1β secretion (Fig 4F). While IL-1β production in HIV-1/M10-4xCTE infected macrophages were lower than that observed with HIV-1 WT-infected MDMs, levels were significantly enhanced over that observed with HIV-1/M10 or mock-infected MDMs (Fig 4F).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Cytoplasmic expression of HIV-1 icRNA induces inflammasome activation in macrophages in a MAVS-independent manner.

(A) Flow cytometry analysis (left panel) of MDMs infected with WT or mutant (M10 or M10-4xCTE) Lai∆envGFP/G viruses (MOI 1) in the presence of SIV3 + /Vpx VLPs, compared to mock (no virus infection). % infection (GFP⁺) in MDMs was quantified at 3 dpi (right panel). (B) Immunoblotting analysis of whole cell lysates from MDMs infected with HIV-1 (WT, M10, or M10-4xCTE constructs) in the absence (untreated, DMSO) or presence of RT inhibitor (EFV, 1 μM) at 3 dpi. HIV-1 proteins (p55^Gag and p24^Gag) were visualized using anti-HIV-1 IgG, β-actin was used as a loading control. (C) RT-qPCR quantification of unspliced (usRNA), and multiply-spliced (Tat/Rev) transcripts (msRNA) (D) in whole cell lysates from MDMs infected with WT, M10, or M10-4xCTE virus ± EFV, at 3 dpi compared to uninfected (mock) lysates. HIV-1 unspliced and spliced transcripts were quantified by RT-qPCR. Production of bioactive type-I IFN (E), and secretion of IL-1β (F) in culture supernatants from MDMs infected with WT, M10, or M10-4xCTE virus ± EFV were harvested at 3 dpi and quantified using type I IFN bioassay or ELISA, respectively. (G, H) Transient knockdown of MAVS, STING and UNC93B1 by siRNA transfection in primary MDMs. Knockdown efficiency was quantified by RT-qPCR (G) at day 2 post-transfection and reported as relative expression normalized to that observed with scramble siRNA (siControl), and protein level analyzed by immunoblotting (H). siRNA-transfected MDMs were infected with Lai∆envGFP/G (MOI 1) for 3 days (I–L). Culture supernatants were harvested and analyzed at 3 dpi for p24^Gag (I), IL-1β (J), and IP-10 (K) secretion by ELISA. (L) Bioactive type I IFN in the culture supernatants from infected MDMs at 3 dpi quantified using interferon bioassay. The values were normalized to that of mock-infected control in each donor. The means ± SEM are shown, with each symbol representing an independent donor. The means ± SEM are shown, and each symbol represents an independent donor. Statistical significance was determined by two-way ANOVA followed by Tukey’s multiple comparisons (C, D), or Dunnett’s post-test, comparing HIV-infected MDMs to uninfected (mock) control (E, F), or with specific siRNAs to that from siControl (K, L). P-values: **** < 0.0001; *** < 0.001; ** < 0.01; * < 0.05; ns: not significant (p ≥ 0.05). The data underlying this figure can be found in S4 Data.

https://doi.org/10.1371/journal.pbio.3003320.g004

Myeloid cells utilize a wide range of pattern recognition receptors to sense pathogen determinants and initiate antiviral responses. Diverse viral RNA sensors, including RLRs such as RIG-I and MDA5 [70,71], TLRs [72,73], and NLRs [21,23,24,74–80], sense and respond to viral RNAs. Since our previous studies had implicated MAVS-dependent induction of type I IFN and ISG expression upon cytoplasmic HIV-1 icRNA expression [56], we sought to determine whether HIV-1 icRNA-induced IL-1β secretion also required MAVS-initiated signaling cascades. To this end, we used siRNAs to transiently knock down the expression of adaptor proteins for cytosolic viral nucleic acid sensors (MAVS and STING) or a chaperone protein (UNC93B1), which are important for innate immune activation and type I interferon responses upon RLR, cGAS and endosomal TLR activation, respectively [81–86]. Downregulation of MAVS, STING and UNC93B1 expression was confirmed by RT-qPCR (Fig 4G), and immunoblotting (Fig 4H). siRNA-treated MDMs were either left uninfected (mock) or infected with Lai∆envGFP/G (MOI = 1) in the absence (DMSO) or presence of EFV (1 μM), followed by quantification of p24^Gag and IL-1β secretion in culture supernatants at 3 dpi (Fig 4I and 4J).

Transient knockdown of MAVS, STING, or UNC93B1 did not affect HIV-1 infection establishment, as quantified by p24^Gag production in the culture supernatants, compared to non-targeted (scramble) control (Fig 4I). Surprisingly, knockdown of MAVS, STING, or UNC93B1 in primary MDMs did not abrogate HIV-1 icRNA-induced IL-1β secretion (Fig 4I). In contrast, IP-10 secretion (Fig 4J) or type I IFN induction (Fig 4K) was markedly diminished upon downregulation of MAVS expression but not STING or UNC93B1, in agreement with our previously published results that HIV-1 icRNA induces type I IFN responses in a MAVS-dependent manner [34,55,56]. These data suggest that cytosolic nucleic acid sensing pathways that require MAVS, STING, or UNC93B1 activities are dispensable for HIV-1 icRNA-mediated inflammasome activation and IL-1β secretion. Collectively, these findings suggest that nuclear export and cytoplasmic expression of HIV-1 icRNA can induce IL-1β production in a MAVS-independent manner, thus underlying the existence of divergent nucleic acid sensing pathways of cytosolic HIV-1 icRNA that trigger type I IFN production and inflammasome activation.

