In this study, we report that the TLR4 ligand, LPS, and TLR3 ligand polyinosinic:polycytidylic acid failed to activate IRF3 or STAT1 in bone marrow–derived macrophages (BMMs) isolated from two independently generated lines of Rosa26-integrated Cas9-expressing C57BL/6J (B6) mice. RNA-sequencing analysis reveals that hundreds to thousands of genes including IFN-stimulated genes were differentially expressed in BMMs from these Cas9 strains compared with B6 upon LPS stimulation. Furthermore, the NF-κB signaling axis and TRIF-mediated necroptosis were also strongly reduced in response to LPS and polyinosinic:polycytidylic acid. In contrast, there were no defects in the responses of BMMs to ligands of the RIG-I, STING, TLR2, TLR9, and IFN receptors. Defects in TLR3 and TLR4 signaling were observed in mice with the B6 but not 129 background, and when Cas9 was integrated at the Rosa26 but not H11 locus. However, integration at the Rosa26 site, CAG promoter–driven Cas9 or eGFP were not individually sufficient to cause the defect. Taken together, the results of this study suggest a putative TRIF-mediated defect in TLR-3/4 signaling in BMMs from commercially available and widely used B6–Cas9–expressing mice.

TLRs, especially the TLR4-mediated signaling system, has been studied extensively. Upon stimulation with bacterial LPS, TLR4 sends a signal through MyD88 and TIR domain–containing adaptor-inducing IFN-β (TRIF) (1, 2) by generating two supramolecular centers known as the “Myddosome” and the “Triffosome” (3, 4). One of the functions of the Myddosome assembly, which is composed of MyD88, TIRAP, IRAK enzymes, TRAF6, and recently identified TBK1, is to activate NF-κB through its downstream activators TAK1, IKKα/β, and IκBα. At steady-state condition, NF-κB binds to its inhibitor protein IκBα and stays in the cytoplasm. Following LPS stimulation, IκBα is heavily phosphorylated by its immediate upstream kinases, leading to its subsequent ubiquitination and proteasomal degradation, which allows NF-κB to translocate to the nucleus. Nuclear NF-κB then triggers expression of proinflammatory genes including those encoding the cytokines TNF-α and IL-6 (5, 6). In addition, the Myddosome also activates MAPK, which allows translocation of AP1 to the nucleus for maximum production of the cytokines.

Triffosome assembly starts with either LPS-induced internalization of TLR4 or with polyinosinic:polycytidylic acid [poly(I:C)]–induced TLR3 activation directly on the endosome membrane (4). The Triffosome sends signals through phosphorylated TRIF, TBK1, IKKε, and IRF3 for the production of type I IFN, which in turn activates STAT1 and regulates expression of IFN-stimulated gene products (ISGs) (6, 7). This TRIF-mediated signal is required not only for the production of type I IFN but also to keep oscillations and sustained activation of NF-κB (8). TRIF is also required for LPS or poly(I:C)-stimulated TRIF–RIPK3–mixed lineage kinase domain-like protein (MLKL)-mediated necroptosis upon caspase inhibition (9, 10).

TBK1 and IKKε can be activated to produce IFN by TRIF-independent pathogen sensing pathways. Sendai virus (SeV) or influenza virus are sensed intracellularly by a RIG-I–mediated pathway (11), whereas double-stranded DNA and bacterial cyclic dinucleotides are sensed by the STING pathway, which can also be activated by the small molecule DMXAA (12), leading to the production of IFN.

CRISPR–Cas9 technology is a powerful tool for genome manipulation in eukaryotic cells. For in vitro experiments, this technique requires delivery of guide RNAs and Cas9 endonuclease into cells either by transfection or transduction. Production of gene-edited immune cells requires either Cas9-expressing cell lines or primary cells isolated from Cas9-expressing mice. A common “safe harbor” integration site used in Cas9-expressing mice is Rosa26 locus (13). This locus is situated on mouse chromosome 6 and consists of three transcripts that were biotyped as long noncoding RNAs with unknown functions (14). The advantage of using Rosa26 locus is that the expression of transgenes are ubiquitous and stable.

We have observed that bone marrow (BM)-derived macrophages (BMMs) from Rosa26-integrated Cas9-expressing C57BL/6J (B6) mice fail to fully activate IRF3, STAT1, and the NF-κB signaling axis following LPS or poly(I:C) stimulation, whereas lipoteichoic acid (LTA) or peptidoglycan (PGN) from Staphylococcus aureus or CpG-ODN-1585–mediated signals were not affected. The zymosan-stimulated NF-κB signaling pattern was similar to that of LPS. We ruled out any possible genetic defects in the IRF3 or STAT1 signaling by monitoring phosphorylation of STAT1 by activating the RIG-I– or STING-mediated pathway to type I IFN production and by stimulating cells with exogenous IFN-β. We propose that there is a putative defect in TRIF signaling in primary BMMs isolated from two commercially available and widely used B6–Cas9 mice. The precise mechanism of this defect is still unknown.

All strains used in this study are available at The Jackson Laboratory (Bar Harbor, ME) and are as follows: B6 (stock no: 000664), Cas9-KR (stock no: 028555), Cas9-FZ (stock no: 026179), Rosa26-Cas9-129 derived (Cas9-129-FZ) (stock no: 024858), Rosa26–lox-stop-lox (LSL)-Cas9-KR (stock no: 028551), Cas9-H11 (stock no: 028239), and Rosa26-LSL-TdTomato (stock no: 007914). Cas9-H11 mice were provided by Dr. Marc Schwartz, Cas9-FZ and Cas9-129-FZ mice were provided by Dr. Kathryn Geiger-Schuller, and Rosa26-LSL-TdTomato mice were provided by Morgan Fleishman, all from the Broad Institute of MIT and Harvard. Animals were housed either in Massachusetts Institute of Technology or at the Broad Institute of MIT and Harvard. Protocols related to animal work were reviewed and approved by Institutional Animal Care and Use Committees.

