Leaked mitochondrial C1QBP inhibits activation of the DNA sensor cGAS

Abstract

Cytosolic DNA from pathogens activates the DNA sensor, cyclic GMP-AMP synthase (cGAS) that produces the second messenger, cyclic GMP-AMP (cGAMP). cGAMP triggers a signal cascade leading to type I interferon (IFN) expression. Host DNA is normally restricted in the cellular compartments of the nucleus and mitochondria. Recent studies have shown that DNA virus infection triggers mitochondrial stress, leading to the release of mitochondrial DNA (mtDNA) to the cytosol and activation of cGAS; however, the regulatory mechanism of mtDNA-mediated cGAS activation is not well elucidated. In this study, we analyzed cGAS protein interactome in mouse RAW264.7 macrophages and found that cGAS interacted with C1QBP. C1QBP is predominantly localized in the mitochondria and leaked into the cytosol during DNA virus infection. The leaked C1QBP bound the NTase domain of cGAS and inhibited cGAS enzymatic activity in cells and in vitro. Overexpression of the cytosolic form of C1QBP inhibited cytosolic DNA-elicited innate immune responses and promoted HSV-1 infection. By contrast, deficiency of C1QBP led to the elevated innate immune responses and impaired HSV-1 infection. Taken together, our study suggests that C1QBP is a novel cGAS inhibitor hidden in the mitochondria.

INTRODUCTION

Cytosolic DNA from infectious microbes triggers innate host defense by activating type I interferon (IFN) expression (1–4). The DNA sensor cyclic GMP-AMP synthase (cGAS, also known as MB21D1) produces cyclic GMP-AMP (cGAMP) once engaging with cytosolic DNA (5). cGAMP is a second messenger that binds to the endoplasmic reticulum membrane protein, stimulator of interferon genes (STING, also known as TMEM173, MPYS, MITA, and ERIS), leading to STING dimerization (6–9). Subsequently, STING recruits TANK-binding kinase 1 (TBK1) to the endoplasmic reticulum and activates TBK1. Activated TBK1 phosphorylates interferon regulatory factors (IRFs), leading to IRF dimerization and nuclear translocation. In the nucleus, IRFs form active transcriptional complexes and activate type I IFN gene expression.

Excessive host DNA can activate the cGAS signaling pathway, leading to aberrant IFN activation and autoimmune diseases, such as Aicardi-Goutieres syndrome (AGS) (10–13). Therefore, cells must keep cGAS inert from host genomic DNA. Host DNA is normally restricted in the cellular compartments of the nucleus and mitochondria; however, mitochondrial DNA (mtDNA) may escape to cytoplasm due to cellular stresses caused by infection, oxidative stress, and apoptosis (14–16). Recent studies showed that the leaked mtDNA is sensed by cGAS and activates IFN expression (14–16). DNA viruses, like HSV-1, have been shown to trigger mitochondrial stress and release mtDNA into the cytosol (16), but whether this benefits the virus remains to be established (17).

Complement C1q binding protein (C1QBP, also known as P32, GC1qR, and HABP1) was first found to bind the C1 complement complex on the cell surface. Following studies showed that C1QBP is ubiquitous and abundantly expressed in the mitochondrial matrix and moderately expressed on the cell surface, cytosol, and nucleus (18–25). C1QBP participates in several biological processes, such as ribosome biogenesis, oxidative stress, autophagy, regulation of apoptosis, and transcriptional regulation (26–30). C1QBP also regulates host defense and pathogen infection. Several DNA viruses, such AdV, EBV, HSV-1 and CMV, exploit C1QBP to facilitate viral infection (19, 31–37), implying a general mechanism. C1QBP may directly or indirectly affect DNA virus infection and pathogenesis. However, how C1QBP regulates DNA virus infection is not clear.

In this study, we first analyzed the cGAS protein complex in macrophages by proteomics and identified C1QBP as a novel cGAS interactor. We further found that C1QBP leaked from the mitochondria to the cytoplasm after HSV-1 infection. More importantly, the leaked C1QBP inhibited cGAS activation in cells and in vitro and facilitated viral infection. Collectively, we identify cytosolic C1QBP as an intrinsic regulator of the cGAS signaling pathway.

MATERIALS AND METHODS

Cells.

HEK293 cells (ATCC, # CRL-1573), RAW 264.7 (ATCC, # TIB-71), and Vero cells (ATCC, # CCL-81), were maintained in Dulbecco’s Modified Eagle Medium (Life Technologies, # 11995–065) containing antibiotics (Life Technologies, # 15140–122) and 10% fetal bovine serum (Life Technologies, # 26140–079). A549 cells (ATCC, # CCL-185) were cultured in RPMI Medium 1640 (Life Technologies, # 11875–093) plus 10% fetal bovine serum and 1 × MEM Non-Essential Amino Acids Solution (Life Technologies, # 11140–050).

Viruses.

HSV-1 KOS strain was purchased from ATCC (#VR-1493). HSV-1-GFP and VACV-Luc were reported before (38, 39). Viral titration was performed as described in a previous study (40). Briefly, Vero cells were infected with a serial diluted HSV-1. After 1 h, the medium was removed and replaced by the DMEM plus 5% FBS and 1% agarose. After 3 d, cells were examined for cytopathic effects to determine TCID₅₀ or were fixed using the methanol–acetic acid (3:1) fixative and stained using a Coomassie blue solution to determine the multiplicity of infection (MOI).

Plasmids.