NLRP1 is required for HIV-1 icRNA-dependent inflammasome activation in macrophages

Besides cytosolic RLRs and endosomal TLRs, NLRs and AIM2-like inflammasomes have been hypothesized to sense HIV-1 genomic RNAs, abortive viral transcripts, and dsDNA (reverse transcription intermediates), respectively [41,43,44,87,88]. Therefore, we sought to determine the potential role of these inflammasome sensors in mediating IL-1β production upon cytoplasmic HIV-1 icRNA expression in macrophages. We used siRNAs to transiently knock down the expression of key inflammasome sensors (AIM2, NLRP1, NLRP3, and CARD8) and executioner caspases (Caspase-1 and Caspase-4) in MDMs, and confirmed significant knockdown of the target RNA expression by RT-qPCR (Fig 5A) and protein expression by immunoblotting (Fig 5B), as well as functional assays (S3A and S3B Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. NLRP1 is required for HIV-1 icRNA-dependent inflammasome activation in macrophages.

(A, B) Transient knockdown of AIM2, NLRP1, NLRP3, CARD8, CASP1, and CASP4 by siRNA transfection in primary MDMs. Knock-down efficiency was determined by RT-qPCR at day 2 post-transfection, and represented as relative mRNA expression of individual targets compared to scramble siControl (A), and protein level analyzed by immunoblotting (B). MDMs were infected with Lai∆envGFP/G (MOI 1) for 3 days in the absence (DMSO) or presence of EFV (C–G). Culture supernatants were harvested at 3 dpi and analyzed for p24^Gag (C), IL-1β (D), and IP-10 (E) secretion by ELISA. (F) Bioactive type I IFN in the culture supernatants from infected MDMs at 3 dpi quantified using interferon bioassay. The values in panels B-E were normalized to that of siControl-transfected and HIV-1-infected cells in each donor and reported as relative expression. (G) Quantitation of LDH release in culture supernatants was measured using the CytoTox cytotoxicity assay, followed by subtraction of background OD from culture media, and normalized to EFV-treatment for each siRNA transduction per donor. The means ± SEM are shown, and each symbol represents an independent donor. Statistical significance was determined by two-way ANOVA followed by Dunnett’s post-test comparing HIV-infected MDMs from siControl to indicated specific siRNA (D, G). P-values: **** < 0.0001; * < 0.05; no symbol: not significant (p ≥ 0.05). The data underlying this figure can be found in S5 Data.

https://doi.org/10.1371/journal.pbio.3003320.g005

To assess the role of inflammasomes in driving HIV-1 icRNA-induced IL-1β secretion, siRNA-transfected MDMs were either mock-infected or infected with Lai∆envGFP/G (MOI = 1) (Fig 5C and 5D). Notably, the knockdown of inflammasome pathways did not alter HIV-1 infection efficiency in MDMs (Fig 5C). Interestingly, knockdown of NLRP1 and caspase-1 expression, but not AIM2, NLRP3, CARD8, or caspase-4, led to a significant reduction in IL-1β secretion in HIV-1-infected MDMs (Fig 5D). In contrast, knockdown of any of the inflammasome components did not impact HIV-1 icRNA-induced IP-10 (Fig 5E) or type I IFN production (Fig 5F). Furthermore, there was a significant enhancement in LDH release at 3dpi from HIV-1-infected MDMs compared to uninfected (mock) or EFV-treated controls (Fig 5G), highlighting the loss of cell integrity and pyroptotic cell death. Moreover, LDH secretion was abrogated upon knockdown of either NLRP1 or caspase-1, comparable to uninfected or EFV-treated conditions (Fig 5G). Collectively, these findings suggest that NLRP1 inflammasome and caspase-1 activation are required for HIV-1 icRNA-induced IL-1β secretion and pyroptotic cell death in macrophages.

Overexpression of NLRP1 enables HIV-1 icRNA-dependent IL-1β secretion in THP-1/PMA macrophages

Since the knockdown of NLRP1 in primary MDMs resulted in attenuated IL-1β secretion in HIV-1-infected MDMs (Fig 5), we next sought to determine if NLRP1 expression alone is enough for inflammasome activation and IL-1β release. We over-expressed NLRP1 in THP-1 monocytes, which do not express detectable levels of endogenous NLRP1 (Fig 6A) to determine if NLRP1 expression rescues HIV-1 icRNA-induced IL-1β secretion. THP-1 cells were transduced with a lentivector expressing full-length NLRP1, and NLRP1 overexpression in THP-1 monocytes was confirmed by immunoblotting (Fig 6A). We next differentiated THP-1 monocytes into macrophage-like cells with phorbol 12-myristate-13-acetate (PMA, 100 nM). Immunofluorescence analysis confirmed a lack of NLRP1 expression in parental THP1/PMA macrophages (S4 Fig). In contrast, cytoplasmic NLRP1 expression was detected upon constitutive NLRP1 overexpression in THP1/PMA/NLRP1 macrophages, similar to MDMs and iMGs (S4 Fig). THP1/PMA macrophages were infected with Lai∆envGFP/G (MOI = 1) in the absence (infection control) or presence of an RT inhibitor (EFV), integrase inhibitor (Ral), and caspase 1/4 inhibitors (zVAD-fmk and VX-765) in the presence of TNFα. Flow cytometry analysis at 3 dpi showed similar levels of infection (represented by GFP expression) of both control THP1/PMA and THP1/PMA/NLRP1 macrophages (Fig 6B). As expected, infection was blocked in the presence of RT and integrase inhibitors but remained unchanged when cells were pre-treated with caspase inhibitors. Intriguingly, IL-1β secretion was significantly enhanced in HIV-1-infected THP1/PMA/NLRP1 macrophages compared to control THP1/PMA macrophages (Fig 6C). Additionally, despite robust infection establishment in the presence of both pan-caspase (z-VAD) and caspase-1 (VX-765) inhibitors (Fig 6B), IL-1β production was attenuated in THP1/PMA/NLRP1 macrophages compared to DMSO-treated infection control (Fig 6C). As a positive control, stimulation of both THP1/parental and THP1/PMA/NLRP1 macrophages with a DPP8/9 inhibitor, Val-boroPro (VbP), a known activator of NLRP1 inflammasome [89], induced robust IL-1β secretion only in NLRP1-overexpressing cells, confirming NLRP1 functionality.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. NLRP1 overexpression in THP-1/PMA macrophages enhances HIV-1 icRNA-induced IL-1 β production.