Cells were grown at 37°C in a humidified incubator supplemented with 5% CO2. BM cells were harvested from 6–12-wk-old female mice and plated on non–tissue culture–treated petri dishes in RPMI-1640 medium (Thermo Fisher Scientific: 21870-092), supplemented with 10% heat-inactivated FBS (VWR: 97068-091), l-glutamine (VWR: 45000-676), penicillin/streptomycin (VWR: 45000-652), MEM nonessential amino acids (VWR: 45000-700), HEPES (VWR: 45000-690), sodium pyruvate (VWR: 45000-710), and β-mercaptoethanol (Thermo Fisher Scientific: 21985023, 1000×). BMMs were generated by growing cells in the presence of M-CSF (10 ng/ml, R&D Systems: 216-MC-025). BMMs were collected and reseeded on assay plates after 5–6 d of differentiation. All experiments were performed at 6–7 d of differentiation. BM-derived dendritic cells (BMDCs) were generated by growing cells in the presence of GM-CSF (20 ng/ml, PeproTech: 315-03). BMDCs were collected and reseeded on day 8, and all experiments were performed on day 9. Cells were stimulated with 100 ng/ml of LPS, 10 μg/ml of DMXAA (a STING agonist), 10 μg/ml of poly(I:C), 100 ng/ml of purified LTA–S. aureus, 2 μg/ml of PGN–S. aureus, 5 μg/ml of zymosan or 2 μM CpG-ODN-1585 (all from InvivoGen: tlrl-peklps, tlr-dmx, tlrl-pic, HMW, tlrl-pslta, tlrl-pgns2, tlrl-zyn, and ODN-1585, respectively), 10 HA units/ml of SeV (Charles River Laboratories: material no. 10100816), and 100 units/ml of IFN-β (R&D Systems: 8234-MB-010/CF) for the indicated times. Necroptosis was analyzed on immunoblots after treating cells with 25 μM of Z-VAD-FMK (APExBIO Technology: A1902), a pan-caspase inhibitor, alone or with LPS/poly(I:C) for 2 h. For TAT-Cre recombinase treatment, cells were incubated with either 100 units/ml (MilliporeSigma: SCR508) or 100 μg/ml (Excellgen: EG-1001) of enzyme at 37°C for 45 min in Opti-MEM medium (Thermo Fisher Scientific: 31985062) (15). Cells were then washed and grew in BMM media for 3 d. On day 3, immunoblotting were performed with cells stimulated with LPS for 2 h. For ELISA, 106 cells/well on six-well plates were stimulated with LPS. Supernatants were collected at 0, 2, 4 and 6 h and used for specific ELISA (BioLegend: TNF-α [430901], IL-6 [431301] and IFN-β [439407]). In one experiment, immortal BMMs, a gift from Dr. Karin Pelka, generated by retroviral transduction (16), were used. Immunoblotting was performed with cells lysed with radioimmunoprecipitation assay buffer (Boston BioProducts: BP-115) in the presence of protease and phosphatase inhibitors (Sigma: complete mini 4693124001 and PhosSTOP 4906837001). Equal amounts of protein were separated by SDS-PAGE and transferred to nitrocellulose membrane (Iblot2 stack) for 7 min using iBlot2 dry blotting system (Thermo Fisher Scientific: IB21001S). Membranes were then blocked with milk and probed with primary Abs followed by HRP-conjugated secondary Ab (Jackson ImmunoResearch Laboratories: 111-035-144 for goat anti-rabbit and 115-035-146 for goat anti-mouse; Santa Cruz Biotechnology: sc-2020 for donkey anti-goat) and were visualized by ECL (VWR: PI32209 and 89168-782), according to the instructions of the manufacturer.

Myddosome assembly and coimmunoprecipitaion experiments were performed following the published protocol (17), except that we used Dynabeads (Thermo Fisher Scientific: 88803). For flow cytometry, BMMs were stimulated with LPS for 30 min and 2 h and labeled with two fluorescence-conjugated anti-TLR4 Abs, allophycocyanin anti-mouse CD284 (TLR4) Ab, and PE anti-mouse TLR4 (CD284)/MD2 complex Ab (BioLegend: 145406 and 117605). Unstimulated cells were used as a control.

All plasmids were obtained from the Genetic Perturbation Platform of the Broad Institute at MIT and Harvard. Lentiviral particles were generated by transfecting plasmid together with viral packaging vectors psPAX2 and VSVg to HEK293T cells (American Type Culture Collection: CRL-3216) using lipofectamine 3000 transfection reagent (Thermo Fisher Scientific: L3000-008). Viral particles were collected 24 h after transfection, and transductions were done on a 2-d-old culture of BMMs. Transduced cells were selected by antibiotics, and immunoblots were performed after stimulating cells with LPS for 2 h.

Abs used in this study were as follows: β-actin (Abcam: ab6276) as a loading control, MyD88 (R&D Systems: AF3109-SP), serum isotype for MyD88 (R&D Systems: AB-108-C), and IRF3 (BioLegend: 655702). The following Abs were from Cell Signaling Technology with their corresponding catalog numbers in brackets: p-STAT1Y701 (9167), STAT1 (9172), p-IRF3 (4947), p-TBK1 (5483), TBK1 (3013), p-IKKα/β (2697), IKKβ (2370), p-TAK1 (4508), TAK1 (4505S), IκBα (9242), p-RIPK3 (91702), RIPK3 (95702), p-MLKL (37333), MLKL (37705), Cas9 (14697), m-Cherry (43590), and GFP (2956).

B6 mice were crossed with Cas9-KR mice to generate heterozygotes, which were then crossed within themselves to generate wild-type littermates and Cas9-KR homozygotes strains. Mice were genotyped with ear-punched DNA using PCR.

BMMs from wild-type (B6), Cas9-FZ, and Cas9-KR were stimulated with/without LPS for 2 h, and total RNA was extracted using RNeasy mini kit (QIAGEN: 74104) following the instruction of the manufacturer and quantified. Then, 1 ng of RNA was used in triplicates for the synthesis of cDNA using the Superscript III reverse transcription system (Thermo Fisher Scientific: 18080093). cDNA quality was checked using High Sensitivity DNA Bioanalyzer chip (Agilent Technology), and 0.15 ng cDNA was subjected to the process of making an amplified library and sequenced on HiSeq 2500 (Illumina) to generate Smart-Seq2 sequencing data. For data analysis, short sequencing reads were aligned using STAR (18) to the mm10 reference genome and used as input in RSEM (19) to quantify gene expression levels. Data were normalized and analyzed using the R software package DESeq2 (20). Gene set enrichment analysis was used to assess differentially expressed pathways (21).

Variations among experiments were shown by performing densitometric analysis of phosphorylated proteins relative to β-ACTIN using ImageJ software from the National Institutes of Health and by doing statistical analysis. Bar graphs were plotted, SDs were introduced, and Student t tests were performed for statistical analysis where p < 0.05 was considered statistically significant. Statistically significant data compared with wild-type (B6) at a particular time point was shown with an asterisk on top of the bar graph.

To study the effect of gene knockouts in mouse immune cells, we initially used BMMs from one of the commercially available Cas9-expressing B6 mice. This strain was originally generated in Drs. Klaus Rajewsky and Ralf Kuhn’s laboratory (15) and hereafter called Cas9-KR. Surprisingly, we observed that phosphorylation of IRF3 and STAT1 were absent in BMMs upon LPS stimulation compared with BMMs derived from B6 control mice (Fig. 1A; B6 versus KR).

FIGURE 1.

BMMs from two independently generated lines of Rosa26-Cas9–expressing mice on B6 genetic background are defective for TLR-3/4 signaling in response to LPS or poly(I:C).

BMMs from B6 and Cas9-KR were stimulated with (A) LPS, (B) SeV, and (C) DMXAA; BMDCs from B6 and Cas9-KR were stimulated with (D) LPS; BMMs from B6, Cas9-FZ, and Cas9-KR were treated with (E) IFN-β; BMMs from B6, Cas9-FZ, and Cas9-129-FZ (PC) were stimulated with (F) LPS and (G) DMXAA; and BMMs from B6, Cas9-KR, and Cas9-FZ were stimulated with (H) poly(I:C) for the indicated times. Immunoblotting was performed for the indicated proteins. All results were representative of three independent experiments.

FIGURE 1.

BMMs from two independently generated lines of Rosa26-Cas9–expressing mice on B6 genetic background are defective for TLR-3/4 signaling in response to LPS or poly(I:C).

