Material availability
All unique and newly generated materials (cell lines, plasmids) are available from the corresponding authors on request.
Cell culture, primary BMDM generation and treatments
iBMDM lines and HEK293T cells (CRL-11268, ATCC) were cultured in DMEM supplemented with glutamine, sodium pyruvate, penicillin–streptomycin and 10% FBS (cDMEM), at 37 °C under 5% CO2. An overview of all cell lines is provided in Supplementary Table 1.
BMDMs were differentiated from bone marrow from tibiae and femurs of C57BL/6J, WT or Ticam2−/− mice (aged 10 weeks, female, provided by L. Li). In brief, bones were surgically removed and cleaned of surrounding tissue, washed once in 70% ethanol, once in PBS, 2.5 mM EDTA and penicillin–streptomycin and stored in PBS, 2.5 mM EDTA and penicillin–streptomycin. The bones were cut open and placed into a sterile 0.6 ml tube with holes at the base. The 0.6 ml tube with bones was placed into a sterile 2 ml tube and the nested tubes were centrifuged at 10,000g for 1 min. Obtained bone marrow pellets were resuspended in PBS + 2.5 mM EDTA and centrifuged at 400g for 3 min followed by RBC lysis using ACK buffer for 1 min at room temperature. Lysis was stopped by adding 10 vol of PBS + 2.5 mM EDTA and cells were pelleted at 400g for 3 min. A total of 4 × 106 cells was seeded per 10 cm suspension culture dish and differentiated in 10 ml cDMEM supplemented with 30% L929 supernatants containing M-CSF for 6 days. Cells were fed with 5 ml additional differentiation medium after 3 days. BMDMs were detached using PBS + 2.5 mM EDTA by incubating for 10 min at room temperature and reseeded for experiments at the required density and numbers in cDMEM supplemented with 10% L929 supernatants. Experiments were performed the next day.
Cells were treated with 10 ng ml−1 LPS O55:B5 (ALX-581-013, Enzo) or 100 ng ml−1 LTA-SA (tlrl-pslta, Invivogen) unless otherwise indicated. Other treatments were performed with 100 ng ml−1 Pam2CSK4 (P2C; tlrl-pm2s, Invivogen), 100 ng ml−1 Pam3CSK4 (P3C; vac-pms, Invivogen) 100 μg ml−1 poly(I:C) (tlrl-pic, Invivogen), 1 µg ml−1 Fla-ST (tlrl-epstfla, Invivogen), 100 ng ml−1 R848 (vac-r848, Invivogen), 5 µM CpG-DNA (ODN1826; tlrl-1826, Invivogen), 50 ng ml−1 mIL-33 (3626-ML, R&D Systems), 50 μg ml−1 DMXAA (tlrl-dmx, Invivogen) or 20 ng ml−1 mIFNγ (315-05, Peprotech). RIG-I–MAVS pathway activation used Sendai virus (SeV, Cantell strain) supplied as allantoic fluid (10100774, Charles River Laboratory). Cells were treated with 20 µM Zimlovisertib/PF-06650833 (HY-19836, MedChemExpress) to inhibit IRAK4 kinase activity. Cells were stimulated with 50 ng ml−1 recombinant mouse TNF (315-01A, Thermo Fisher Scientific) to activate NF-κB activity or kept in the presence of 200 ng ml−1 mouse sTNFRII (426-R2-050, R&D Systems) to block autocrine TNF signalling. To deplete PI(4,5)P2, cells were treated with 10 µM ionomycin (I24222, Invitrogen). The MyD88 inhibitors MPP (13863, Cayman Chemical) and TSI-13-57 (BD01861091, BLD Pharm) were used at 5 µM for in vitro experiments. Secretion of cytokines was blocked by treatment with 5 µg ml−1 brefeldin A (420601, BioLegend) and TLR endocytosis was blocked by treatment with 10 µM hydroxy Dynasore (Dyngo-4a; HY-13863, MedChemExpress).
Bacteria and virus infections
All bacterial strains used in this study are summarized in Supplementary Table 1. Staphylococcus aureus (strain SA113), S. Typhimurium (strain SL1344) and E. coli K12 (strain MG1655) were cultured on Luria Broth (LB; 12-780-052, Invitrogen) agar (1.5% w/v; 214010, BD) at 37 °C. Listeria monocytogenes (strain 10403S) was cultured on brain–heart infusion (BHI; 237500, BD) agar at 37 °C. Lactococcus lactis (strain NCTC 6681) was cultured on BHI agar at 37 °C under 5% CO2.
To prepare bacteria for infection, S. aureus, S. Typhimurium and E. coli colonies were inoculated into liquid LB and grown overnight at 37 °C. Listeria monocytogenes colonies were inoculated into liquid BHI and grown overnight at 30 °C without shaking. Lactococcus lactis colonies were inoculated into liquid BHI supplemented with 0.042% (w/v) NaHCO3 and grown overnight at 37 °C. Stationary phase S. aureus and E. coli in liquid cultures were then subjected to two back dilutions in liquid LB to enrich for highly viable, logarithmic phase cultures for infections. Salmonella enterica serovar Typhimurium was back diluted twice into LB supplemented with 300 mM NaCl and grown in a closed container. Lactococcus lactis was back diluted once into BHI supplemented with 0.042% (w/v) NaHCO3 and L. monocytogenes overnight cultures were used right away.