(A) Immunoblotting analysis of NLRP1 expression in THP-1 monocytes transduced with either non-targeting (scramble) control or NLRP1-expressing lentivector. β-actin was used as a loading control. (B–D) THP1/PMA macrophages were primed with TNFα (10 ng/ml) and then infected with Lai∆envGFP/G (MOI 1) in the presence of HIV inhibitors (EFV and Ral) or caspase inhibitors (zVAD-fmk and VX-765). (B) % infection (GFP⁺ cells) was quantified by flow cytometry at 3 dpi, and (C) IL-1β secretion was analyzed by ELISA. (D, E) THP-1/PMA macrophages were primed (TNFα, 10 ng/ml) and then infected with Lai∆envGFP/G WT, M10, or ∆Gag-pol (MOI 1) for 3 days. (D) % infection (GFP⁺) was determined by flow cytometry, and (E) IL-1β secreted in culture supernatants was measured by ELISA at 3 dpi. Knockdown of MAVS or STING expression in THP1/NLRP1 monocytes transduced with lentiviral vectors expressing scramble shRNA control (scramble), or shRNA sequences against MAVS or STING confirmed by RT-qPCR (F) and immunoblotting (G). (H–J) Infection of primed (TNFα, 10 ng/ml) THP1/PMA macrophages with Lai∆envGFP/G (MOI 1) in the absence (no drug treatment, NT) or presence of HIV-1 RT inhibitor (EFV) for 3 days. (H) Flow cytometry analysis at 3 dpi showing infection (% GFP expression) in scramble (control) and knockdown cell lines. (I) Culture supernatants were harvested at 3 dpi and analyzed for IL-1β production by ELISA. (J) Quantitation of LDH release in the culture supernatants at 3 dpi was measured using the CytoTox cytotoxicity assay. The means ± SEM from 3 independent experiments are shown. Statistical significance was determined by two-way ANOVA followed by Dunnett’s post-test comparing WT HIV-infected Control or NLRP1-expressing THP-1/PMA macrophages in the absence or presence of indicated inhibitors (C) or infections with HIV mutants (E), or WT HIV-infected or VbP-treated THP-1/PMA macrophages expressing scramble shRNA (control) to those expressing specific shRNAs targeting either MAVS or STING (I–J). P-values: **** < 0.0001; * < 0.05; no symbol: not significant (p ≥ 0.05). The data underlying this figure can be found in S6 Data.

https://doi.org/10.1371/journal.pbio.3003320.g006

To confirm the requirement of cytoplasmic HIV-1 icRNA sensing for NLRP1- inflammasome activation, we infected parental and THP1/PMA/NLRP1 macrophages with Lai∆envGFP/G WT, M10, or ∆Gag-Pol (PR-deficient) viruses (MOI = 1) and observed relatively similar levels of infection between all three viruses when analyzed by flow cytometry (Fig 6D). HIV-1 (WT or ∆Gag-pol) infected THP1/PMA/NLRP1 macrophages primed with TNFα, but not TNFα -primed HIV-1/M10-infected macrophages, secreted robust levels of IL-1β (Fig 6E), thus confirming the requirement of nuclear export of HIV-1 icRNA for PR-independent NLRP1 activation-induced IL-1β secretion. Since IL-1β release only examines the ensemble effects of HIV-1 icRNA-induced inflammasome activation in infected THP1/PMA/NLRP1 macrophages, we sought to examine cell-intrinsic caspase-1 activation in infected cells by flow cytometry. FLICA staining showed robust upregulation of activated caspase-1 in virus-infected THP1/PMA/NLRP1 macrophages (S5A and S5B Fig), which was significantly attenuated upon pre-treatment with RT inhibitor (EFV) (S5B Fig, right panel). Interestingly majority of cells with active caspase-1 were GFP-deficient or uninfected cells (S5A Fig), suggesting that HIV-1 icRNA expression not only triggers caspase-1 activation and infected cell death but also transfer inflammatory responses to bystander cells.