BMMs from B6 and Cas9-KR were stimulated with (A) LPS, (B) SeV, and (C) DMXAA; BMDCs from B6 and Cas9-KR were stimulated with (D) LPS; BMMs from B6, Cas9-FZ, and Cas9-KR were treated with (E) IFN-β; BMMs from B6, Cas9-FZ, and Cas9-129-FZ (PC) were stimulated with (F) LPS and (G) DMXAA; and BMMs from B6, Cas9-KR, and Cas9-FZ were stimulated with (H) poly(I:C) for the indicated times. Immunoblotting was performed for the indicated proteins. All results were representative of three independent experiments.

Close modal

Despite the loss of phosphorylated IRF3, phosphorylation of TBK1 was maintained at a level similar to B6 BMMs (Fig. 1A; see Supplemental Fig. 1 for quantitation of proteins for (Fig. 1). We hypothesize that TBK1 phosphorylation in Cas9-KR BMMs occurs downstream of Myddosome formation and is not TRIF mediated, which was based on a recent finding that upon LPS stimulation, the activation of TBK1 was initiated by the Myddosome complex (22).

We then sought to determine whether the defect in IRF3 activation we observed in LPS-treated Cas9 BMMs was present when cells were treated with other stimuli that also induce type I IFN production. To test this, we activated the RIG-I or STING pathways in B6 BMMs or Cas9-KR BMMs through SeV infection or treatment with DMXAA, both of which require TBK1 for IRF3 phosphorylation and type I IFN production. We did not observe any difference in phosphorylation of STAT1, IRF3, or TBK1 between B6 and Cas9-KR cells (Fig. 1B, 1C). This confirms that there is no genetic defect in RIG-I/STING–mediated activation of IRF3/STAT1 in Cas9-KR BMMs. In addition, BMDCs from Cas9 mice also display a defect in LPS-induced phosphorylation of IRF3 and STAT1 (Fig. 1D).

To determine if the observed defect in the TLR4 signaling pathway was specific to this single Cas9-expressing mouse strain, we tested the response to LPS in BMMs derived from a second independently derived Cas9 strain, originally produced in Dr. Feng Zhang’s laboratory (referred to hereafter as Cas9-FZ) (23). The Cas9 transgene present in both strains is similar: both are integrated at the Rosa26 locus and expressed eGFP. Differences between the two transgenes include their specific generation methods and the method of eGFP expression (15, 23). The parental Cas9 (PC)-FZ mouse was a chimeric strain on 129 background (hereafter called Cas9-129-FZ or PC), which after efficient backcrossing with B6, generated Cas9-FZ.

We treated all cells with 100 units/ml of IFN-β and followed the activation of STAT1. Both Cas9-FZ and Cas9-KR BMMs responded normally to this stimulation compared with B6 BMMs (Fig. 1E). This result more definitively ruled out any genetic defect that affects IFN-β–STAT1 signaling.

Upon stimulation with LPS (Fig. 1F), but not DMXAA (Fig. 1G), Cas9-FZ BMMs displayed abrogated IRF3 and STAT1 phosphorylation, similar to our previous results in Cas9-KR BMMs (Fig. 1A versus (Fig. 1F). Lack of secretion of IFN-β in LPS-stimulated Cas9 BMMs was confirmed by ELISA analysis (Supplemental Fig. 2). Surprisingly, the PC/Cas9-129-FZ BMMs did not display any defects in the TLR4 signaling pathway as evident by the phosphorylation of IRF3 and STAT1 compared with the B6 BMMs (Fig. 1F; B6 versus PC). Further, LPS stimulation activated TBK1 in all samples, similar to Cas9-KR BMMs (Fig. 1A, 1F). Together, these results suggest that LPS-stimulated activation of IRF3 and STAT1 is defective in BMMs derived from Rosa26-integrated Cas9-expressing mice of B6 background, despite effective phosphorylation of TBK1.

We next wanted to know if poly(I:C)-stimulated and TLR3/TRIF-mediated activation of IRF3 was defective in the BMMs of Cas9-KR and Cas9-FZ compared with the B6 sample. As expected, Fig. 1H), poly(I:C) did not lead to TRIF-mediated signaling, based on the lack of phosphorylation of TBK1 as well as IRF3.

As we established that Cas9-expressing BMMs did not produce IFN-β, we wanted to investigate whether the difference in signaling pathways affects downstream processes. We performed RNA sequencing (RNA-seq) to unbiasedly characterize gene expression levels of the entire transcriptome. We analyzed RNA-seq data for BMMs collected from B6, Cas9-FZ, and Cas9-KR mice at baseline or 2 h post–LPS stimulation. We found that there is a drastic difference in gene expression following LPS stimulation between Cas9-FZ and B6 mice, with thousands of genes significantly differentially expressed between the two strains (Fig. 2A). There is also a large magnitude of difference between Cas9-KR and B6 mice following stimulation (Fig. 2B). We confirmed that these differences are consistent across biological replicates (Fig. 2C, 2D). Pathway enrichment analysis reveals that many immune processes are significantly different between Cas9-FZ and B6 strains. Two representative gene sets are shown (Fig. 2E, 2F), and all data are available for analysis (accession number GSE184551 at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE184551).

FIGURE 2.

RNA-seq reveals drastic differences in transcriptomic responses to LPS stimulation between mouse strains.

Volcano plot showing differentially expressed genes between (A) Cas9-FZ and B6 and (B) Cas9-KR and B6 2 h post–LPS stimulation. Significantly differentially expressed genes (log2 fold change >1, adjusted p value <0.01) were colored red. (C and D) Heatmap showing the top differentially expressed genes between Cas9-FZ and B6 and between Cas9-KR and B6, respectively. Row-normalized gene expression values were shown. (E and F) Gene set enrichment analysis for Cas9-FZ versus B6. Two representative gene sets were shown.

FIGURE 2.

RNA-seq reveals drastic differences in transcriptomic responses to LPS stimulation between mouse strains.

Volcano plot showing differentially expressed genes between (A) Cas9-FZ and B6 and (B) Cas9-KR and B6 2 h post–LPS stimulation. Significantly differentially expressed genes (log2 fold change >1, adjusted p value <0.01) were colored red. (C and D) Heatmap showing the top differentially expressed genes between Cas9-FZ and B6 and between Cas9-KR and B6, respectively. Row-normalized gene expression values were shown. (E and F) Gene set enrichment analysis for Cas9-FZ versus B6. Two representative gene sets were shown.

Close modal

To understand the source of the defect in TLR-3/4 signaling in Cas9 BMMs, we examined different steps of the pathway in B6 and Cas9 BMMs. We first examined immunoprecipitates of MyD88 to determine whether Cas9 BMMs can form the Myddosome complex (17) upon stimulation with LPS, as read out by coimmunoprecipitation of TBK1 with Myd88. Indeed, TBK1 coimmunoprecipitated with MyD88 (Fig. 3A) in B6 BMMs and BMMs from both Cas9-expressing mice, consistent with the notion that LPS-induced phosphorylation of TBK1 in Cas9 BMMs (Fig. 1A) was indeed Myd88 mediated (22) and not dependent on TRIF (this is in contrast to our results showing a lack of TBK1 phosphorylation, which is known to be TRIF dependent, following poly(I:C) treatment in Cas9 BMMs) (Fig. 1H).