Human C1QBP cDNA was synthesized and cloned into pCMV-3Tag-8 to generate C1QBP-FLAG and C1QBP-HA. Human cGAS cDNA was also cloned into pCMV-3Tag-8 to produce cGAS-FLAG and cGAS-HA. Point mutations and deletions of C1QBP and cGAS were constructed using a Q5® Site-Directed Mutagenesis Kit (New England Biolabs, # E0554S).

Antibodies.

Primary antibodies: Anti-α-tubulin (Cell Signaling Technology, # 2144), anti-FLAG (Sigma, # F3165), anti-HA (Cell Signaling Technology, # 3724), anti-TBK1 (Cell Signaling Technology, # 3504S], anti-phospho-TBK1 (Ser172) (Cell Signaling Technology, # 5483S), anti-GST (Cell Signaling Technology, # 2624S), anti-His (Thermo Fisher Scientific, # MA1–21315), anti-human cGAS (Signaling Technology, # 15102S), anti-mouse cGAS (Signaling Technology, # 31659), and anti-C1QBP (Cell Signaling Technology, # 6502S).

Secondary antibodies: Goat anti-Mouse IgG-HRP [Bethyl Laboratories, # A90–116P, WB (1:10,000)], Goat anti-Rabbit IgG-HRP [Bethyl Laboratories, # A120–201P, WB (1:10,000)], Alexa Fluor 594 Goat Anti-Mouse IgG (H+L) [Life Technologies, # A11005, IFA (1:200)], Alexa Fluor 488 Goat Anti-Rabbit IgG (H+L) [Life Technologies, # A11034, IFA (1:200)].

Sample preparation, Western blotting, and immunoprecipitation.

Approximately 1 X 10⁶ cells were lysed in 500 µl of tandem affinity purification (TAP) lysis buffer [50 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 100 mM NaCl, 0.5% Nonidet P40, 10% glycerol, the Complete EDTA-free protease inhibitor cocktail tablets (Roche, # 11873580001)] for 30 min at 4 °C. The lysates were then centrifuged for 30 min at 14,000 rpm. Supernatants were collected and mixed with the Lane Marker Reducing Sample Buffer (Thermo Fisher Scientific, # 39000).

Western blotting and immunoprecipitation were performed as described in a previous study (41). Briefly, samples (10–15 μl) were loaded into Mini-Protean TGX Precast Gels, 15 well (Bio-Rad, # 456–103), and run in 1 × Tris/Glycine/SDS Buffer (Bio-Rad, # 161–0732) for 60 min at 140 V. Protein samples were transferred to Immun-Blot PVDF Membranes (Bio-Rad, # 162–0177) in 1 × Tris/Glycine Buffer (Bio-Rad, # 161–0734) at 70 V for 60 min. PVDF membranes were blocked in 1 × TBS buffer (Bio-Rad, # 170–6435) containing 5% Blotting-Grade Blocker (Bio-Rad, # 170–6404) for 1 h. After washing with 1 × TBS buffer for 30 min, the membrane blot was incubated with the appropriately diluted primary antibody in antibody dilution buffer (1 × TBS, 5% BSA, 0.02% sodium azide) at 4 °C for 16 h. Then, the blot was washed three times with 1 × TBS (each time for 10 min) and incubated with secondary HRP-conjugated antibody in antibody dilution buffer (1:10,000 dilution) at room temperature for 1 h. After three washes with 1 × TBS (each time for 10 min), the blot was incubated with Clarity Western ECL Substrate (Bio-Rad, # 170–5060) for 1–2 min. The membrane was removed from the substrates and then exposed to the Amersham imager 600 (GE Healthcare Life Sciences, Marlborough, MA).

For immunoprecipitation, 2% of cell lysates were saved as an input control, and the remainder was incubated with 5–10 μl of the indicated antibody plus 20 μl of Pierce Protein A/G Plus Agarose (Thermo Fisher Scientific, # 20423) or 10 μl of EZview Red Anti-FLAG M2 Affinity Gel (Sigma, # F2426). After mixing end-over-end at 4 °C overnight, the beads were washed 3 times (5 min each wash) with 500 μl of lysis buffer. All co-IP experiments were performed by transfection into HEK293 cells.

Protein purification from E. coli and pull-down assays.

The dMTS mutant was cloned into pGEX-5X-3 (GE Healthcare, # 28–9545-55) to fuse with a GST tag. cGAS was cloned into pET28b(+) (Novagen, # 69865–3) to fuse with a His tag. These constructs were transformed into BL21 (DE3) E. coli (New England Biolabs, # C2527I) and cultured in LB broth at 20 °C. IPTG (0.4 mM) was added to induce protein expression. The GST Protein Interaction Pull-Down Kit (ThermoFisher Scientific, # PI21516) was used for GST-tagged protein purification and GST pull-down assays. The His-Spin Protein Miniprep kit (Zymo Research, # P2002) was used for His-tagged protein purification.

Immunofluorescence assay.

Cells were cultured in the Lab-Tek II CC2 Chamber Slide System 4-well (Thermo Fisher Scientific, # 154917). After the indicated treatment, the cells were fixed and permeabilized in cold methanol for 10 min at −20 °C. Then, the slides were washed with 1 × PBS for 10 min and blocked with Odyssey Blocking Buffer (LI-COR Biosciences, # 927–40000) for 1 h. The slides were incubated in Odyssey Blocking Buffer with appropriately diluted primary antibodies at 4 °C for 16 h. After 3 washes (10 min per wash) with 1 × PBS, the cells were incubated with the corresponding Alexa Fluor conjugated secondary antibodies (Life Technologies) for 1 h at room temperature. The slides were washed three times (10 min each time) with 1 × PBS and counterstained with 300 nM DAPI for 1 min, followed by washing with 1 × PBS for 1 min. After air-drying, the slides were sealed with Gold Seal Cover Glass (Electron Microscopy Sciences, # 3223) using Fluoro-gel (Electron Microscopy Sciences, # 17985–10). Images were captured and analyzed using a Revolve Microscope (Discover Echo Inc.). The quantitated colocalization was determined by fluorescence intensity plot and the Coloc2 function of ImageJ.