To investigate whether NLRP1-mediated cytosolic HIV-1 icRNA sensing is independent of other cytosolic nucleic acid sensing pathways in myeloid cells, we stably knocked down expression of MAVS or STING in NLRP1⁺ THP-1/PMA cells by lentiviral transduction of shRNAs, and confirmed knockdown of MAVS and STING expression by RT-qPCR (Fig 6F) and immunoblotting (Fig 6G). Compared to control NLRP1⁺ THP-1/PMA cells, knockdown of MAVS or STING in NLRP1⁺ THP-1/PMA macrophages did not impact infection efficiency (Fig 6H). While knockdown of STING led to slightly lower level of secreted IL-1β compared to control (scramble) cells, knockdown of either adaptors still resulted in robust IL-1β production at 3 dpi compared to uninfected (mock) condition (Fig 6I). Similar to control NLRP1⁺ THP-1/PMA cells, pre-treatment with an RT inhibitor (EFV) markedly reduced IL-1β secretion in both cell lines, thus confirming the requirement for de novo viral icRNA sensing via NLRP1, independently of MAVS or STING expression. Lastly, quantitation of LDH release in the culture supernatants of HIV-1-infected NLRP1⁺ THP-1/PMA cells at 3 dpi corroborated the IL-1β data (Fig 6I), with comparable levels of LDH release into the culture supernatants of HIV-1 (WT)-infected control, MAVS or STING knockdown cells (Fig 6J).

Even though CARD8 has been reported to sense HIV-1 protease activity and trigger inflammasome activation in myeloid cells [52,54,90], knockdown of human CARD8 did not abrogate IL-1β secretion in HIV-1-infected primary macrophages (Fig 5C). To further investigate whether protease-dependent CARD8 activation is linked with cytoplasmic HIV-1 icRNA-dependent NLRP1 activation, PMA-differentiated THP-1/parental macrophages were primed with TNFα (10 ng/ml), and then infected with Lai∆envGFP/G WT or protease-inactive mutant (ΔPR) virus (MOI = 1) for 48 hours, followed by medium change and addition of HIV-1 RT inhibitor rilpivirine (RPV, 5 μM) or DMSO (vehicle control) for 24 hours (S5C–S5E Fig). Cells were harvested and analyzed for infection efficiency based on GFP expression by flow cytometry (S5C Fig), and the culture supernatants were harvested to measure IL-1β production by ELISA (S5D Fig) and LDH release (S5E Fig). Treatment of THP-1/PMA cells at 48 h post-infection with RPV resulted in dramatically increased IL-1β secretion and cell death in cells infected with WT but not ΔPR virus, consistent with the previous reports demonstrating CARD8 cleavage and inflammasome activation following NNRTI-induced Gag-pol dimerization and HIV-1 PR activation [91,92]. To investigate the role of CARD8 in NLRP1-mediated HIV-1 icRNA sensing, we transduced THP-1/control or THP-1/NLRP1 cells with control or CARD8-targeting shRNA (S5F–S5J Fig). Immunoblotting analysis confirmed the successful knockdown of CARD8 (S5F Fig) while having minimal effects on NLRP1 expression (S5G Fig). shRNA-transduced PMA-differentiated THP-1 cells were subsequently primed with TNFα (10 ng/mL) and infected with Lai∆envGFP/G WT virus (MOI = 1) for 72 h, or treated with VbP (100 μM) for 24 h. Cells and culture supernatants were harvested at the indicated time points, and GFP expression in cells was analyzed by flow cytometry (S5H Fig), IL-1β secretion by ELISA (S5I Fig), and LDH release (S5J Fig). Analysis of both IL-1β and LDH secretion confirmed that shRNA knockdown of human CARD8 does not abrogate either IL-1β or LDH release in both infected and VbP-stimulated cells. These findings suggest that functionally active CARD8 is dispensable for NLRP1-mediated HIV-1 icRNA sensing and inflammasome activation. Collectively, these data suggest that NLRP1 expression in THP-1/PMA macrophages leads to cytoplasmic HIV-1 icRNA sensing and IL-1β secretion independent of MAVS, STING, or CARD8 activation.

NLRP1 detects cytosolic HIV-1 icRNA expression

It was recently reported that NLRP1 binds viral dsRNAs generated during positive-strand RNA virus infection [93]. Positive-strand RNA viruses generate dsRNA replication intermediates in the cytoplasm of infected cells with regions of extensive base pairing [94–96]. Although dsRNA is not thought to be generated in HIV-1-infected cells, HIV-1 unspliced or icRNAs contain extensive secondary structures with RNA duplex regions limited to select sites in the viral genome, such as the rev response element, the 5′ untranslated region including the trans-activation response element and dimerization domain, and a hairpin structure in gag-pol orf that mediates a frameshift during synthesis of the Gag-Pol polyprotein [97,98]. We sought to investigate if HIV-1 infection of THP-1/PMA macrophages results in the expression of dsRNAs. To this end, we infected THP-1/PMA macrophages with Lai∆envGFP/G (WT) or Rev-deficient (M10) (MOI = 1) and processed for immunofluorescence analysis for dsRNA expression at 3 dpi with anti-dsRNA specific antibody J2, previously shown by us and others to detect dsRNAs, of at least 40 base pairs in size [99,100]. Immunofluorescence analysis of GFP and dsRNA expression (Fig 7A) and quantification of mean fluorescent intensity per cell (Fig 7B) showed that, while infection efficiency based on GFP expression between WT and M10/HIV-1 was similar (left panel), dsRNA puncta (highlighted by white arrowheads) are only expressed in HIV-1 WT but not M10 (right panel)-infected cells or in EFV-treated HIV-1-infected cells. Cytoplasmic dsRNA puncta was also detected in Lai/YU2env-infected MDMs and iMGs (S6A and S6B Fig). Even though direct co-localization of dsRNA and NLRP1 was not observed, distinct cytoplasmic expression of dsRNA suggested that positive dsRNA staining presumably corresponds to intron-containing viral usRNAs instead of msRNAs.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Cytoplasmic expression of HIV-1 icRNA in macrophages is detected by NLRP1.