FIGURE 3.

Dissection of TLR-3/4–mediated signaling pathway upon LPS or poly(I:C) stimulation.

(A) Myddosome assembly: Indicated BMMs were stimulated with LPS for the indicated times and immunoprecipitated with anti-MyD88 Ab. Immunoblotting was performed for both MyD88 and TBK1 proteins. The corresponding immunoblots of whole cell extracts (WCE) were shown. (B and C) Post-Myddosome signaling: Indicated BMMs were stimulated with (B) LPS or (C) poly(I:C) for the indicated times, and cell lysates were analyzed by immunoblotting for the indicated proteins. (D) TLR4 internalization: (upper panel) BMMs were stimulated with LPS for the indicated times and surface stained with two TLR4-specific fluorescent Abs to study endocytosis by flow cytometry. Lower panel, Change of mean fluorescence intensity of TLR4 over time with both Abs were shown. (E) TRIF-mediated necroptosis: Indicated BMMs were treated with LPS or poly(I:C) in the presence or absence of Z-VAD-FMK for 2 h. Cell lysates were used for immunoblot analysis for the indicated proteins. All results were representative of three independent experiments.

FIGURE 3.

Dissection of TLR-3/4–mediated signaling pathway upon LPS or poly(I:C) stimulation.

(A) Myddosome assembly: Indicated BMMs were stimulated with LPS for the indicated times and immunoprecipitated with anti-MyD88 Ab. Immunoblotting was performed for both MyD88 and TBK1 proteins. The corresponding immunoblots of whole cell extracts (WCE) were shown. (B and C) Post-Myddosome signaling: Indicated BMMs were stimulated with (B) LPS or (C) poly(I:C) for the indicated times, and cell lysates were analyzed by immunoblotting for the indicated proteins. (D) TLR4 internalization: (upper panel) BMMs were stimulated with LPS for the indicated times and surface stained with two TLR4-specific fluorescent Abs to study endocytosis by flow cytometry. Lower panel, Change of mean fluorescence intensity of TLR4 over time with both Abs were shown. (E) TRIF-mediated necroptosis: Indicated BMMs were treated with LPS or poly(I:C) in the presence or absence of Z-VAD-FMK for 2 h. Cell lysates were used for immunoblot analysis for the indicated proteins. All results were representative of three independent experiments.

Close modal

We next determined the state of the NF-κB signaling cascade downstream of the Myddosome complex by examining phosphorylation of TAK1 and IKKα/β as well as the levels of total IκBα. We observed a very rapid activation of NF-κB signaling pathway in Cas9-KR and Cas9-FZ BMMs within 30 min of LPS stimulation, similar to B6 BMMs (Fig. 3B; see Supplemental Fig. 3 for quantitation of proteins for (Fig. 3), which is consistent with a recent study (24). However, at 1 h after LPS stimulation, phospho-TAK1 and phospho-IKKα/β were reduced, and total IκBα was elevated in Cas9 BMMs when compared with B6 samples. This early and unsustained activation of NF-κB in Cas9 BMMs is reminiscent of defects in TRIF signaling (25).

To directly assay the inability of Cas9 BMMs to activate TRIF, we stimulated cells with poly(I:C), a ligand for TLR3, which only uses TRIF for NF-κB activation. Cas9-KR and Cas9-FZ BMMs had strongly reduced TLR3-mediated phosphorylation of TAK1, IKKα/β, and higher total IκBα levels compared with B6 BMMs, indicating that TRIF activation is defective in Cas9-expressing BMMs (Fig. 3C).

Loss of TRIF was previously shown to attenuate LPS-mediated activation of TLR4 as TRIF was responsible for a second wave of activation to maintain NF-κB oscillations (8). The decrease in phosphorylation of TAK1 and IKKα/β and the suppressed/unsustained activation of NF-κB in Cas9-expressing BMMs was consistent with a defective activation. Finally, consistent with the less sustained NF-κB activation, Cas9 BMMs displayed very low induction of TNF-α and IL-6 production in response to LPS as measured by ELISA (Supplemental Fig. 2B–E).

As TRIF requires endocytosis (26) of TLR4, we determined if this aspect of TRIF signaling was intact in Cas9 BMMs. All BMMs were stimulated with LPS for 30 min and 2 h. We observed that both Cas9-KR and Cas9-FZ BMMs expressed similar levels of surface TLR4 as B6 BMMs at steady state and efficiently internalized TLR4 upon LPS stimulation (Fig. 3D).

Finally, we wanted to learn if the TRIF-mediated necroptosis pathway was functional in Cas9 BMMs upon caspase inhibition or not. Upon LPS or poly(I:C) stimulation and blocking caspase activity, TRIF interacts with RIPK3 through its receptor-interacting protein kinase homotypic interaction motif (RHIM) domain and subsequently RIPK3 activates MLKL to initiate necroptosis (10, 27). We stimulated both B6 and Cas9 BMMs with LPS or poly(I:C) in the presence or absence of Z-VAD-FMK, a pan-caspase inhibitor, for 2 h and followed the activation of RIPK3 and MLKL by immunoblotting. Consistent with the defect in TRIF (10, 28), phosphorylation of both RIPK3 and MLKL was reduced following stimulation in the presence of caspase inhibitor in Cas9 BMMs (Fig. 3E) compared with B6 BMMs.

To determine if signaling via cell surface TLRs other than TLR3/4 are affected, we have treated cells with three widely used TLR2 ligands: purified bacterial products, LTA (LTA–S. aureus) and PGN (PGN–S. aureus), or the yeast product zymosan, an insoluble cell wall preparation (2931). TLR2 forms heterodimers with either TLR1 or TLR6 and sends signals through MyD88 to activate NF-κB. Purified LTA–S. aureus led to patterns of activation for IKKα/β and T-IκBα that were similar in Cas9-expressing and B6 cells, peaking at 30 min after stimulation (Fig. 4A, 4D). PGN–S. aureus–stimulated cells also showed no difference in the activation of IKKα/β, which peaked at 1 h (Fig. 4B, 4E). In contrast, the activation pattern of zymosan-stimulated cells was reduced in Cas9 BMM, with a strikingly similar pattern to that of LPS (Fig. 4C, 4F versus (Fig. 3B), suggesting that the zymosan preparation contained mannan, which triggered TLR4 signaling (32, 33).

FIGURE 4.

Activation of TLR2.

BMMs from B6, Cas9-FZ, and Cas9-KR were stimulated with TLR2 ligands: (A) purified LTA–S. aureus, (B) PGN–S. aureus, and (C) zymosan, respectively, for the indicated times, and immunoblotting was performed for the indicated proteins. All results were representative of three independent experiments. The corresponding densitometric analysis of the phosphorylated proteins compared with β-ACTIN were presented in (D), (E), and (F), respectively.

FIGURE 4.

Activation of TLR2.

BMMs from B6, Cas9-FZ, and Cas9-KR were stimulated with TLR2 ligands: (A) purified LTA–S. aureus, (B) PGN–S. aureus, and (C) zymosan, respectively, for the indicated times, and immunoblotting was performed for the indicated proteins. All results were representative of three independent experiments. The corresponding densitometric analysis of the phosphorylated proteins compared with β-ACTIN were presented in (D), (E), and (F), respectively.