Real-time PCR.

Total RNA was prepared using the RNeasy Mini Kit (Qiagen, # 74106). 500 ng of RNA was reverse transcribed into cDNA using the QuantiTect reverse transcription kit (Qiagen, # 205311). For one real-time reaction, 10 µl of SYBR Green PCR reaction mix (Eurogentec) including 100 ng of the synthesized cDNA plus an appropriate oligonucleotide primer pair were analyzed on a 7500 Fast Real-time PCR System (Applied Biosystems). The comparative Ct method was used to determine the relative mRNA expression of genes normalized by the housekeeping gene GAPDH. The primer sequences: mouse Gapdh, forward primer 5`- GCGGCACGTCAGATCCA −3`, reverse primer 5`- CATGGCCTTCCGTGTTCCTA −3`; mouse Ifnb1, forward primer 5`- CAGCTCCAAGAAAGGACGAAC −3`, reverse primer 5`- GGCAGTGTAACTCTTCTGCAT −3`; mouse Cxcl10 (Ip10), forward primer 5`- CCAAGTGCTGCCGTCATTTTC −3`, reverse primer 5`- GGCTCGCAGGGATGATTTCAA −3`; mouse Ccl5 (Rantes), forward primer 5`- GCTGCTTTGCCTACCTCTCC −3`, reverse primer 5`-TCGAGTGACAAACACGACTGC-3`; HSV-1 VP16, forward primer 5`- GGACTGTATTCCAGCTTCAC −3`, reverse primer 5`- CGTCCTCGCCGTCTAAGTG −3`.

Plasmid transfection.

HEK293 and A549 cells were transfected using Lipofectamine 3000 or Lipofectamine LTX Transfection Reagent (Life Technologies, # L3000015) according to the manufacturer’s protocol.

CRISPR/Cas9.

The single guide RNA (sgRNA) targeting sequences: mouse C1qbp sgRNA: 5`- CGTACGCTGAGCAAACCGAA −3`. The sgRNA was cloned into lentiCRISPR v2 vector (42) (Addgene). The lentiviral construct was transfected with psPAX2 and pMD2G into HEK293T cells using PEI. After 48 h, the media containing lentivirus were collected. The targeted cells were infected with the media containing the lentivirus supplemented with 10 μg/ml polybrene. Cells were selected with 10 μg/ml puromycin for 14 days. Single clones were expanded for knockout confirmation by Western blotting.

Purification of protein complexes

Affinity purification coupled with mass spectrometry experiments were performed as previously described (43). For protein purification, RAW264.7 cell lines stably expressing FLAG-tagged cGAS were collected and lysed in 10 ml of TAP buffer (44). Cell lysates were precleared with 50 μl of protein A/G resin before the addition of 20 μl of anti-FLAG resin (Sigma, # F2426) and incubation for 16 hr at 4°C on a rotator. The resin was washed 3 times and transferred to a spin column with 40 μl of 3 X FLAG peptide for 1 hr at 4°C on a rotator. The purified complexes were loaded onto a 4–15% NuPAGE gel. The gels were stained with a SilverQuest staining kit (Invitrogen), and lanes were excised for mass spectrometry analysis by the Taplin Biological Mass Spectrometry Facility (Harvard Medical School, Boston, MA).

Mass spectrometry

Samples were reconstituted in 5 – 10 µl of HPLC solvent A (2.5% acetonitrile, 0.1% formic acid). A nano-scale reverse-phase HPLC capillary column was created by packing 5 µm C18 spherical silica beads into a fused silica capillary (100 µm inner diameter x ~12 cm length) with a flame-drawn tip. After equilibrating the column each sample was loaded via a Famos auto sampler (LC Packings, San Francisco CA) onto the column. A gradient was formed and peptides were eluted with increasing concentrations of solvent B (97.5% acetonitrile, 0.1% formic acid). As peptides eluted, they were subjected to electrospray ionization and then entered into an LTQ Velos ion-trap mass spectrometer (Thermo Fisher, San Jose, CA). Peptides were detected, isolated, and fragmented to produce a tandem mass spectrum of specific fragment ions for each peptide. Dynamic exclusion was enabled such that ions were excluded from reanalysis for 30 s. Peptide sequences (and hence protein identity) were determined by matching protein databases with the acquired fragmentation pattern by the software program, Sequest (Thermo Fisher, San Jose, CA). The human IPI database (Ver. 3.6) was used for searching. Precursor mass tolerance was set to +/− 2.0 Da and MS/MS tolerance was set to 1.0 Da. A reversed-sequence database was used to set the false discovery rate at 1%. Filtering was performed using the Sequest primary score, Xcorr and delta-Corr. Spectral matches were further manually examined and multiple identified peptides (>=2) per protein were required.

SAINT analysis of AP-MS data

Four biological repeats were performed for each protein complex. The resulting data are presented in Supplemental Table 1. Proteins found in the control group were considered as non-specific binding proteins. The SAINT algorithm (http://sourceforge.net/projects/saint-apms) was used to evaluate the MS data (45). The default SAINT options were low Mode = 1, min Fold = 0, and norm = 0. The SAINT scores computed for each biological replicate were averaged (AvgP) and reported as the final SAINT score. A SAINT score of AvgP ≥0.89 was considered a true positive BioID protein with an estimated FDR of ≤2%. Proteins with SAINT score <0.89 are considered as non-specific binding proteins. We manually removed ribosomal proteins from the final HCIP list because these proteins are prone to associate with RNA-binding proteins.