(A) Immunofluorescence analysis of PMA-differentiated THP-1 parental (control) macrophages infected with Lai∆envGFP/G (WT, M10, or WT in the presence of RT inhibitor (EFV), MOI 1), and stained with anti-dsRNA (J2) antibody. The establishment of productive HIV-1 infection in THP1/PMA macrophages was visualized by GFP expression, and nucleus is stained with DAPI. The white arrowheads indicate positive staining of dsRNA puncta. Scale bar represents 5 μm. (B) Quantification of relative mean fluorescent intensity of GFP⁺ cells and dsRNA⁺ puncta. The means ± SEM are shown, and each dot corresponds to an individual image representing a total of 60−80 cells per image per infection condition. (C–F) THP-1/NLRP1-PMA macrophages infected with WT or Rev-deficient (M10) Lai∆envGFP/G, MOI = 1) were stained with anti-NLRP1 antibody (C) or anti-p24^Gag antibody (E) followed by hybridization with Quasar 670-smFISH probes targeting HIV-1 usRNA at day 3 pi. Data are representative of 2 independent experiments. Bar = 10 μm. Quantification of HIV-1 usRNA colocalization with NLRP1 (D) or HIV-1 p24^Gag (F) determined from at least 20 independent fields (representative fields shown in E and G, respectively). Approximately 50−100 infected cells were analyzed. The means ± SEM are shown, and each symbol represents an independent field/cell. (G, H) THP-1/parental and THP-1/NLRP1 monocytes infected with Lai∆envGFP/G (MOI 2) were harvested at 3 dpi, followed by fractionation into cytosolic and nuclear fractions. RT-qPCR quantification of HIV-1 usRNA (G), and tat/rev (msRNA) (H) in cytosolic fractions immunoprecipitated with anti-NLRP1 antibody or IgG control. Immunoprecipitated viral RNA copy numbers were determined using standard curves for both usRNA and msRNA, and the percent IP (immunoprecipitated usRNA or msRNA transcripts detected from each IP condition using anti-NLRP1 antibody or control IgG to the input amount in each cell type) is reported as mean ± SEM, where each symbol represents an independent experiment. (I) Schematic of the diversity of cytosolic HIV-1 icRNA sensing and innate immune activation pathways in myeloid cells. Image was created in BioRender. Jalloh, S. (2025) https://BioRender.com/q0g4tr1. Statistical significance was determined by one-way ANOVA followed by Dunnett’s post-test comparing EFV-treated HIV-WT infection to untreated WT or M10 (B), or IP from anti-NLRP1 antibody in NLRP1-overexpressing THP-1 to control THP-1 cells (G, H). Alternatively, statistical significance was determined by one-way ANOVA followed by Tukey’s multiple comparisons test for uninfected (mock), HIV-1/WT, and HIV-1/M10 infected THP1/NLRP1-PMA macrophages (D, F). P-values: **** < 0.0001; ** < 0.01; no symbol: not significant (p ≥ 0.05). The data underlying this figure can be found in S7 Data.

https://doi.org/10.1371/journal.pbio.3003320.g007

To determine if NLRP1 co-localizes with HIV-1 usRNA, we utilized single-molecule RNA FISH (smFISH) to specifically label HIV-1 usRNAs using Quasar670-conjugated oligonucleotides (S1 Table) in virus-infected THP-1/NLRP1-PMA macrophages, in tandem with immunofluorescence microscopy to detect endogenous NLRP1 expression (Fig 7C), and HIV-1 p24^Gag (Fig 7E). While HIV-1 usRNA was peripherally localized near the plasma membrane in wild-type HIV-1-infected cells (Fig 7C and 7E), HIV-1 usRNA localization was restricted to the nucleus in HIV-1/M10-infected macrophages (Fig 7F). Interestingly, co-localization of NLRP1 and p24^Gag was observed with HIV-1 usRNAs in wild-type virus-infected cells, but not in cells infected with Rev-deficient (M10) HIV-1 (Fig 7C and 7E), suggesting that de-novo transcribed cytosolic HIV-1 icRNAs (and not the incoming genome) are co-localized with NLRP1 in productively infected cells. To determine whether NLRP1 can bind to HIV-1 icRNAs, THP-1/control, and THP-1/NLRP1 monocytes were infected with Lai∆envGFP/G (MOI = 2), followed by fractionation of cellular lysates to nuclear and cytoplasmic fractions (Fig 7G and 7H). RNPs in cytoplasmic fractions were incubated with control IgG or anti-NLRP1 antibody, and co-immunoprecipitation of usRNA (Fig 7G) or msRNA (tat/rev) (Fig 7H) was quantified by RT-qPCR. Analysis of cytoplasmic HIV-1 usRNA or msRNA co-immunoprecipitated by anti-NLRP1 antibody compared to control IgG revealed preferential enrichment of HIV-1 usRNA in THP1/NLRP1 cells compared to the parental THP1 cells (Fig 7G). In contrast, no significant difference in immunoprecipitation of HIV-1 msRNA by NLRP1 antibody was observed between NLRP1-overexpressing cells compared to parental cells (Fig 7H). These findings suggest that NLRP1 interacts with HIV-1 usRNA, resulting in inflammasome activation and IL-1β secretion in myeloid cells.

Ethics statement

This research has been determined to be exempt by the Institutional Review Board of the Boston University Medical Center since it does not meet the definition of human subjects research, as all human samples were collected in an anonymous fashion and no identifiable private information was collected.