Close modal

Similar to TLR3, TLR9 functions through endosome (34), which can be stimulated by synthetic unmethylated CpG-ODN. Synthetic CpG-ODNs are divided mainly into three classes, A, B and C, based on the activation of the IRF7–IFN-α pathway, NF-κB pathway, or both, respectively (3436).

Using the class A immunostimulatory CpG oligonucleotide (CpG-ODN-1585), we observed that both B6 and Cas9 BMMs displayed similar levels of STAT1 activation after 4 h of incubation (Fig. 5A, 5B), as expected for a pathway that does not rely on TRIF.

FIGURE 5.

Stimulation of BMMs with CpG-ODN-1585.

BMMs from B6, Cas9-FZ, and Cas9-KR were stimulated with TLR9 ligand CpG-ODN-1585 for the indicated times and immunoblotting of the indicated proteins was performed (A). Results were representative of three independent experiments. Densitometric scanning of pSTAT1 compared with β-ACTIN was shown (B).

FIGURE 5.

Stimulation of BMMs with CpG-ODN-1585.

BMMs from B6, Cas9-FZ, and Cas9-KR were stimulated with TLR9 ligand CpG-ODN-1585 for the indicated times and immunoblotting of the indicated proteins was performed (A). Results were representative of three independent experiments. Densitometric scanning of pSTAT1 compared with β-ACTIN was shown (B).

Close modal

We next sought to determine which aspect of Cas9 expression in BMMs was responsible for this defect in TRIF signaling. The components of the transgene that we tested include the integration site, the promoter, the expression of Cas9 itself, and the eGFP reporter tag. Both Cas9-expressing strains we have used in this study used the Rosa26 locus as the integration site, and the transgene was under the control of the CAG promoter.

To examine the effect of Cas9 in a different locus on LPS- or poly(I:C)-induced activation of NF-κB and TRIF-mediated signaling, we compared the previously described Cas9-KR BMMs to BMMs derived from the mouse strain in which the Cas9 transgene was integrated on chromosome 11 at the Igs2-H11 locus (Cas9-H11) (37) and was also driven by CAG promoter. In addition to containing a different integration site for Cas9, the transgene in this strain also lacks any fluorescence marker, allowing us to determine whether CAG promoter–driven Cas9 expression was sufficient to inhibit TRIF activation.

We studied the activation of TLR3/4 signaling molecules upon stimulation with LPS or poly(I:C). We observed that phosphorylation of STAT1, IKKα/β, and expression of total IκBα were similar between Cas9-H11 BMMs and B6 BMMs, whereas Cas9-KR BMMs showed TRIF-mediated defects as expected (Fig. 6A, 6B, 6E–H). These results indicate that neither CAG promoter nor Cas9 expression themselves contribute to the TRIF-mediated signaling defect.

FIGURE 6.

Rosa26 integration site, CAG promoter–driven Cas9, or eGFP proteins independently do not contribute to TRIF signaling defect in BMMs from two independently generated lines of Rosa26-Cas9–expressing mice on B6 genetic background.

(A and B) BMMs from B6, Cas9-KR, and CAG promoter–driven and H11 locus–integrated Cas9-expressing B6 mice were stimulated with (A) LPS or (B) poly(I:C) for the indicated times, and immunoblotting was performed for the indicated proteins. (C) BMMs from B6 mice were transduced with lentiviral particles generated from expression plasmids containing empty vector (EV), Cas9 (C), GFP (G), or combined Cas9/GFP (CG) and treated with LPS for 2 h. Cell lysates were used for immunoblotting for the indicated proteins. (D) Immortalized macrophages and Cas9-expressing-immortalized macrophages were treated with LPS for the indicated times, and cell lysates were used for immunoblotting for the indicated proteins. All results were representative of two independent experiments. Densitometric analysis of the phosphorylated proteins compared with β-ACTIN was presented in panels (E)–(L).

FIGURE 6.

Rosa26 integration site, CAG promoter–driven Cas9, or eGFP proteins independently do not contribute to TRIF signaling defect in BMMs from two independently generated lines of Rosa26-Cas9–expressing mice on B6 genetic background.

(A and B) BMMs from B6, Cas9-KR, and CAG promoter–driven and H11 locus–integrated Cas9-expressing B6 mice were stimulated with (A) LPS or (B) poly(I:C) for the indicated times, and immunoblotting was performed for the indicated proteins. (C) BMMs from B6 mice were transduced with lentiviral particles generated from expression plasmids containing empty vector (EV), Cas9 (C), GFP (G), or combined Cas9/GFP (CG) and treated with LPS for 2 h. Cell lysates were used for immunoblotting for the indicated proteins. (D) Immortalized macrophages and Cas9-expressing-immortalized macrophages were treated with LPS for the indicated times, and cell lysates were used for immunoblotting for the indicated proteins. All results were representative of two independent experiments. Densitometric analysis of the phosphorylated proteins compared with β-ACTIN was presented in panels (E)–(L).

Close modal

To further determine the effect of LPS stimulation on proteins expressed as a result of random integration, we transduced B6 BMMs with lentiviral particles containing empty vector, Cas9, eGFP, and combined Cas9-GFP (Fig. 6C, 6I, 6J). In parallel, we immortalized B6 BMMs (iMac) and transduced iMac with Cas9-containing lentiviral particles to generate Cas9-expressing immortalized BMMs (iMac-Cas9) and stimulated both with LPS (Fig. 6D, 6K, 6L). In both cases, we did not observe any difference in activation pattern. It ruled out any possible defect in TRIF signaling because of the expression of Cas9 or eGFP.

As we have not identified any cause of TRIF signaling defects in the above-mentioned scenarios, we aimed to determine whether expression of a transgene at the Rosa26 locus, as opposed to its integration, was responsible for disrupting TRIF signaling. To test this, we used BMMs derived from another Cas9-KR strain, in which the protein Cas9 was not expressed. We used a commercially available mouse strain containing LSL–Cas9 transgene at the Rosa26 locus (which have not been crossed to CMV-Cre lines to generate Cas9-KR). LSL cassette prevents expression of Cas9 (and eGFP). This expression can also be released in vitro by treating LSL containing Cas9 BMMs with TAT-Cre recombinase enzyme. As a control, we used Rosa26-LSL-TdTomato (38), in which TdTomato transgene was driven by CAG promoter, to monitor the efficacy of TAT-Cre enzyme.

We stimulated BMMs from B6, Cas9-KR, and LSL-Cas9-KR with LPS and DMXAA for 2 h and performed immunoblotting for phosphorylated and total IRF3, TBK1, and STAT1 (Fig. 7A, 7C–E). Although Cas9-KR displayed abrogated IRF3 and STAT1 phosphorylation, surprisingly, LSL-Cas9-KR had similar levels of IRF3 and STAT1 activation as B6. These results led us to hypothesize that removal of LSL cassette may play a role in altering TRIF signaling.

FIGURE 7.

Expression of transgene from the Rosa26 locus does not disrupt TRIF signaling.