In vitro cGAS enzymatic assays

Purified recombinant cGAS-His was mixed with ctDNA, ATP, GTP, and different amounts of dMTS-GST in vitro at 37 °C for 1 h. Then, the production of cGAMP was determined by the cGAMP ELISA kit (Cayman Chemical).

Statistical analysis.

The sample size was sufficient for data analyses. Data were statistically analyzed using the software GraphPad Prism 9. Significant differences between the indicated pairs of conditions are shown by asterisks (* P <0.05; ** P <0.01; *** P <0.001; **** P <0.0001).

RESULTS

Proteomic analysis of cGAS protein complex in macrophages.

To seek new regulators of cGAS, we performed affinity purification coupled with mass spectrometry (AP-MS) (46) to analyze the cGAS protein interaction network. We examined cGAS complexes in RAW264.7 macrophages because this cell line is physiologically relevant for the study of innate immune responses. Mouse cGAS tagged with a FLAG epitope was transfected into RAW264.7 macrophages to generate stable cell lines. After we obtained the stable cell line by puromycin selection, cells were stimulated with or without calf thymus DNA (ctDNA). The cGAS protein complexes were purified using an anti-FLAG antibody and analyzed by mass spectrometry (Fig. 1A). The AP-MS was biologically repeated twice. To efficiently reduce false positives in AP-MS, we adopted the well-established statistical method SAINT (45) with a database of 24 protein complexes purified from RAW264.7 cells (Supplemental Table 1). Using a stringent statistical SAINT score cutoff of 0.89 (P < 0.01), we identified five high-confidence candidate interacting proteins (HCIPs), including histones and IFI202 (Fig. 1B). It is well-established that the histone-containing nucleosomes bind cGAS (47–51). IFI202 is one of the mouse homologs of human IFI16 that is implicated in the DNA sensing pathway and interacts with cGAS (52). These known interactions substantiate the high quality of our cGAS protein interaction network. We also found two new interactions with C1QBP and IMPDH2; however, these interactions are independent of ctDNA stimulation. Next, we performed co-immunoprecipitation (co-IP) to validate the two new cGAS interactors. We transfected cGAS with C1QBP or IMPDH2 into HEK293 cells. Co-IP showed that cGAS interacted with C1QBP and IMPDH2 (Figs. 1C and 1D).

Fig. 1. Analysis of cGAS protein complex.

(A) Schematics of AP-MS for cGAS protein complex in RAW247.6 cells. (B) cGAS protein interaction network. Legends are indicated. HCIP: high-confidence candidate interacting proteins. (C) FLAG-tagged IMPDH2 (IMPDH2-FLAG) and HA-tagged cGAS (cGAS-HA) were co-transfected into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with anti-FLAG antibody, then blotted with the indicated antibodies. (D) FLAG-tagged cGAS (cGAS-FLAG) and HA-tagged C1QBP (C1QBP-HA) were co-transfected into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with anti-FLAG antibody, then blotted with the indicated antibodies.

Dissect C1QBP-cGAS protein interaction

We chose C1QBP for further investigation because of its pro-viral role in DNA virus infection. We first examined the interaction between endogenous C1QBP and cGAS. RAW264.7 macrophages were infected with HSV-1, then the cell lysates were immunoprecipitated with the anti-C1QBP antibody. As shown in Figure 2A, HSV-1 infection promoted the interaction between endogenous C1QBP and cGAS. Next, we wanted to determine which form of C1QBP was responsible for cGAS interaction. It is well-known that the N-terminal 73 residues of C1QBP contain a mitochondrial targeting signal (MTS) (53) (Fig. 2B). Full-length C1QBP migrates at approximately the same position as the MTS deletion mutant (dMTS) when expressed in mammalian cells (Fig. 2C). Co-IP found that C1QBP and the dMTS mutant interacted with cGAS (Fig. 2C). We also expressed and purified GST-tagged dMTS and His-tagged cGAS from bacteria. GST pull-down assays found that the dMTS mutant interacted with cGAS in vitro (Fig. 2D). Next, we determined the domains in C1QBP responsible for cGAS interaction. We created a panel of C1QBP mutants (Fig. 2E). Co-IP showed that the region of amino acids 74–220 in C1QBP was sufficient for cGAS binding (Fig. 2F). Overall, these data suggest that HSV-1 infection facilitates C1QBP-cGAS interaction and the 74–220 region of C1QBP is responsible for the interaction with cGAS.

Fig. 2. C1QBP interacts with cGAS.

(A) RAW264.7 macrophages were infected with 1 MOI of HSV-1 for 16 h. Then, the cell lysates were subjected to immunoprecipitation and immunoblotting with the indicated antibodies to detect endogenous protein interactions. (B) Schematic of wild type C1QBP protein and the mutant lacking the mitochondrial targeting signal (MTS). (C) cGAS-HA was co-transfected with FLAG-tagged C1QBP or the dMTS mutant into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with the anti-FLAG antibody, then blotted with the indicated antibodies. (D) Purified recombinant His-tagged cGAS (cGAS-His) was mixed with GST or GST-tagged dMTS (dMTS-GST) in vitro at 4 °C for 16 h. Then, GST pull-down was performed and blotted as indicated. (E) Schematic of C1QBP truncation mutants. (F) cGAS-HA was co-transfected with the indicated FLAG-tagged C1QBP mutants into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with anti-FLAG antibody, then blotted with the indicated antibodies.