At the time of infection, bacteria were centrifuged at 10,000g for 3 min, supernatants discarded, pellets washed with PBS twice and resuspended in cDMEM without antibiotics. Bacterial suspensions were serially diluted and plated to confirm the MOI. To initiate synchronized infections, the target cells (0.5 × 106 cells per well of a 24-well plate) were placed onto ice for 5 min, the culture supernatants removed, exposed to 1 ml bacteria at the indicated MOI and centrifuged at 750g for 10 min. Infected cells were incubated for 1 h at 37 °C under 5% CO2 to allow for bacterial uptake, after which the cells were washed with warm PBS and the medium was replaced with cDMEM supplemented with 100 μg ml−1 gentamicin (15750060, Gibco) to remove and eliminate remaining extracellular bacteria. Infections were allowed to proceed at 37 °C with 5% CO2 for the remaining time.
For SeV (Cantell strain) infection, SeV was diluted in DMEM supplemented with 1% FBS for an MOI of 0.00001. Cells were exposed to SeV for 1 h at 37 °C with 5% CO2. After 1 h, cells were washed twice with PBS, the medium was replaced with 1 ml cDMEM and infection incubated at 37 °C with 5% CO2 for the remaining time.
Lentiviral transductions
Plasmid transfections were performed by using a 10× mix in Opti-MEM I reduced serum medium (31985062, Thermo Fisher Scientific) containing DNA plasmids and Lipofectamine 2000 transfection reagent (11668019, Thermo Fisher Scientific) at a 1:2 ratio according to the manufacturer’s instructions. For lentiviral packaging, HEK293T cells were transfected with equimolar ratios of the transfer plasmid, pMD2.G (Addgene, 12259) and psPAX2 (Addgene, 12260, both gifts from D. Trono). The medium was replaced after 16 h with cDMEM containing 5 mM sodium butyrate (B5887, Sigma-Aldrich) and the cells were left to produce lentiviral particles for over 24 h. The virus-containing supernatant was filtered through a 0.43-μm syringe filter and supplemented with 5 μM polybrene (TR-1003, Sigma-Aldrich). The target cells were resuspended in virus-containing medium and spinfected for 1 h at 1,250g and 30 °C. After centrifugation, 1 ml cDMEM was added and the cells left to rest.
Myddosome flow cytometry sorting and RNA-seq analysis
Historically, most flow cytometers use a 488 nm (blue) laser as standard configuration for SSC detection because the 488 nm wavelength provides robust sensitivity to intracellular refractive index heterogeneity and μm-scale structures. We therefore assessed the formation of cytosolic myddosomes by flow cytometry using SSC. After stimulation, endogenously GFP-tagged MyD88 in MyD88–AGF iBMDMs forms abundant μm-scale cytoplasmic complexes that increase intracellular heterogeneity and granularity. The accumulation of these protein-dense assemblies results in measurable changes in SSC.
SSC shift was used to separate myddosome-containing cells from non-responding bystander cells. To do so, PAMP-stimulated cells were collected, washed in PBS, resuspended in PBS + 1% FCS to a concentration of 4 × 106 cells per ml and passed through a 35 µm nylon mesh strainer. Subsequently, cells were sorted using a BD FACSMelody Cell Sorter using BD FACSChorus software (BD Biosciences) by gating on singlets and SSChigh versus SSClow cells.
To assess separation of subpopulations, we sorted cells as described to obtain 1 × 106 cells for each bin and collected the cells into 5 ml flow cytometry tubes containing 0.5 ml cDMEM, pelleted the cells by centrifugation at 400g, washed them once with cDMEM and subsequently plated onto six-well glass-bottom imaging plates (P06-1.5H-N, Cellvis). Cells were centrifuged at 400g for 5 min to speed up adhering and immediately imaged to test for presence/absence of cytosolic myddosomes. Alternatively, collected cells were lysed and analysed by immunoblotting.