Viruses

HIV-1 replication-competent molecular clone, Lai/YU-2env, single-round GFP expressing reporter virus constructs, Lai∆envGFP (GFP in place of the nef orf), and rev-deficient mutant Lai∆envGFP M10, have been described previously [56]. Lai∆envGFP∆Gag-pol proviral plasmid, includes an additional deletion between N-terminus of CA and IN, and has been described previously [56]. The Lai∆envGFP M10-4xCTE construct was generated by PCR-amplifying CTE sequence from the plasmid, pCR4-CTE (Addgene #36868) using the following primers (forward: TTTTTGGATCCACTATAGGGCGAATTGAATTTAGCG; reverse: GGATAACAAT TTCACACAGG AAACAGCTAT GAC) and inserted into the Lai∆envGFP M10 proviral plasmid using BclI and NheI restriction enzymes. Replication-competent viruses were derived from HEK293T cells via calcium phosphate-mediated transient transfection of Lai/YU-2env. Single-round viruses pseudotyped with VSV-G were generated from HEK293T cells via co-transfection of Lai∆envGFP proviral plasmids and VSV-G expression plasmid (pHCMV-G). For generation of Lai∆envGFP/M10, Lai∆envGFP/M10-4XCTE or Lai∆envGFPΔGag-Pol, the HIV-1 Gag-Pol packaging plasmid (psPAX2) was co-transfected with the rev or gag-deficient proviral plasmids. SIV_mac Vpx containing VLPs were generated from HEK293T cells via co-transfection of SIV3⁺, and VSV-G expression plasmid, as described previously [56]. NLRP1-expressing lentivector (LV-NLRP1) was derived by recombining the pDONR-NLRP1 plasmid (Addgene #166823, [89]) with the pLenti CMV puro DEST w118-1 plasmid (Addgene #17452, [122]) via LR clonase (Gateway LR Clonase II Enzyme mix, Invitrogen # 11791020). Lentiviruses expressing NLRP1 (or vector control) were produced via co-transfection of HEK293T cells with NLRP1 expressing lentivector, psPAX2 and pHCMV-G. Virus-containing cell culture supernatants were harvested 2 days post-transfection, passed through 0.45 µm filters, and purified and concentrated by ultracentrifugation on a 20% sucrose cushion (24,000 rpm and 4°C for 2 h with a SW32Ti or SW28 rotor (Beckman Coulter)). Virus pellets were resuspended in PBS, aliquoted, and stored at −80°C until use. The capsid content of HIV-1 was determined by a p24^gag ELISA and virus titer was measured on TZM-bl by measuring β-gal activity as previously described [56].

Cells and viruses

HEK293T (ATCC) were maintained in DMEM (Gibco) containing 10% heat-inactivated FBS (Gibco) and 1% pen/strep (Gibco). THP-1 and TZMbl cell lines (NIH AIDS reagent program) were maintained in DMEM (Gibco) containing 10% FBS and 1% pen/strep. All cell lines are regularly tested for mycoplasma contamination and confirmed negative. Where indicated, THP-1 cells were stimulated with PMA (100 nM, Sigma-Aldrich) for 48 hours and treated with 10 ng/mL TNFα (Peprotech) for the duration of infection. To generate shRNA or NLRP1 expressing THP-1 cells, cells were transduced with lentivectors via spinoculation in the presence of polybrene (10 µg/ml, Sigma-Aldrich) before selection and maintenance with 2 µg/mL puromycin (Invivogen), plus 400 µg/mL hygromycin (Gibco) where indicated.

Isolation and differentiation of primary human macrophages

Peripheral blood mononuclear cells were isolated from de-identified leukopacks obtained from NY Biologics as described previously [123]. Monocytes were positively isolated using antibody-coated beads against CD14 (Miltenyi, Cat#130-050-201). Monocytes were cultured in 20 ng/mL M-CSF (Peprotech, catalog #300-25B), heat-inactivated 10% human AB serum (Sigma-Aldrich), 1% pen/strep in RPMI1640 (Gibco) for 5–6 days for differentiation into macrophages as previously described [56].

Generation of human iPSC-derived microglia

Two iPSC lines, BU1 and BU3 (herein referred to as iMG #1 and iMG #2, respectively) were first differentiated toward a mesodermal, hematopoietic lineage on a matrigel-coated 6-well plate in mTeSR (Stem Cell Technologies). Next, differentiation was initiated by culturing cells in a cocktail of growth factors, as described in [124], maintained in a hypoxic incubator (5% O₂). Cells were continually fed every 2 days and maintained in hypoxia up to day 6, varying the components and concentrations of the differentiation cocktail [124]. On day 6, the media composition was changed for the next 6 days (day 6–12), with the addition of cytokines, maintaining in hypoxia condition. On day 8, the cells are fed and returned to normoxia for the rest of the differentiation. At day 12, hematopoietic cells are harvested from the culture supernatant. This population typically contains 25%–65% (average 43%) CD34+CD45+ progenitor cells. Subsequently, the hematopoietic cells harvested are differentiated toward microglia lineage using STEMdiff Microglia Differentiation Kit and STEMdiff Microglia Maturation Kit.