HSV-1 infection causes the mitochondrial leakage of C1QBP and co-localization with cGAS in the cytoplasm

To examine subcellular localization of cGAS and C1QBP, we transfected FLAG-tagged cGAS, C1QBP or the dMTS mutant into A549 cells. Immunofluorescence assays (IFA) found that full-length C1QBP showed mitochondrial localization, whereas the dMTS mutant was mainly expressed in the cytoplasm (Fig. 3A). cGAS co-localized with the dMTS mutant but not full-length C1QBP in the cytoplasm (Fig. 3B). Furthermore, we examined the subcellular localization of cGAS and C1QBP in A549 cells after HSV-1 infection. As shown in Figure 3C, HSV-1 infection caused the increased cytosolic localization of C1QBP. By contrast, transfection of calf thymus DNA (ctDNA) had little effect on the mitochondrial localization of C1QBP (Fig. 3C). Further subcellular fractionation corroborated that HSV-1 infection induced mitochondrial leakage of C1QBP in RAW264.7 macrophages (Fig. 3D). Interestingly, the RNA virus vesicular stomatitis virus (VSV) also caused C1QBP mitochondrial leakage (Fig. 3E). Taken together, our data suggest that viral infection causes the mitochondrial leakage of C1QBP and promotes the cytosolic interaction between C1QBP and cGAS.

Fig. 3. C1QBP leaks from the mitochondria during HSV-1 infection and co-localized with cGAS in the cytosol.

(A) A549 cells were transfected with FLAG-tagged C1QBP, dMTS, or cGAS. After 48 h, cells were stained with MitoTracker CMXRos (red), anti-FLAG (green), and DAPI nuclear stain (blue). The red and green fluorescence intensity was plotted according to the white line in the merged image. The quantitated colocalization was determined by Pearson correlation coefficient (Pearson’s r) and shown on next to the fluorescence intensity plot. Thirty cells (ten cells from each experiment) were analyzed by ImageJ to quantify the colocalization between green and red channel. (B) cGAS-HA was transfected with FLAG-tagged C1QBP or the dMTS mutant into A549 cells. After 48 h, cells were fixed and stained with anti-FLAG (red), anti-HA (green), and DAPI nuclear stain (blue). The red and green fluorescence intensity was plotted according to the white line in the merged image. The quantitated colocalization was determined by Pearson’s r and shown on next to the fluorescence intensity plot. Thirty cells (ten cells from each experiment) were analyzed by ImageJ to quantify the colocalization between green and red channel. (C) A549 cells were mock-infected, infected with HSV-1 for 8 h, or stimulated with ctDNA for 4 h. Then, cells were stained with MitoTracker CMXRos (red), anti-C1QBP (green), and DAPI nuclear stain (blue). Arrows indicate the leaked cytosolic C1QBP proteins that are not localized in the mitochondria in the inset of boxed area. The red and green fluorescence intensity was plotted according to the white line in the inset image. The quantitated colocalization was determined by Pearson’s r and shown on next to the fluorescence intensity plot. Thirty cells (ten cells from each experiment) were analyzed by ImageJ to quantify the colocalization between green and red channel. (D) Subcellular fractionation of RAW264.7 cells infected with and without HSV-1 for 8 h. Cell lysates were blotted as indicated. (E) Subcellular fractionation of RAW264.7 cells infected with and without VSV for 8 h. Cell lysates were blotted as indicated.

Cytosolic C1QBP suppresses cGAS-mediated innate immune response.

We hypothesized that the leaked C1QBP inhibits cGAS activation as the majority of C1QBP localizes in the mitochondria while cGAS resides in the cytoplasm and nucleus. To test the hypothesis, we first examined the effect of the dMTS mutant, which lacks the mitochondrial targeting signal and represents a leaked form of C1QBP, on cGAS-mediated ISRE reporter activity. Different doses of C1QBP and the dMTS mutant were transfected into HEK293 cells together with an ISRE reporter. As expected, the dMTS mutant, but not the full-length C1QBP (residing at the mitochondria), inhibited cGAS-induced reporter activity in a dose-dependent manner (Fig. 4A). To further examine the role of C1QBP in cGAS-mediated innate immune response, we overexpressed the dMTS mutant in RAW264.7 macrophages. The dMTS mutant impaired TBK1 phosphorylation induced by HSV-1 (Fig. 4B). Consistently, it also inhibited the mRNA expression of IFNβ, IP-10, and RANTES induced by HSV-1 (Figs. 4C–4E). Furthermore, the dMTS mutant inhibited ctDNA-induced IFNβ mRNA expression (Fig. 4F).

Fig. 4. Cytosolic C1QBP inhibits cGAS-mediated innate immune responses and facilitates HSV-1 infection.

(A) 200 ng of cGAS was transfected with different doses (50, 100, and 200 ng) of C1QBP or the dMTS mutant together with 20 ng of pRL-SV40 (Renilla luciferase as an internal control) and 200 ng of pISRE-Luc into HEK293 cells. After 48 h, cells were collected, and the ratio of firefly luciferase to Renilla luciferase was calculated to determine the relative reporter activity. All experiments were repeated three times. The P value was calculated by one-way ANOVA followed by Dunnett’s multiple comparisons test (** P < 0.01, **** P < 0.0001). (B) RAW264.7 macrophages stably transfected with vector and dMTS were infected with 1 MOI of HSV-1 for indicated times. Cell lysates were collected and blotted as indicated. Band densitometry was calculated by Image J. The ratio of phosphorylated TBK1 to total TBK1 in each lane was indicated. (C–E) RAW264.7 macrophages stably transfected with vector and dMTS were infected with 1 MOI of HSV-1 for 8h. Real-time PCR was performed to determine the relative mRNA levels of IFNβ (C), IP-10 (D), and RANTES(E). All experiments were biologically repeated three times. Data represent means ±s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (* P < 0.05, ** P < 0.01). (F) RAW 264.7 macrophages stably transfected with vector and the dMTS mutant were stimulated with 1 μg/ml ctDNA for 4h. Real-time PCR was performed to determine the relative IFNβ mRNA levels. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (* P < 0.05).