To analyse myddosome-induced transcriptional changes by RNA-seq, eight 10-cm tissue-culture-treated dishes were seeded each with 2.0 × 107 MyD88–AGF iBMDMs and the cells stimulated with 10 ng ml−1 LPS. After 6 h, cells were washed with PBS and prepared for sorting as described above. All of the subsequent steps were carried out at 4 °C. First, LPS-stimulated cells were sorted into responder and bystander populations based on SSChigh and SSClow bins. Subsequently, LPS-treated cells and untreated controls were passed through the flow cytometer without gating to obtain pooled samples for comparison. Then, 1 × 106 cells for each sample/replicate/bin were collected into 5 ml flow cytometry tubes containing 0.5 ml TRIzol LS (10296028, Thermo Fisher Scientific). Total RNA was extracted according to the manufacturer’s instructions. RNA pellets were resuspended in 100 μl 1× TurboDNase buffer and potential gDNA contaminants were removed by adding 5 μl TurboDNase (AM2238, Thermo Fisher Scientific) and incubating at 37 °C for 45 min. RNA was repurified by phenol–chloroform extraction and ethanol precipitation. In brief, RNA solutions were transferred to 1.5 ml low bind tubes, and 100 μl double-distilled H2O and 200 μl (1 vol) of phenol:chloroform:isoamyl alcohol (25:24:1, v/v; 15593031, Thermo Fisher Scientific) were added. The samples were thoroughly mixed and centrifuged at 12,000g for 5 min. The aqueous phase (about 250 μl) was transferred to a new tube, and 1 vol of chloroform was added. The samples were thoroughly mixed and centrifuged again, and chloroform extraction repeated two more times. Finally, the RNA was precipitated by adding 1 ml ice-cold 100% ethanol and 1 μl glycol blue (5 μg; AM9515, Thermo Fisher Scientific) and centrifugating at 21,000g for 1 h at 4 °C. The obtained RNA pellets were washed twice with 80% ethanol, dried for 10 min at room temperature and resuspended in 100 μl double-distilled H2O. The RNA quantity and quality were assessed by spectrophotometry (NanoDrop One, Thermo Fisher Scientific) and confirmed by fragment analysis showing RNA-integrity numbers of >7.1 for all of the samples (performed by Genewiz). RNA-seq and initial data processing was performed by Genewiz and further data analysis was performed using the Galaxy Server60. Adapter trimming was conducted using Cutadapt (Galaxy v.3.7), applying a minimum read length of 20 bp and Phred quality score of >20. Reads were filtered using fastp (Galaxy v.1.1.0) with poly(G) trimming. Trimmed paired-end reads were aligned to the mouse reference genome (GRCm38, mm10) using HISAT2 (Galaxy v.2.2.2). Alignment statistics were evaluated with MultiQC, showing mapping rates exceeding 90%. Gene-level expression quantification was performed using featureCounts (Galaxy v.2.1.1). Additional alignment quality control was performed using RSeQC, including gene body coverage analysis (Galaxy v.5.0.3), and read distribution across genomic features (Galaxy v.5.0.3). Duplicate reads were evaluated using Picard MarkDuplicates (Galaxy v.3.1.1.0), and chromosomal mapping statistics were generated using SAMtools idxstats (Galaxy v.2.0.8), together demonstrating low duplication levels, even gene body coverage, predominant exon mapping and high-quality paired-end alignments. Differential expression analysis was conducted using DESeq2 (Galaxy v.2.11.40.8). The samples were grouped by condition with four replicates per group. Low-count genes were prefiltered before normalization. PCA and MA plots were generated to assess sample relationships and differential expression patterns. Genes with Padj < 0.01 and |log2[FC]| > 1 were considered to be differentially expressed. Per-gene z scores were calculated, and heat maps were generated using Morpheus (https://software.broadinstitute.org). Hierarchical clustering was performed using 1 − Pearson correlation as the distance metric and average linkage as the clustering method. RNA-seq data are available from NIH Gene Expression Omnibus (GEO) under identifier GSE321692.
Genome-wide CRISPR KO screen for myddosome assembly regulators
To express Cas9 in the parental MyD88–AGF iBMDMs, the sgRNA expression cassette was removed from pLentiCRISPR-v2 (52961, Addgene; a gift from F. Zhang)61 by digesting with KpnI-HF (R3142, NEB) and EcoRI-HF (R3101, NEB) and religating with annealed repair oligonucleotides that restore the restriction sites using Quick Ligation Kit (M2200, NEB). Next, the resulting plasmid was digested with BamHI-HF (R3136, NEB) and RsrII (R0501, NEB) to remove the PuroR gene and a bsr open reading frame (ORF) conferring blasticidin resistance amplified from pLenti-dCAS-VP64_Blast (61425, Addgene; a gift from F. Zhang)55 inserted using Gibson assembly to obtain pLenti-CRISPR-Blast-no-sgRNA. MyD88–AGF WT iBMDM were transduced with lentiviral particles and selected with 15 μg ml−1 blasticidin S (A1113903, Thermo Fisher Scientific). Successful transduction was validated by immunoblotting.
To generate an NF-κB activity reporter plasmid, a DNA sequence containing a synthetic poly(A) signal, a transcriptional pause site (to reduce background expression) and a minimal promoter was synthesized by gBlock (Integrated DNA Technologies, IDT) and a pentameric NF-κB response element (5×NF-κB-RE) was added to the promoter through annealed oligonucleotides by performing overlap extension PCR using Q5 DNA polymerase (M0491, NEB). An mScarlet-3 ORF was synthesized by gBlock and the two fragments were inserted into EcoRI-HF- and BamHI-HF-digested pLenti-P2A-PuroR62 by Gibson assembly (E2611, NEB). The resulting pLenti-mScarlet3-NF-κB-reporter plasmid was confirmed by whole-plasmid sequencing (Plasmidsaurus) and transduced into MyD88–AGF WT + Cas9 iBMDMs with lentiviral particles. Transduced cells were stimulated with 50 ng ml−1 recombinant TNF for 6 h, sorted for uniform NF-κB-induced mScarlet3 expression and rested for >1 week, in the presence of 15 μg ml−1 blasticidin S to maintain Cas9 expression until fluorescence had faded.