Infections

Cells were spinoculated with HIV-1 (1 h at RT and 1,100g) at the indicated multiplicity of infection, cultured for 2−3 h at 37°C, washed to remove unbound virus particles, and cultured in fresh media for 2−3 days. Cells were seeded at a density of 0.5 million MDMs or iMGs/well or 0.25 million THP1/PMA macrophages in 12-well plates. For indicated experiments, cells were pre-treated for 30 min with inhibitors at 37°C before infection and maintained throughout the culture. Inhibitors used in this study are a reverse transcription inhibitor efavirenz (1 µM, NIH AIDS Research and Reference Reagent Program, catalog #4624), an integrase inhibitor raltegravir (30 µM or 60 µM, Selleck Chemicals, catalog #50-615-1), protease inhibitors; saquinavir, darunavir, and lopinavir (1 µM, NIH AIDS Research and Reference Reagent Program), spironolactone (30 µM; Selleck Chemicals), or KPT-335 (0.1 µM, verdinexor; Selleck Chemicals), a pan-caspase inhibitor zVAD-fmk (30 µM, Sigma, catalog #219007), a caspase-1 inhibitor VX-765 (1 µM, InvivoGen, catalog #inh-vx765-1), ValboroPro NLRP1/CARD8 inducer VbP (10 or 100 µM, InvivoGen, #tlrl-vbp-10). Where indicated, TNFα was added (10 ng/ml) as a priming signal in THP-1/PMA cells immediately after washing unbound virus or at 24 h before harvest. At the indicated time points, culture supernatants were harvested for ELISA, and cells were collected for RNA or protein analysis.

siRNA transfection and functional knockdown

MDMs were seeded overnight at a density of 0.5 or 0.25 million cells per well in 12- or 24-well plates, respectively. Scramble control siRNAs or specific siRNAs against indicated host targets were transfected at 50 nM for 2 days using the TransIT-X2 dynamic delivery system (Mirus, #MIR6000). At day 2 post-transfection, cells were either infected with HIV-1 (as described above), or stimulated as follows to determine functional knockdown: AIM2 (cells were treated with ultra-pure LPS (100 ng/m, Invivogen) for 2 h, followed by transfection with linearized DNA (1 μg/mL) for 4 h); NLRP1 and CARD8 (cells were primed with Pam3CSK4 (0.5 μg/mL, InvivoGen) for 4 h, followed by stimulation with Val-boroPro (10 μM, InvivoGen) for 24 h; NLRP3 and caspase-1 (cells were primed with ATP (5 mM, Thermo Scientific) for 6 h, followed by activation with nigericin (10 μM) for 60 min; Caspase-4 (cells were transfected with ultra-pure LPS (5 μg/ml, InvivoGen) for 6 hrs; UNC93B1 (cells were treated with Resiquimod (5 μg/ml, Invivogen) for 24 h. At the indicated timepoints, culture supernatants were harvested for IL-1β and IP-10 ELISA, and cells were lysed for RNA or protein analysis.

Quantitative RT-PCR and viral RNA analysis

At 48−72 h post-infection, cells were harvested for RNA isolation using RNeasy RNA isolation kits (QIAGEN, catalog # 74104) and cDNA was synthesized using oligo dT primers and Superscript III RT (Invitrogen, catalog #18080-051). cDNA corresponding to 100 ng of RNA was analyzed by qRT-PCR using SYBR green (Thermo Scientific, catalog # FERK0241) to quantify host mRNA levels and HIV transcripts using the primer sets listed in Table 1. The data were normalized to input (for immunoprecipitation) or GAPDH control (as infection control). The C_T value was normalized to that of GAPDH and represented as a relative value to mock or drug control, where indicated, using the 2^−∆∆C_T method as previously described [125,126].

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. qRT-PCR primers.

https://doi.org/10.1371/journal.pbio.3003320.t001

ELISA

Secreted IL-1β or IP-10 in culture supernatants was measured using a commercially available Human IL-1 beta/IL-1F2 DuoSet ELISA kit (R&D Systems, catalog #DY201) or Human IP-10 ELISA Set (BD Biosciences), respectively, according to directions provided by the manufacturer. HIV-1 p24^Gag levels in the culture supernatants were assessed using an in-house p24^Gag ELISA [56].

Flow cytometry

Cells were analyzed using BD LSRII or Cytek Aurora instruments with the help of Boston University Flow Cytometry Core who provided instrumentation and technical support. Data analysis was performed using FlowJo software (FlowJo). For cell surface antigen staining, cells were pre-chilled at 4°C for 30 min prior to staining. Antibody solution was added at indicated dilution in 2% NCS (normal calf serum, Invitrogen, catalog #26170043) in PBS (catalog #14190-250) for 30 min at 4°C. Cells were washed once with 2% NCS/PBS and fixed in 4% paraformaldehyde (PFA) in PBS (Boston Bioproducts) for at least 30 min at 4°C. For intracellular antigen staining, cells were fixed with 4% PFA, and permeabilized using Perm/Wash Buffer (BD Biosciences, catalog #554723) for at least 15 min at RT. Cells were stained at the indicated dilution of antibody in Perm/Wash buffer for 30 min at 4°C. Cells were washed once with Perm/Wash and resuspended in 2% NCS/PBS.

Antibodies

To characterize iMGs following differentiation, the following antibodies were used; mouse anti-CD11b (BioLegend; #301410; 1:20), mouse anti-CD45 (BD Biosciences; #555483; 1:10), mouse anti-P2RY12 (Biolegend, #392105, 1:50), mouse anti-TREM2 (Biolegend, #824803, 1:50). Intracellular p24^Gag was stained using FITC-conjugated mouse anti-p24^Gag monoclonal antibody (KC57; Beckman Coulter, catalog # 6604665, 1:25) as previously described [126]. For all experiments, cell viability was assessed using Zombie NIR Fixable Viability Kit (Biolegend, #423106, 1:250).