(A) BMMs from B6, Cas9-KR, and Rosa26-LSL-Cas9-KR (L) were stimulated with the indicated stimuli for 2 h, and immunoblotting was performed for the indicated proteins. (B) BMMs from B6, Rosa26-LSL-Cas9-KR, and Rosa26-LSL-TdTomato were treated with TAT-Cre recombinase from two different sources and stimulated with LPS for 2 h. Cas9-KR BMMs were used as a negative control. Cell lysates were used for immunoblot analysis for the indicated proteins. Results were representative of two independent experiments. Densitometric data of phosphorylated proteins over β-ACTIN from panels (A) and (B) were presented in panels (C)–(E) and panels (F) and (G), respectively.

FIGURE 7.

Expression of transgene from the Rosa26 locus does not disrupt TRIF signaling.

(A) BMMs from B6, Cas9-KR, and Rosa26-LSL-Cas9-KR (L) were stimulated with the indicated stimuli for 2 h, and immunoblotting was performed for the indicated proteins. (B) BMMs from B6, Rosa26-LSL-Cas9-KR, and Rosa26-LSL-TdTomato were treated with TAT-Cre recombinase from two different sources and stimulated with LPS for 2 h. Cas9-KR BMMs were used as a negative control. Cell lysates were used for immunoblot analysis for the indicated proteins. Results were representative of two independent experiments. Densitometric data of phosphorylated proteins over β-ACTIN from panels (A) and (B) were presented in panels (C)–(E) and panels (F) and (G), respectively.

Close modal

To test this, we compared BMMs from LSL-Cas9-KR and LSL-TdTomato mice and used B6 BMMs as a control. We used recombinant TAT-Cre recombinase obtained from two different vendors to remove the stop cassette in vitro during BMM differentiation followed by stimulation with LPS. Although we observed recombination in both of our samples as evident by the expressions of eGFP and TdTomato (Fig. 7B), recombinase treatment did not suppress IRF3 or STAT1 activation. Moreover, one of the enzymes itself caused downregulation of STAT1 phosphorylation upon LPS stimulation on control B6 BMMs (Fig. 7B, 7F, 7G). It suggests that release of Cas9 (and eGFP) protein or integration of Cas9-eGFP transgene at Rosa26 locus in B6 mice is at least not creating any disruption to activate the TRIF-mediated signaling defect.

In addition, to study the above-mentioned aspects, we generated a batch of wild-type littermate, wild-type Cas9-KR heterozygote and homozygote Cas9-KR. BMMs from these strains were stimulated with LPS and poly(I:C) to study both Myd88- and TRIF-dependent pathways. BMMs from Cas9-expressing homozygotes showed similar protein activation patterns as that of Cas9-KR (Supplemental Fig. 4A–D), suggesting a permanent TRIF-mediated genetic defect is present in Rosa26-Cas9–expressing B6 mice.

In this study, we found that BMMs derived from B6 mouse strains, with a Cas9-eGFP transgene integrated at the Rosa26 locus, had defects in TLR-3/4–induced TRIF-mediated signaling. We used BMMs isolated from two independently generated, commercially available Rosa26-integrated Cas9-expressing mice. We found that LPS (TLR4) or poly(I:C) (TLR3) stimulation did not activate IRF3/STAT1 in Cas9 BMMs, even though RIG-I– and STING-mediated pathways were not affected, and an exogenous addition of IFNβ generated similar activation of STAT1 in all samples, excluding genetic defects downstream of IRF3 activation. We further established the presence of a defect in TRIF signaling in Cas9-expressing BMMs, which also led to suppression of the NF-κB signaling axis in Cas9 BMMs compared with B6. We were surprised to find phosphorylation of TBK1 in LPS-stimulated Cas9 BMMs and not in poly(I:C)-stimulated Cas9 BMMs, which prompted us to show that the activation of TBK1 was Myddosome derived and not TRIF mediated. We also have shown that LPS- or poly(I:C)-stimulated Cas9-expressing cells failed to activate proteins responsible for TRIF-mediated necroptosis upon pan-caspase inhibition. While studying another cell surface TLR, TLR2, we observed that BMMs from all samples responded normally to stimulation with two TLR2 agonists, purified LTA–S. aureus, and PGN–S. aureus. Cas9 BMMs responded differently to a third TLR2 ligand Zym compared with B6 cells. Although the major component of zymosan is β-glucan, it could also contain mannan, mannoprotein, and chitin. There is evidence that mannan can signal through TLR4 (32, 33). As the zymosan stimulation pattern was exactly like that of LPS, it may be due to mannan-mediated TLR4 activation. This result also provides additional confirmation of a TRIF-related defect in Cas9 BMMs.

TRIF is an adaptor protein for TLR3/4 signaling. It consists of an N-terminal domain, a proline-rich region containing TBK1 (for activation of IRF3), TRAF6 and TRAF2 binding sites, a TIR domain, which binds to the C-terminal region of TLR3 or TIR-containing protein TRAM, and a RHIM containing C-terminal domain, the region responsible for apoptosis or necroptosis (10, 39, 40). The current model for TRIF-mediated IRF3 phosphorylation involves a multistep interaction with TBK1 (41, 42), by which (1) TRIF recruits and activates TBK1 (2), activated TBK1 phosphorylates TRIF at its pLxIS site (3), IRF3 is recruited to TRIF and phosphorylated by TBK1 (4), and phosphorylated IRF3 dissociates from the complex and dimerizes to form the functional transcription factor. Because IRF3 was not phosphorylated when cells were stimulated through TLR3 or TLR4, which act through TRIF, we hypothesize that some step in TRIF-dependent activation of IRF3 is defective in Cas9-expressing BMMs. We also showed that LPS- or poly(I:C)-stimulated Cas9-expressing cells failed to activate RIPK3 and MLKL proteins responsible for TRIF-mediated necroptosis upon pan-caspase inhibition. During the necroptosis process, the RHIM region of TRIF recruits and phosphorylates RIPK3, which subsequently activates MLKL and initiates necroptosis. Because RIPK3 and IRF3 are not phosphorylated in Cas9 BMMs, we conclude that TRIF is likely unable to carry out these two critical functions.

We further established from RNA-seq analysis that because of a lack of IFN-β expression, a substantial numbers of genes including ISGs were downregulated in Cas9 BMMs compared with B6 upon LPS stimulation. These RNA-seq results also highlight that biological studies carried out on the Cas9-FZ and Cas9-KR strains should be interpreted with caution as these strains can have a very different immune response compared with B6.

The Cas9-expressing mouse strains used were generated by crossing Rosa26-LSL-Cas9/eGFP mice with Cre-expressing mice to remove the LSL sequence and allow expression of Cas9/eGFP. Mice were then crossed to B6 for multiple generations to remove Cre and create a uniform B6 background. Although we found a defect in TLR/TRIF signaling in Cas9-FZ that were crossed to B6, we were surprised that Cas9-129-FZ mice on the 129 background had no TLR signaling defect. A similar defect was observed in Cas9-KR, which was originally on the B6 background. The results suggested that the B6 background was important for the phenotype, and because Cas9-H11 mice did not show the signaling defect, the Rosa26 integration site was also critical. However, TAT-Cre recombinase–mediated removal of LSL from Rosa26-LSL-Cas9-KR macrophages did not block TRIF signaling, suggesting that the defect does not appear rapidly and that the defect results from longer-term effects of the modified locus on development and TLR signaling.