To validate the functionality of the NF-κB reporter and Cas9, cells were electroporated with recombinant sgRNAs against Tlr2, Tlr4, Myd88 or non-targeting control (IDT) using the Neon Transfection System (Thermo Fisher Scientific). In brief, 1.2 × 106 cells were resuspended in 120 μl buffer R, 2 μl 100 µM sgRNA was added and the cells electroporated using 100 μl electroporation tips, 1,400 V, 2 pulses, 10 ms each. Cells were immediately placed into wells of a six-well plate containing 3 ml cDMEM. The medium was replaced after 4 h with fresh cDMEM + 15 μg ml−1 blasticidin S to remove dead cells and maintain Cas9 expression. Then, 48 h later, the cells were collected and the KO efficiency was assessed by flow cytometry (TLR2/4) or immunoblot analysis (MyD88). Moreover, 0.2 × 106 cells were seeded per well of a black-wall 96-well plate (12 wells for each genotype) and left to adhere for >4 h. Cells were then treated in triplicates with LPS, LTA and TNF in 200 μl Fluobrite DMEM (A1896701, Gibco) supplemented with 20 mM HEPES or left unstimulated. Plates were sealed and induction of mScarlet-3 fluorescence measured over 6 h on a Tecan Spark 10 M Multi-Mode Plate Reader heated to 37 °C (operated using Tecan SparkControl software). Fluorescence was recorded every 10 min and the resulting kinetic analysed with GraphPad/PRISM. At the end point, cells were collected, pooled from the replicate wells and single-cell fluorescence induction was analysed by flow cytometry.
To execute the screen, we used the recently published Julianna library34. Plasmids of the pooled Library were obtained from Addgene (247028, a gift from J. Doench and D. Root) and propagated in ElectroMAX Stbl4 Competent Cells (11635018, Thermo Fisher Scientific) according to the depositor’s instructions. After DNA preparation by Maxiprep (740414.50, Macherey-Nagel), the plasmids were further purified to remove any contaminants. In brief, around 200 μg plasmid DNA (pDNA) was placed into a 1.5 ml tube and 1 vol 2× protein digestion buffer (final concentration 10 mM Tris pH 8.0, 1 mM EDTA and 0.5% SDS) was added. pDNA solution was supplemented with proteinase K to a final concentration of 0.1 mg ml−1 (25530049, Thermo Fisher Scientific) and incubated at 37 °C for 30 min. Subsequently, pDNA was purified by phenol–chloroform extraction and ethanol precipitation (see above) and resuspended to a concentration of 1 μg μl−1 in double-distilled H2O. Library virus production followed the same protocol as described above using eight 10-cm dishes with HEK293T cells. Virus was produced in 7 ml cDMEM + 5 mM butyrate per dish for 72 h. Virus-containing supernatants were collected, centrifuged for 3 min at 400g to remove cellular debris, filtered through a 0.43 µm syringe filter, aliquoted and stocks were stored at −80 °C until use.
Viral tittering before library creation optimized the transduction conditions. Increasing amounts of Julianna library virus stocks (0–300 μl) were added to 3 × 106 cells in cDMEM with 5 μg ml−1 polybrene in a constant final volume of 2 ml per well of a 12-well plate. Transduction was performed by centrifugation at 1,250g for 1 h at 30 °C. After spinfection, 2 ml of cDMEM was added. The next day, 0.1 × 106 cells from each condition were replated into two wells of a six-well plate each and half the wells supplemented with 4 μg ml−1 puromycin. Test selection was carried out over 5 days. At the end of the selection, viable cells were counted and transduction efficiency calculated as the ratio of viable cells from the conditions with or without puromycin. The target was around 30% transduction efficiency to ensure a maximum of 0.5–1 infections per cell and that the final enrichment analysis can correlate guide abundance to a single perturbation63.
To guarantee statistical robustness during enrichment analysis, the downstream screening assays should maintain >500 cells per sgRNA. For our library creation, a total of 4.8 ml of virus stock was used to transduce 1.5 × 108 cells. Cells were transduced as described above and, 24 h after transduction, the cells were reseeded onto six 15 cm tissue culture dishes. After 1 day, selection was started using 4 μg ml−1 puromycin and 15 μg ml−1 blasticidin S. The selection medium was replaced every 2 days and cells expanded as necessary resulting in 15 confluent 15-cm tissue culture dishes with around 1 × 109 cells after 1 week, when untransduced controls had died—marking the successful creation of iBMDM MyD88–AGF + 5×NFkB-RE-mScarlet3 + Cas9 +Julianna library cells (Julianna cells).