Imaging

For MDMs and THP-1, cells were cultured and infected on coverslips (LabTekII) and were washed and fixed with 4% PFA. Cells were then permeabilized with 0.1% TritonX100 in PBS, and stained as follows. Permeabilized cells were incubated overnight at 4°C with rabbit polyclonal antibody directed against C-terminal NLRP1 (Sigma, #HPA064431) at 1:500 dilution, followed by secondary antibody staining using goat anti-rabbit-AF568 (Invitrogen, #A21069, 1:200) for 1 h at RT. For dsRNA staining [127], a mouse anti-dsRNA (clone J2, Cell Signaling, #76651) antibody was used at 1:500 dilution overnight, followed by goat anti-mouse-AF647 (Invitrogen, #21237, 1:200) for visualization. For nuclear staining, 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) was used at 200 ng/ml for 15 min at RT. After each step, cells were washed three times in PBS. Images were acquired using a Leica SP5 confocal microscope and analyzed using ImageJ software.

Immunoblot analysis

To assess expression of host and viral proteins, cell lysates containing 15–30 µg total protein were separated by SDS-PAGE, transferred to nitrocellulose membranes, and the membranes were blocked with a 1:1 mix of PBS/Li-Cor Odyssey Blocking Buffer (Li-Cor, catalog # 927-70001) in PBS (Invitrogen, catalog #14190-250) for 1–2 h at RT. Blots were then probed with the following primary antibodies overnight at 4°C; human HIV-Ig (NIH AIDS reagent program, #3957, 1:1000), mouse anti-SAMHD1 (Abcam, # ab67820, 1:1000), rabbit anti-phospho-SAMHD1 (Cell Signaling, #15,038, 1:1000), mouse anti-NLRP1 (Biolegend, #6709802, 1:1000), rabbit anti-caspase-1 (Sigma, #06503-I, 1:1000), rabbit anti-CARD8 (Abcam, #ab194585, 1:1000), rabbit anti-MAVS (Invitrogen, #PA5-17256, 1:1000), rabbit anti-STING (Cell Signaling, #13647, 1:1000). Blots were washed 3 times for 5 min with PBS-T (0.05% Tween-20 in PBS). Specific staining was visualized with secondary antibodies, goat anti-mouse-IgG-DyLight 680 (Invitrogen, #35518, 1:20000), goat anti-rabbit-IgG-DyLight 800 (Invitrogen, #SA5-35571, 1:20000), or goat anti-human DyLight 800 (Rockland, #609-145-123) for 1–3 h at RT. As loading controls, β-actin was probed using mouse anti-β-actin (Invitrogen, #AM4302, 1:5000). Membranes were scanned with an Odessy scanner (Li-Cor).

IFN bioassay

Levels of bioactive type I IFN secreted from HIV-1-infected macrophages and microglia were measured using a HEK293 ISRE-luc reporter cell line [126,128]. Briefly, HEK293 ISRE-luc cells (8 × 10⁴) were incubated with cell culture supernatants for 21 h. Cells were lysed in BrightGlo lysis buffer (Promega, catalog #E2661), and luciferase activity in the cell lysates was analyzed with Bright-Glo Luciferase System (Promega, catalog #E2620). Serial dilutions of recombinant IFN_⍺2 ranging from 200 to 0.39 units/ml (PBL Interferon Source, catalog #11100-1) were used for generating a standard curve.

LDH release assay

Release of LDH in culture medium was quantified using the CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega, #PR-G1780) according to the manufacturer’s instructions. All samples were lysed in 0.5% Triton X-100, and LDH release was normalized to mock (uninfected or unstimulated) controls.

FLICA assay

PMA-differentiated THP1 cells were either treated with virus-free media (Mock) or infected with Lai∆envGFP/G WT (MOI = 1) in the absence (DMSO) or presence of RT inhibitor, Efavirenz (EFV, 1 μM) for 3 days. As positive control, cells were transfected with lipopolysaccharide (LPS, 1 μg) for 24 h, followed by treatment with Nigericin (Ngc, 20 μM) for 30 min. Cells were then stained with FLICA Caspase-1 (ImmunoChemistry Technologies, #9,122) for 1 h, following manufacturer’s instructions. Cells were then collected, washed three times, fixed with 4% paraformaldehyde, and analyzed by flow cytometry. Percent infection was quantified based on GFP expression, and caspase-1 activation determined by YVAD-660 expression.

RNA-immunoprecipitation (RNA-IP)

Infected cells were washed twice with PBS, and where indicated, followed by cellular fractionation using PARIS kit (Invitrogen, #AM192), according to the manufacturer’s guide. Whole cell lysates or cytoplasmic fractions were immunoprecipitated following a recently described protocol [129]. Briefly, protein A Dynabeads (Invitrogen) were conjugated with human anti-NLRP1 antibody (Novus Biologicals, #NB100-56148SS, diluted at 10 μg per 50 μL beads) for 1 hour while rotating at RT. Unfractionated cells were lysed in NET-2 buffer (50 mM Tris-HCl, 150 mM NaCl, 0.05% NP-40), and fractionated samples were lysed using cellular fractionation buffer supplied in PARIS kit. After bead preparation, NLRP1-conjugated beads were incubated with cell lysates at 4°C for 2 h while rotating. This is followed by magnetic bead separation, and washing thrice in NET-2 buffer. After RNA extraction using the TRIzol reagent (Invitrogen), RNA samples were treated with DNase using Turbo DNA-free kit (Invitrogen, #AM1907), and cDNA synthesis, followed by qRT-PCR was carried out as described above.

smFISH

Statistics

All the statistical analysis was performed using GraphPad Prism 9. P-values were calculated using one-way ANOVA followed by the Tukey–Kramer post-test or the Dunnett’s post-test. Symbols represent, *: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p < 0.0001. No symbol: not significant (p ≥ 0.05).