Cytosolic C1QBP increases host susceptibility to viral infection

We examined the effects of the dMTS mutant on viral infection. Wild type RAW264.7 cells and cells expressing the dMTS mutant were infected with HSV-1 carrying GFP (HSV-1-GFP) for 16 h. As shown in Figure 5A, the C1QBP dMTS mutant promoted viral infection activity evidenced by the increased GFP positive cells. However, the dMTS failed to increase HSV-1 infection in cGAS knockout RAW264.7 macrophages (Fig. 5A), suggesting that the C1QBP dMTS mutant regulates viral infection via cGAS. Furthermore, overexpression of dMTS increased viral mRNA expression (Fig. 5B) and the production of viral particles (Fig. 5C). Interestingly, the C1QBP dMTS mutant also enhanced the infection of a vaccinia reporter virus carrying a luciferase gene (Fig. 5D). The combined data suggest that the leaked cytosolic C1QBP inhibits cGAS activity and facilitates DNA virus infection.

Fig. 5. Cytosolic C1QBP facilitates DNA virus infection.

(A) Wild type and cGAS knockout (KO) RAW264.7 macrophages stably expressing vector or dMTS were infected with 5 MOI of HSV-1-GFP for 16 h. Cells expressing GFP were examined and counted under a fluorescence microscope. The relative infection was determined by the ratio of positive cells. Data represent means ±s.d. of three independent experiments. (> 200 cells were counted in each field and five fields were counted per experiment). The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (**** P < 0.0001, ns: not significant). (B) RAW264.7 macrophages stably transfected with vector and dMTS were infected with the indicated MOI of HSV-1 for 16 h. Then, cells were collected for RNA extraction. The RNA levels of HSV-1 VP16 were determined by real-time PCR. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (* P < 0.05, ** P < 0.01). (C) RAW264.7 macrophages stably transfected with vector and dMTS were infected with 0.01 MOI of HSV-1 for the indicated days. Titers of culture supernatants containing HSV-1 were determined on Vero cells, and plaques were enumerated. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (* P < 0.05, ** P < 0.01, *** P < 0.001). (D) RAW264.7 macrophages stably transfected with vector and dMTS were infected with 1 MOI of VACV carrying a luciferase gene (VACV-Luc) for 16 h. Luciferase activities were measured to determine the relative infection activity. Data represent means ±s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Sidak’s multiple comparisons test (* P < 0.05, **** P < 0.0001).

Knockout of C1QBP sensitizes cGAS-mediated innate immune responses.

To examine the effect of C1QBP deficiency on cGAS-mediated innate immunity, we generated two C1QBP knockout RAW264.7 cell lines by CRISPR (Fig. 6A). We also used two cGAS knockout RAW264.7 cell lines as the controls (Fig. 6B). As predicted, ctDNA-induced IFNβ, RANTES, and IP10 mRNA expression was impaired in the cGAS knockout cell lines (Fig. 6C). However, ctDNA induced comparable mRNA expression of IFNβ, RANTES, and IP10 in C1QBP knockout cells (Fig. 6C), consistent with the finding ctDNA transfection has no effect on C1QBP leakage. Next, we examined the effect of C1QBP on HSV-1-induced innate immune responses in wild type, cGAS knockout cells, and C1QBP knockout cells. Real-time PCR found that deficiency of C1QBP increased the mRNA levels of IFNβ, IP-10, and RANTES induced by HSV-1 while ablation of cGAS abolished mRNA expression of these genes (Figs. 6D–6F). Consistently, TBK1 and IRF3 phosphorylation was also enhanced in C1QBP knockout cells (Fig. 6G), suggesting that knockout of C1QBP promotes HSV-1 infection-induced cGAS signaling. To further corroborate the role of C1QBP, we reconstituted the dMTS mutant in the C1QBP knockout macrophages (Fig. 6H). We found that the dMTS rescued the innate immune response phenotype in C1QBP knockout cell, reducing the mRNA expression of IFNβ, IP-10, and RANTES induced by HSV-1 (Fig. 6I–6K). Taken together, these data suggest that C1QBP deficiency increases cGAS-induced innate immune responses.

Fig. 6. C1QBP deficiency increases HSV-1-induced innate immune response.

(A) The lysates of wild type and two C1QBP KO RAW264.7 cell lines were blotted as indicated. (B) The lysates of wild type and two cGAS KO RAW264.7 cell lines were blotted as indicated. (C) RAW264.7 macrophages, two cGAS KO cell lines, and two C1QBP KO cell lines were stimulated with 1 μg/ml ctDNA for 4 h. Real-time PCR was performed to determine the relative mRNA levels of IFNβ, IP10, and RANTES. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Tukey’s multiple comparisons test (* P < 0.05). (D-F) RAW264.7 macrophages, two cGAS KO cell lines, and two C1QBP KO cell lines were infected with 1 MOI of HSV-1 for indicated times. Real-time PCR was performed to determine the relative mRNA levels of IFNβ (D), IP10 (E) and RANTES (F). Data represent means ±s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Tukey’s multiple comparisons test (* P < 0.05, ** P < 0.01). (G) C1QBP WT and KO cells were infected with 1 MOI of HSV-1 for indicated times. Cell lysates were collected and blotted as indicated. Band densitometry was calculated by Image J. The ratio of phosphorylated TBK1 to total TBK1 and phosphorylated IRF3 to total IRF3 in each lane was indicated. (H) The lysates of wild type RAW264.7 macrophages, cGAS KO, and C1QBP KO reconstituted with dMTS were blotted as indicated. (I-K) Wild type RAW264.7 macrophages, C1QBP KO cells, and C1QBP KO cells reconstituted with dMTS were infected with 1 MOI of HSV-1 for indicated times. Real-time PCR was performed to determine the relative mRNA levels of IFNβ (I), IP10 (J) and RANTES (K). All experiments were biologically repeated three times. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Tukey’s multiple comparisons test (* P < 0.05).