From these, 0.9 × 109 cells were seeded on a total of eighteen 15-cm tissue culture dishes and left to rest overnight. Cells on six dishes each were either treated with 10 ng ml−1 LPS with 100 ng ml−1 LTA or left untreated. After 6 h of PAMP stimulation, cells were collected by detaching with 8 ml PBS + 2.5 mM EDTA per dish and all cells from one condition were pooled into 50 ml tubes. Cells were pelleted by centrifugation at 400g for 5 min, washed with PBS and fixed by resuspending the pellets in 100 ml of PBS + 0.04% BSA + 4% PFA and left to fix at room temperature for 15 min while mixing continuously with end-over-end rotation. Fixed cells were pelleted by centrifugation at 400g for 5 min, resuspended in 20 ml PBS + 0.04% BSA (cell density of around 1.5 × 107 cells per ml) passed through a 70 µm nylon mesh strainer, aliquoted into 5 ml flow cytometry tubes and stored at 4 °C. Cells were sorted on a BD FACSMelody Cell Sorter as described above by gating on singlets and NF-κBlowSSClow cells (around 15% lowest for each). We obtained around 7.5–10 × 106 cells per condition and these were collected into PBS + 1 mM EDTA + 30% FBS. All 3 × 108 cells from the untreated condition were saved as input control.
Genomic DNA was extracted by resuspending the cells at a density of 1.0 × 107 in nucleus lysis buffer (A7941, Promega) and adding proteinase K to a final concentration of 0.1 mg ml−1. The samples were thoroughly mixed by vortexing and incubated 55 °C overnight to reverse PFA cross-linking of chromatin. The next morning RNase A (A7973, Promega) was added to a final concentration of 20 µg ml−1, the samples mixed and incubated at 37 °C for 30 min. The samples were chilled on ice, one-third vol protein precipitation solution (A7951, Promega) was added and the samples were mixed by vortexing until they turned visibly cloudy. Proteins were precipitated by incubating on ice for 5 min followed by centrifugation at 4 °C and 16,000g for 5 min. The cleared supernatants were transferred into 1.5 ml tubes, and DNA was precipitated by adding 1 vol isopropanol. Tubes were inverted 50 times to mix and incubated at −20 °C for 15 min. Genomic DNA was pelleted by centrifugation at 4 °C and 21,000g for 20 min, washed twice with 70% ethanol, air dried and resuspended in double-distilled H2O to a concentration of 1 μg μl−1.
Next, the locus encoding the sgRNA was amplified by PCR using the following primers Fwd: 5′-GGACTATCATATGCTTACCGTAA-3′; rev: 5′-GTTGATAACGGACTAGCCTTAT-3′ and Q5 DNA polymerase. As each diploid nucleus contains around 6 pg of gDNA and 500× sgRNA coverage needs to be maintained, the PCRs used 210 μg of gDNA from the control sample and all gDNA from the LPS/LTA-treated and enrichment samples (about 70 μg each). For comparison, the locus was also amplified from 80 ng input pDNA. Each PCR reaction used 2.5 µg DNA per 50 μl reaction resulting in linear amplification over 28 cycles (annealing temperature, 58 °C; and extension time, 30 s). The resulting amplicons were gel purified and extracted using the crush and soak method. In brief, amplicon bands were excised and crushed into bits by passing through a 20 G needle. Crushed gel bits were soaked in 3 vol of soak buffer (300 mM sodium acetate + 1 mM EDTA pH 8.0) at 4 °C overnight with end-over-end rotation. The next day, gel suspensions were filtered through a 40 µm nylon mesh strainer and 3 vol of ice-cold 100% ethanol was added. The samples were inverted 50 times, incubated at −80 °C for 10 min and centrifuged at 7,000g for 2 h at 4 °C. DNA pellets were washed twice with 70% ethanol, air-dried and resuspended in 100 μl double-distilled H2O. To ensure sufficient DNA purity, the samples were further purified using phenol–chloroform re-extractions as described above. The final samples (library input control, untreated control, LPS-treated and enriched, LTA-treated and enriched) were sent to Genewiz to perform library preparation and high-throughput sequencing.