We thank Dr. Jonathan C. Kagan and the members of the Hacohen laboratory for feedback on the project.

The gene sets presented in this article have been submitted to the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE184551) under accession number GSE184551.

This work was supported by National Institutes of Health/National Institute of Allergy and Infectious Diseases Grant U19 AI133524 and a gift from Arthur, Sandra, and Sarah Irving that funded a David P. Ryan, MD Endowed Chair in Cancer Research for N.H.

The online version of this article contains supplemental material.

Abbreviations used in this article

B6

C57BL/6J

BM

bone marrow

BMDC

BM-derived dendritic cell

BMM

BM-derived macrophage

ISG

IFN-stimulated gene product

LSL

lox-stop-lox

LTA

lipoteichoic acid

MLKL

mixed lineage kinase domain-like protein

PC

parental Cas9

PGN

peptidoglycan

poly(I:C)

polyinosinic:polycytidylic acid

RHIM

receptor-interacting protein kinase homotypic interaction motif

RNA-seq

RNA sequencing

SeV

Sendai virus

TRIF

TIR domain–containing adaptor-inducing IFN-β

1.
Yamamoto
M.
,
S.
Sato
,
K.
Mori
,
K.
Hoshino
,
O.
Takeuchi
,
K.
Takeda
,
S.
Akira
.
2002
.
Cutting edge: a novel Toll/IL-1 receptor domain-containing adapter that preferentially activates the IFN-beta promoter in the Toll-like receptor signaling.
J. Immunol.
169
:
6668
6672
.
2.
Kawai
T.
,
S.
Akira
.
2010
.
The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors.
Nat. Immunol.
11
:
373
384
.
3.
Lin
S.-C.
,
Y.-C.
Lo
,
H.
Wu
.
2010
.
Helical assembly in the MyD88-IRAK4-IRAK2 complex in TLR/IL-1R signalling.
Nature
465
:
885
890
.
4.
Fitzgerald
K. A.
,
J. C.
Kagan
.
2020
.
Toll-like Receptors and the Control of Immunity.
Cell
180
:
1044
1066
.
5.
Kawai
T.
,
S.
Akira
.
2007
.
Signaling to NF-kappaB by Toll-like receptors.
Trends Mol. Med.
13
:
460
469
.
6.
Fitzgerald
K. A.
,
D. C.
Rowe
,
B. J.
Barnes
,
D. R.
Caffrey
,
A.
Visintin
,
E.
Latz
,
B.
Monks
,
P. M.
Pitha
,
D. T.
Golenbock
.
2003
.
LPS-TLR4 signaling to IRF-3/7 and NF-kappaB involves the toll adapters TRAM and TRIF. [Published erratum appears in 2003 J. Exp. Med. 198: 1450].
J. Exp. Med.
198
:
1043
1055
.
7.
Yamamoto
M.
,
S.
Sato
,
H.
Hemmi
,
K.
Hoshino
,
T.
Kaisho
,
H.
Sanjo
,
O.
Takeuchi
,
M.
Sugiyama
,
M.
Okabe
,
K.
Takeda
,
S.
Akira
.
2003
.
Role of adaptor TRIF in the MyD88-independent toll-like receptor signaling pathway.
Science
301
:
640
643
.
8.
Cheng
Z.
,
B.
Taylor
,
D. R.
Ourthiague
,
A.
Hoffmann
.
2015
.
Distinct single-cell signaling characteristics are conferred by the MyD88 and TRIF pathways during TLR4 activation.
Sci. Signal.
8
:
ra69
.
9.
He
S.
,
Y.
Liang
,
F.
Shao
,
X.
Wang
.
2011
.
Toll-like receptors activate programmed necrosis in macrophages through a receptor-interacting kinase-3-mediated pathway.
Proc. Natl. Acad. Sci. USA
108
:
20054
20059
.
10.
Kaiser
W. J.
,
H.
Sridharan
,
C.
Huang
,
P.
Mandal
,
J. W.
Upton
,
P. J.
Gough
,
C. A.
Sehon
,
R. W.
Marquis
,
J.
Bertin
,
E. S.
Mocarski
.
2013
.
Toll-like receptor 3-mediated necrosis via TRIF, RIP3, and MLKL.
J. Biol. Chem.
288
:
31268
31279
.
11.
Kawai
T.
,
S.
Akira
.
2008
.
Toll-like receptor and RIG-I-like receptor signaling.
Ann. N. Y. Acad. Sci.
1143
:
1
20
.
12.
Chen
Q.
,
L.
Sun
,
Z. J.
Chen
.
2016
.
Regulation and function of the cGAS-STING pathway of cytosolic DNA sensing.
Nat. Immunol.
17
:
1142
1149
.
13.
Zambrowicz
B. P.
,
A.
Imamoto
,
S.
Fiering
,
L. A.
Herzenberg
,
W. G.
Kerr
,
P.
Soriano
.
1997
.
Disruption of overlapping transcripts in the ROSA beta geo 26 gene trap strain leads to widespread expression of beta-galactosidase in mouse embryos and hematopoietic cells.
Proc. Natl. Acad. Sci. USA
94
:
3789
3794
.
14.
Li
S.
,
L. X.
Chen
,
X. H.
Peng
,
C.
Wang
,
B. Y.
Qin
,
D.
Tan
,
C. X.
Han
,
H.
Yang
,
X. N.
Ren
,
F.
Liu
, et al
2018
.
Overview of the reporter genes and reporter mouse models.
Animal Model. Exp. Med.
1
:
29
35
.
15.
Chu
V. T.
,
T.
Weber
,
R.
Graf
,
T.
Sommermann
,
K.
Petsch
,
U.
Sack
,
P.
Volchkov
,
K.
Rajewsky
,
R.
Kühn
.
2016
.
Efficient generation of Rosa26 knock-in mice using CRISPR/Cas9 in C57BL/6 zygotes.
BMC Biotechnol.
16
:
4
15
.
16.
De Nardo
D.
,
D. V.
Kalvakolanu
,
E.
Latz
.
2018
.
Immortalization of Murine Bone Marrow-Derived Macrophages.
In
Macrophages: Methods in Molecular Biology
,
Vol. 1784
.
G.
Rousselet
.
Humana Press
,
New York, NY
, p.
35
49
.
17.
Tan
Y.
,
J. C.
Kagan
.
2018
.
Biochemical Isolation of the Myddosome from Murine Macrophages.
Methods Mol. Biol.
1714
:
79
95
.
18.
Dobin
A.
,
C. A.
Davis
,
F.
Schlesinger
,
J.
Drenkow
,
C.
Zaleski
,
S.
Jha
,
P.
Batut
,
M.
Chaisson
,
T. R.
Gingeras
.
2012
.
STAR: ultrafast universal RNA-seq aligner.
Bioinformatics
29
:
15
21
.
19.
Li
B.
,
C. N.
Dewey
.
2011
.
RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome.
BMC Bioinformatics
12
:
323
.
20.
Love
M. I.
,
W.
Huber
,
S.
Anders
.
2014
.
Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.
Genome Biol.
15
:
550
.
21.
Subramanian
A.
,
P.
Tamayo