C1QBP deficiency impairs DNA virus infection.

To examine the effects of C1QBP deficiency on viral infection, we compared viral infection in wild type RAW264.7 cells, cGAS knockout, and C1QBP knockout cells. These cells were infected with HSV-1 carrying a GFP (HSV-1-GFP) for 16 h. As shown in Figure 7A, knockout of cGAS increased HSV-1 infection, but viral infection activity in C1QBP knockout cells was much lower than in wild type cells. C1QBP deficiency also reduced the expression of viral RNA (Fig. 7B) and the production of viral particles (Fig. 7C). Consistently, knockout of cGAS increased VACV infection (Fig. 7D). By contrast, ablation of C1QBP reduced VACV infection (Fig. 7D). Furthermore, reconstitution of dMTS restored HSV-1 and VACV infection in C1QBP knockout cells (Figs. 7E–7F). Taken together, these data suggest that C1QBP deficiency sensitizes cGAS activity and impairs infection with the HSV-1 and vaccinia DNA viruses.

Fig. 7. Knockout of C1QBP reduces DNA virus infection.

(A) RAW264.7 macrophages, two cGAS KO cell lines, and two cC1QBP KO cell lines were infected with 5 MOI of HSV-1-GFP for 16 h. Cells expressing GFP were examined and counted under a fluorescence microscope. The relative infection was determined by the ratio of positive cells. Data represent means ± s.d. of three independent experiments (> 200 cells were counted in each field and five fields were counted per experiment). The P value was calculated by one-way ANOVA followed by Dunnett’s multiple comparisons test (* P < 0.05, **** P < 0.0001). (B) C1QBP WT macrophages and two KO cell lines were infected with the indicated MOI of HSV-1 for 16 h. Then, cells were collected for RNA extraction. The RNA levels of HSV-1 VP16 were determined by real-time PCR. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Dunnett’s multiple comparisons test (* P < 0.05). (C) C1QBP WT macrophages and two KO cell lines were infected with 0.01 MOI of HSV-1 for the indicated days. Titers of culture supernatants containing HSV-1 were determined on Vero cells. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by two-way ANOVA followed by Dunnett’s multiple comparisons test (* P < 0.05). (D) RAW264.7 macrophages, two cGAS KO cell lines, and two C1QBP KO cell lines were infected with VACV-Luc for 16 h. Luciferase activities were measured to determine the relative infection activity. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by one-way ANOVA followed by Dunnett’s multiple comparisons test (* P < 0.05, **** P < 0.0001). (E) Wild type RAW264.7 macrophages, C1QBP KO cells, and C1QBP KO cells reconstituted with dMTS were infected with 5 MOI of HSV-1-GFP for 16 h. Cells expressing GFP were examined and counted under a fluorescence microscope. The relative infection was determined by the ratio of positive cells. Data represent means ± s.d. of three independent experiments (> 200 cells were counted in each field and five fields were counted per experiment). The P value was calculated by one-way ANOVA followed by Tukey’s multiple comparisons test (* P < 0.05, ** P < 0.01). (F) Wild type RAW264.7 macrophages, C1QBP KO cells, and C1QBP KO cells reconstituted with dMTS were infected with VACV-Luc for 16 h. Luciferase activities were measured to determine the relative infection activity. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by one-way ANOVA followed by Tukey’s multiple comparisons test (**** P < 0.0001).

C1QBP interacts with cGAS NTase domain and inhibits cGAS enzymatic activity

To determine the mechanism by which C1QBP inhibits cGAS, we examined which domain of cGAS interacted with C1QBP. We created a panel of cGAS mutants (Fig. 8A). Co-IP showed that the nucleotidyltransferase (NTase) domain of cGAS was required for the interaction with C1QBP (Fig. 8B). Furthermore, the NTase domain interacted with the 77–220 region of C1QBP (Fig. 8C), suggesting that C1QBP might interfere with cGAS enzymatic activity. Thus, we performed in vitro cGAS enzymatic assays. The in vitro assays found that the C1QBP dMTS mutant inhibited cGAMP production by cGAS (Fig. 8D), concluding that C1QBP inhibits cGAS enzymatic activity.

Fig. 8. C1QBP inhibits cGAS enzymatic activity.

(A) Schematic of cGAS mutants. NTase: Nucleotidyltransferase domain. (B) HA-tagged cGAS (cGAS-HA) or the indicated mutant was co-transfected with FLAG-tagged C1QBP (C1QBP-FLAG) into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with anti-FLAG antibody, then blotted with the indicated antibodies. (C) HA-tagged C1QBP mutant (aa 74–220) was co-transfected with the FLAG-tagged NTase domain of cGAS (NTase-FLAG) into HEK293 cells. After 48 h, cells were harvested and immunoprecipitated with anti-FLAG antibody, then blotted with the indicated antibodies. (D) In vitro cGAS enzymatic assays. Purified recombinant cGAS-His (10 nM) was mixed with ctDNA, ATP, GTP, and different amounts of dMTS-GST (5, 25, 50 nM) in vitro at 37 °C for 1 h. Then, the production of cGAMP was determined by ELISA. Data represent means ± s.d. of three independent experiments. All experiments were repeated three times. The P value was calculated by one-way ANOVA followed by Dunnett’s multiple comparisons test (* P < 0.05, *** P < 0.001, **** P < 0.0001). (E) Model of C1QBP-mediated cGAS inhibition.