All further data analysis was performed using Galaxy Server. Adapter trimming was performed using Cutadapt (Galaxy v.5.2) with the 5′ adapter sequence 5′-TTGTGGAAAGGACGAAACACCG-3′, discarding untrimmed reads and filtering for a minimum read length of 20 bp. This step selectively retained reads containing the Julianna plasmid-derived sequence upstream of the sgRNA cassette and ensures that reads began directly with the 20 bp sgRNA protospacer. Quality assessment of trimmed reads was performed using FastQC (Galaxy v.0.74), incorporating a custom adapter reference containing the sgRNA scaffold sequence immediately 3′ of the protospacer (5′-GTTTAAGAGCTA-3′). The sgRNA scaffold adapter was detected in over 90% of reads beginning at position 21, confirming correct library structure and trimming.
sgRNA abundance quantification was performed using MAGeCK count (Galaxy v.0.5.8.4) with trimmed and filtered fastq files and the corresponding sgRNA library annotation to generate a count summary and normalized counts. The sequencing depth was sufficient across all samples, with the minimum sgRNA coverage exceeding 350×. Mapping efficiency to the sgRNA library exceeded 87%, zero-count sgRNAs represented <0.1% of the library and Gini indices below 0.05 indicated that a high library complexity was maintained through all steps. sgRNA- and gene-level enrichment analyses were performed using the MAGeCK test (Galaxy v.0.5.8.1). Analyses were conducted for LPS versus untreated, LTA versus untreated as well as plasmid input versus untreated samples, producing ranked gene and sgRNA enrichment summaries. Guide dropout for essential genes was compared against datasets from ref. 64. Gene lists were annotated to highlight non-targeting controls and known TLR signalling factors. Enrichment score distributions were visualized using GraphPad/Prism after log-transformation of positive selection scores.
RNA-seq analysis of primary WT and Ticam2
−/− BMDMs
RNA was extracted from 0.5 × 106 cells differentiated and treated as described above using PureLink RNA Mini Kit (12183025, Thermo Fisher Scientific). During purification, on-column DNase treatment using PureLink DNase (12185010, Thermo Fisher Scientific) was performed to remove contaminating gDNA according to the manufacturer’s instructions. BMDMs were derived from three WT and three Ticam2−/− mice representing biological replicates per genotype/treatment condition. The RNA quantity and quality was assessed by spectrophotometry (NanoDrop One, Thermo Fisher Scientific).
RNA-seq and initial data processing were performed by Plasmidsaurus. In brief, raw reads were filtered using fastp v.0.24.0 with poly(X) tail trimming, 3′ quality-based tail trimming, Phred quality score > 15, and a minimum length > 50 bp. Filtered reads were aligned to the reference genome (GRCm38, mm10) using STAR aligner v.2.7.11 with non-canonical splice junction removal, followed by coordinate sorting using samtools v.1.22.1. PCR and optical duplicates were removed using UMI-based deduplication with UMIcollapse v.1.1.0. Gene-level expression quantification was performed using featureCounts (subread package v.2.1.1) with strand-specific counting, multi-mapping read fractional assignment, exons and three prime UTR as the feature identifiers, and grouped by gene_id. Differential expression analysis was done with edgeR v.4.0.16 using filtering for low-expressed genes. Sample-sample correlations and PCA were calculated on normalized counts (trimmed mean of M-values, TMM) using Pearson correlation.
For creating heat maps, DEGs were extracted from the dataset with a threshold of Padj < 0.05 and |log2[FC]| > 1. Subset analysis of PRGs, SRGs and ISGs was performed using comparisons against gene lists compiled from refs. 25,26,27,59 and Hallmark IFNα/IFNγ response datasets (MSigDB)65. Per-gene z scores were calculated, and heat maps were generated using Morpheus (Broad Institute). Hierarchical clustering was performed using 1 − Pearson correlation as the distance metric and average linkage as the clustering method. RNA-seq data are available from NIH GEO under identifier GSE321707.
Generation of KO cells using CRISPR–Cas9
Guide RNA (gRNA) sequences targeting the genes of interest were designed using CRISPick63. Complementary DNA oligonucleotides were annealed and cloned into BsmBI-v2-digested (R0739, NEB) pLentiCRISPR-v2 and transduced into target cells using lentiviral particles. After selection with 5 µg ml−1 puromycin, the sgRNA-efficiency was assessed by Sanger sequencing of the gDNA surrounding the sgRNA targeting site followed by CRISPR ICE analysis (EditCo Bio). Knockout pools were subcloned by serial dilution into 96-well plates. The obtained clones were screened for absence of the target proteins by immunoblotting. For creating ∆Ticam1 cells, KO was reconfirmed by ICE analysis with scores over 95% for each clone, as no antibody is available. For each cell line, over five clones with confirmed absence of target proteins were pooled before undergoing a final screen using CRISPR ICE analysis and immunoblotting. All crRNA sequences and CRISPR ICE genotyping primers are provided in Supplementary Tables 2, 3.
Generation of p65–mScarlet3-expressing cells
To generate an mScarlet3-p65 expression plasmid, the mScarlet3 ORF was amplified from pLenti-mScarlet3-NF-κB-reporter plasmid and the p65 ORF from RelA-cFlag-pcDNA3 (20012, Addgene; a gift from S. Smale)66 using Q5 DNA polymerase and cloned into pLenti-P2A-Zeo1 using Gibson assembly. Successful cloning was confirmed by whole-plasmid sequencing. The obtained pLenti-mScarlet3-p65-P2A-Zeo plasmid was transduced into target cells with lentiviral particles and the cells selected with 200 µg ml−1 Zeocin (R25001, Thermo Fisher Scientific). Surviving cells were sorted for uniform mScarlet3-p65 expression and successful transduction validated by immunoblotting. To validate functionality of the reporter, cells were treated with recombinant TNF for 1 h and live imaged by confocal microscopy, ensuring a homogenous and comparable response between cell lines.