,
V. K.
Mootha
,
S.
Mukherjee
,
B. L.
Ebert
,
M. A.
Gillette
,
A.
Paulovich
,
S. L.
Pomeroy
,
T. R.
Golub
,
E. S.
Lander
,
J. P.
Mesirov
.
2005
.
Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles.
Proc. Natl. Acad. Sci. USA
102
:
15545
15550
.
22.
Tan
Y.
,
J. C.
Kagan
.
2019
.
Innate Immune Signaling Organelles Display Natural and Programmable Signaling Flexibility.
Cell
177
:
384
398.e11
.
23.
Platt
R. J.
,
S.
Chen
,
Y.
Zhou
,
M. J.
Yim
,
L.
Swiech
,
H. R.
Kempton
,
J. E.
Dahlman
,
O.
Parnas
,
T. M.
Eisenhaure
,
M.
Jovanovic
, et al
2014
.
CRISPR-Cas9 knockin mice for genome editing and cancer modeling.
Cell
159
:
440
455
.
24.
Bagaev
A. V.
,
A. Y.
Garaeva
,
E. S.
Lebedeva
,
A. V.
Pichugin
,
R. I.
Ataullakhanov
,
F. I.
Ataullakhanov
.
2019
.
Elevated pre-activation basal level of nuclear NF-κB in native macrophages accelerates LPS-induced translocation of cytosolic NF-κB into the cell nucleus.
Sci. Rep.
9
:
4563
.
25.
Zhou
J.
,
T.
Sun
,
S.
Jin
,
Z.
Guo
,
J.
Cui
.
2020
.
Dual feedforward loops modulate type I interferon responses and induce selective gene expression during TLR4 activation.
iScience.
23
:
100881
.
26.
Kagan
J. C.
,
T.
Su
,
T.
Horng
,
A.
Chow
,
S.
Akira
,
R.
Medzhitov
.
2008
.
TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon-beta.
Nat. Immunol.
9
:
361
368
.
27.
Murphy
J. M.
2020
.
The Killer Pseudokinase Mixed Lineage Kinase Domain-Like Protein (MLKL).
Cold Spring Harb. Perspect. Biol.
12
:
a036376
.
28.
Sanjo
H.
,
J.
Nakayama
,
T.
Yoshizawa
,
H. J.
Fehling
,
S.
Akira
,
S.
Taki
.
2019
.
Cutting Edge: TAK1 Safeguards Macrophages against Proinflammatory Cell Death.
J. Immunol.
203
:
783
788
.
29.
Mukherjee
S.
,
S.
Karmakar
,
S. P. S.
Babu
.
2016
.
TLR2 and TLR4 mediated host immune responses in major infectious diseases: a review.
Braz. J. Infect. Dis.
20
:
193
204
.
30.
Roeder
A.
,
C. J.
Kirschning
,
R. A.
Rupec
,
M.
Schaller
,
G.
Weindl
,
H. C.
Korting
.
2004
.
Toll-like receptors as key mediators in innate antifungal immunity.
Med. Mycol.
42
:
485
498
.
31.
Su
S.-C.
,
K.-F.
Hua
,
H.
Lee
,
L. K.
Chao
,
S.-K.
Tan
,
H.
Lee
,
S.-F.
Yang
,
H.-Y.
Hsu
.
2006
.
LTA and LPS mediated activation of protein kinases in the regulation of inflammatory cytokines expression in macrophages.
Clin. Chim. Acta
374
:
106
115
.
32.
Żelechowska
P.
,
E.
Brzezińska-Błaszczyk
,
S.
Różalska
,
J.
Agier
,
E.
Kozłowska
.
2020
.
Mannan activates tissue native and IgE-sensitized mast cells to proinflammatory response and chemotaxis in TLR4-dependent manner.
J. Leukoc. Biol.
109
:
931
942
.
33.
Tada
H.
,
E.
Nemoto
,
H.
Shimauchi
,
T.
Watanabe
,
T.
Mikami
,
T.
Matsumoto
,
N.
Ohno
,
H.
Tamura
,
K.-i.
Shibata
,
S.
Akashi
, et al
2002
.
Saccharomyces cerevisiae- and Candida albicans-derived mannan induced production of tumor necrosis factor alpha by human monocytes in a CD14- and Toll-like receptor 4-dependent manner.
Microbiol. Immunol.
46
:
503
512
.
34.
Wagner
H.
2011
.
New vistas on TLR9 activation.
Eur. J. Immunol.
41
:
2814
2816
.
35.
Vollmer
J.
,
A. M.
Krieg
.
2009
.
Immunotherapeutic applications of CpG oligodeoxynucleotide TLR9 agonists.
Adv. Drug Deliv. Rev.
61
:
195
204
.
36.
Ahmad-Nejad
P.
,
H.
Häcker
,
M.
Rutz
,
S.
Bauer
,
R. M.
Vabulas
,
H.
Wagner
.
2002
.
Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors at distinct cellular compartments.
Eur. J. Immunol.
32
:
1958
1968
.
37.
Hippenmeyer
S.
,
Y. H.
Youn
,
H. M.
Moon
,
K.
Miyamichi
,
H.
Zong
,
A.
Wynshaw-Boris
,
L.
Luo
.
2010
.
Genetic mosaic dissection of Lis1 and Ndel1 in neuronal migration.
Neuron
68
:
695
709
.
38.
Madisen
L.
,
T. A.
Zwingman
,
S. M.
Sunkin
,
S. W.
Oh
,
H. A.
Zariwala
,
H.
Gu
,
L. L.
Ng
,
R. D.
Palmiter
,
M. J.
Hawrylycz
,
A. R.
Jones
, et al
2009
.
A robust and high-throughput Cre reporting and characterization system for the whole mouse brain.
Nat. Neurosci.
13
:
133
140
.
39.
Ullah
M. O.
,
M. J.
Sweet
,
A.
Mansell
,
S.
Kellie
,
B.
Kobe
.
2016
.
TRIF-dependent TLR signaling, its functions in host defense and inflammation, and its potential as a therapeutic target.
J. Leukoc. Biol.
100
:
27
45
.
40.
Kaiser
W. J.
,
M. K.
Offermann
.
2005
.
Apoptosis induced by the toll-like receptor adaptor TRIF is dependent on its receptor interacting protein homotypic interaction motif.
J. Immunol.
174
:
4942
4952
.
41.
Liu
S.
,
X.
Cai
,
J.
Wu
,
Q.
Cong
,
X.
Chen
,
T.
Li
,
F.
Du
,
J.
Ren
,
Y.-T.
Wu
,
N. V.
Grishin
,
Z. J.
Chen
.
2015
.
Phosphorylation of innate immune adaptor proteins MAVS, STING, and TRIF induces IRF3 activation.
Science
347
:
aaa2630
.
42.
Abe
H.
,
J.
Satoh
,
Y.
Shirasaka
,
A.
Kogure
,
H.
Kato
,
S.
Ito
,
T.
Fujita
.
2020
.
Priming phosphorylation of TANK-binding kinase 1 by IκB kinase β is essential in Toll-like receptor 3/4 signaling.
Mol. Cell. Biol.
40
:
e00509
e00519
.

N.H. has equity in BioNTech and has equity in and advises Danger Bio. The other authors have no financial conflicts of interest.

This article is distributed under the terms of the CC BY-NC-ND 4.0 Unported license.

Supplementary data