DISCUSSION

In this study, we found that C1QBP negatively regulates cGAS-mediated innate immune responses. C1QBP predominantly localizes in the mitochondria; viral infection causes the release of C1QBP from the mitochondria to the cytosol. Subsequently, the leaked cytosolic C1QBP binds cGAS and inhibits cGAS activation (Fig. 8E). C1QBP is a multifunctional and multicompartmental protein (54). Many studies focus on mitochondrial, nuclear and secreted C1QBP. Cytosolic C1QBP is much less studied, and significant levels of cytoplasmic localization are generally observed under pathological conditions, such as viral infection. Our study showed that C1QBP leaked from the mitochondria during viral infection, and cytosolic C1QBP impaired cGAS activation. We found that C1QBP interacted with the NTase domain of cGAS and inhibited cGAS activity in a dose-dependent fashion in vitro, suggesting C1QBP might directly block cGAS activation sites. Interestingly, the mature form of C1QBP, i.e. the MTS deletion, is highly acidic with a calculated pI of 3.96. Thus, it is also plausible that C1QBP competes with DNA for cGAS binding. These possible mechanisms warrant further investigation.

Several RNA viruses utilize C1QBP to promote viral infection (55–57). For example, the Rubella virus capsid interacts with C1QBP, which leads to the promotion of viral replication (56, 57). Recently, C1QBP has been shown to suppress viral RNA-induced innate immunity by disruption of the interaction between retinoic acid-inducible gene I (RIG-I) and mitochondrial antiviral signaling protein (MAVS) (58). DNA viruses are also known to exploit C1QBP to facilitate viral infection. For example, C1QBP has been found to interact with several DNA virus proteins, including core protein V of adenovirus (AdV) (19), EBNA-1 of Epstein–Barr virus (EBV) (31, 37), P22 of hepatitis B virus(59), IE63 and open reading frame P (ORFP) of herpes simplex virus 1 (HSV-1) (32, 60) and the kinase pUL97 of cytomegalovirus (CMV) (33). pUL97 interacts with C1QBP to facilitate nuclear export of viral capsids (33). These studies imply that C1QBP may directly or indirectly affect DNA virus infection and pathogenesis. But how C1QBP regulates DNA virus infection is not clear, and whether C1QBP modulates DNA-mediated innate immunity is unknown. Our study proposes a general mechanism by which leaked C1QBP can suppress cGAS activation and impair type I IFN production, thereby facilitating DNA virus infection.

During infection, DNA viruses expose their DNA in the cytosol, which can activate cGAS-mediated innate immune response. To evade host innate immunity, several viral proteins have been found to engage with cGAS to sabotage host defense (61, 62). The ORF52 of KSHV and its homologs of MHV68, RRV and EBV have been shown to inhibit cytosolic DNA sensing by directly inhibiting cGAS enzymatic activity through a mechanism involving both cGAS binding and DNA binding (62). Recently, N-terminally truncated cytoplasmic isoforms of KSHV latency-associated nuclear antigen (LANA) interacts with cGAS and antagonizes cGAS signaling (61). The UL41 and VP22 of HSV-1 abrogate type I IFN production by degrading cGAS and inhibiting cGAS enzymatic activity, respectively (63, 64). In addition to viral DNA, leaked mtDNA also activates cGAS and induces host innate immune responses (14–16). For example, deficiency of the transcription factor A mitochondrial (TFAM) promotes mitochondrial stress and the leakage of mtDNA, resulting in activating cGAS and inducing a type I interferon response (65). DNA virus infection also triggers mitochondrial stress and the release of mtDNA to the cytosol (16). Interestingly, RNA viruses, such as dengue virus, also elicits a cGAS‐STING response due to the release of oxidized mtDNA into the cytosol caused by viral infection (66). Our study showed that DNA virus infection causes mitochondrial C1QBP leak to inhibit cGAS activation, adding another layer of complexity of viral infection and host defense.

Aberrant activation of the cGAS-STING pathway causes autoimmune diseases, including Aicardi–Goutières syndrome and certain forms of lupus, like systemic lupus erythematosus (10–13). Aberrant inflammation via cGAS-STING also contributes to aging-related neurodegenerative conditions, such as Parkinson’s disease and amyotrophic lateral sclerosis (ALS) (67–69). Mutations of STING cause a life-threatening autoinflammatory condition, termed STING-associated vasculopathy with onset in infancy (SAVI) (70, 71). Thus, the cGAS-STING pathway is not only critical in infectious diseases but also can be pathogenic in primary immune disorders when the pathway is aberrantly activated. mtDNA has been implicated to at least partially contribute to some autoimmune disorders by activating cGAS-STING pathway (72). It would be interesting to know whether C1QBP is also leaked during these conditions and whether the leaked C1QBP provides protection by reducing mtDNA-induced IFN production. Overall, our study will not only reveal a viral strategy to evade immune surveillance but also suggest a potential host protective mechanism that may modulate aberrant activation of innate immunity under pathological conditions.

KEY POINTS.

• HSV-1 infection causes the mitochondrial leakage of C1QBP.

• C1QBP inhibits cGAS enzymatic activity.

• C1QBP suppresses interferon production and facilitates viral infection.