Generation of TRAM-expressing cells
To generate TRAM WT and mutant expression plasmids, the TRAM ORF within pLenti-mTRAM-HA-P2A-Puro was first rendered CRISPR-resistant using site-directed mutagenesis and the resistance cassette was swapped for a neo (neomycin resistance) gene to confer mammalian resistance to G418 (Geneticin). This resulted in pLenti-mTRAM-HA-P2A-Neo and successful cloning was confirmed by whole-plasmid sequencing. Next, the TRAM ORF was mutated using site-directed mutagenesis to generate the G2A, 3X, E180A, C114H and D88A/E89A mutants. To generate expression plasmids for chimeric TRAM versions, the TRAM ORF was removed from pLenti-mTRAM-HA-P2A-Neo by restriction digest with XhoI (R0146, NEB) and XbaI (R0145, NEB). The TRAM ORF was amplified from the original plasmid using Q5 DNA polymerase but removing its N-terminal plasma membrane-binding domain (amino acids 1–39) and adding the sequence encoding the first 15 amino acids of either Fyn or Lyn by primer extension. Moreover, the sequence encoding the TIRAPPBM (amino acids 1–102) was amplified from pcDNA3-mTIRAP (J.C.K. Laboratory). Finally, the empty backbone and amplicons were assembled using Gibson assembly, resulting in pLenti-P2A-Neo plasmids containing the ORFs of Fyn–TRAM, Lyn–TRAM or TIRAPPBM–TRAM and successful cloning was confirmed by whole-plasmid sequencing. The obtained plasmids were transduced into MyD88–AGF ∆Ticam2 or ∆Tirap/Ticam2 iBMDMs with lentiviral particles and the cells selected with 1.5 mg ml−1 G418 (ant-gn-5, Invivogen). Successful transduction was validated by immunoblotting.
Generation of TIRAP-overexpressing cell lines
To generate TIRAP WT and ∆TIR mutant expression plasmids, the TIRAP ORF was amplified from pcDNA3-mTIRAP (J.C.K. laboratory) with primers adding a C-terminal HA-tag and either amplifying the entire ORF or truncating the last 156 amino acids (∆TIR) using Q5 DNA polymerase and inserted into BamHI-HF and XhoI-digested pLenti-P2A-PuroR using Gibson assembly. This resulted in pLenti-mTIRAP-HA-P2A-Puro or pLenti-mTIRAP(∆TIR)-HA-P2A-Puro, which were confirmed by whole-plasmid sequencing. The obtained plasmids were transduced into iBMDM MyD88–AGF WT target cells with lentiviral particles and the cells selected with 5 µg ml−1 puromycin. Successful transduction was validated by immunoblotting and overexpression was assessed after RNA extraction and RT–qPCR.
Generation of PH-PLCδ–mScarlet3-expressing PI(4,5)P2-reporter cell lines
To generate a plasmid expressing a marker for PI(4,5)P2 localization in live-cells, the ORF of the PIP2-binding domain of human PLCδ1 (PH-PLCδ1; amino acids 1–170) was synthesized by gBlock (Integrated DNA Technologies, IDT), the mScarlet3 ORF was amplified from pLenti-mScarlet3-NF-κB-reporter plasmid using Q5 DNA polymerase and the fragments were inserted into BamHI-HF-digested pLenti-P2A-PuroR using Gibson assembly. Successful cloning was confirmed by whole-plasmid sequencing. The obtained pLenti-PH-PLCδ1-mScarlet3-P2A-Puro plasmid was transduced into all target cells (Supplementary Table 1) with lentiviral particles and the cells selected with 5 µg ml−1 puromycin (R25001, Thermo Fisher Scientific). Surviving cells were sorted for homogenous mScarlet3 fluorescence across cell lines and successful transduction validated by fluorescence imaging.
Implementing CRISPR-SAM for activating Tirap and Tram genes in mouse macrophages
We implemented the CRISPR-SAM system55 in mouse macrophages. This three-component gene activation (CRISPRa) tool can boost transcription of endogenous genes. To do so, MyD88–AGF iBMDMs were transduced with pLenti-dCAS-VP64_Blast using lentiviral particles and selected with 15 μg ml−1 blasticidin S. Successful transduction was validated by immunoblotting.
Next, complementary DNA oligonucleotides encoding the crRNA sequences of the activation gRNAs (Supplementary Table 2) were annealed and cloned into BsmBI-v2-digested pXPR_502 (96923, Addgene; a gift from J. Doench and D. Root)67 encoding the modified gRNA scaffold as well as the MS2-P65-HSF1 activator complex. Obtained plasmids were transduced into MyD88–AGF +dCas9 iBMDMs using lentiviral particles and selection performed using 15 μg ml−1 blasticidin and 5 μg ml−1 puromycin. However, transductions of iBMDMs with pXPR_502 were unproductive. Note that our standard lentivirus production procedure yielded 1,000× lower titres for pXPR_502 when compared side by side with a control (pLentiCRISPR-v2; determined using Lenti-X GoStix Plus; 631281, Takara). Furthermore, the hPGK promoter driving the activator expression appeared to be non-functional (inactivated) in iBMDMs as we were unable to detect any transcripts by qPCR.
We decided to abandon pXPR_502 and instead transfer the cassettes encoding the modified sgRNA and the MS2-P65-HSF1 activator complex into the pLentiCRISPR-v2 backbone, which produced superior titres. First, the Cas9 ORF in pLentiCRISPR-v2 was replaced with the activator complex encoding ORF from pXPR_502. For this, pLentiCRISPR-v2 was digested with AgeI-HF (R3552, NEB) and BamHI-HF, the activator ORF amplified from pXPR_502 using Q5 polymerase and Gibson assembly performed to obtain an intermediate plasmid now expressing the MS2-P65-HSF1 activator complex from an EF1α promoter. Second, the hU6 and sgRNA cassette in pLentiCRISPR-v2 was replaced with the modified sgRNA expression cassette from pXPR_502. For this, the intermediate plasmid was digested with PacI (R0547, NEB) and EcoRI-HF, the hU6/modified sgRNA cassette from pXPR_502 amplified using Q5 polymerase and the final pLenti-CRISPRa-empty plasmid was assembled by Gibson assembly. Successful cloning was confirmed by whole-plasmid sequencing. Next, the annealed oligos encoding the activation sgRNAs for Tirap or Ticam2 were inserted and the resulting plasmids or EV transduced into MyD88–AGF + dCas9 iBMDMs either individually or in binary combinations using lentiviral particles and cells selected using 15 μg ml−1 blasticidin and 5 μg ml−1 puromycin. This resulted in successful activation of Tirap and/or Ticam2 genes as detected by RT–qPCR and immunoblotting.
Flow cytometry and cell sorting
For analysis of TLR surface localization, 2 × 106 iBMDMs were seeded per well of a six-well plate and treated as described above. Cells were collected by scraping in ice-cold PBS, washed twice with PBS, passed through a 35 µm nylon mesh strainer (352235, Corning) and resuspended in PBS containing mouse TruStain FcX PLUS (anti-mouse CD16/32; 156604, BioLegend) and primary antibodies against TLRs or isotype control (Supplementary Table 4). Cells were stained by incubation at 4 °C for 30 min in the dark, washed twice with ice-cold PBS and analysed on a LSRFortessa flow cytometer using FACSDiva software (BD Biosciences). The proportion of surface localized TLRs was determined by calculating the MFI using FlowJo v.10.8.2 and comparing with untreated cells and TLR-KO controls after gating for single cells. Analysis of NF-κB reporter activity also used live cells that were stained with LIVE/DEAD Fixable Violet Dead Cell Stain Kit (L34963, Thermo Fisher Scientific) according to the manufacturer’s instructions. The induction of the fluorescent NF-κB reporter was plotted as histograms of MFI using FlowJo v.10.8.2 after gating for live cells and single cells.
For sorting cells based on fluorescent protein expression, cells were collected, washed in PBS, resuspended in PBS + 1% FCS to a concentration of 4 × 106 cells per ml and passed through a 35 µm nylon mesh strainer. Subsequently, cells were sorted into 500 μl complete medium on a BD FACSMelody Cell Sorter using BD FACSChorus software (BD Biosciences) by gating on singlets and fluorescent-protein-positive cells. Immediately after sorting, cells were seeded into six-well plates with complete medium.
SDS–PAGE and immunoblotting
For immunoblotting, 1 × 106 cells were seeded per well of a 24-well plate and stimulated as described above. Cells were washed with ice-cold PBS and lysed for 15 min on ice in 250 μl RIPA buffer (25 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with protease inhibitors (Protease Inhibitor Cocktail set III, EDTA free; 539134, Millipore) and 1 mM activated sodium orthovanadate (567540, Sigma-Aldrich), 5 mM sodium fluoride (201154, Sigma-Aldrich) and 3 mM β-glycerophosphate (G9422, Sigma-Aldrich) phosphatase inhibitors. The lysates were cleared by centrifugation at full speed for 10 min at 4 °C in a table-top centrifuge. Then, 15–20 μg of total protein per sample was mixed with 5× loading dye (125 mM Tris-HCl pH 6.8, 10% SDS, 50% glycerol, 0.06% bromophenol blue) supplemented with 5% β-mercaptoethanol, denatured at 95 °C for 10 min and separated by SDS–PAGE. For immunoblots against phosphorylated proteins, cells were washed with PBS and then lysed in 250 μl 1× loading dye supplemented with 5% β-mercaptoethanol. The cell lysates were homogenized by several passages through a 29 G needle, denatured at 95 °C for 10 min and separated by SDS–PAGE using home-made running buffer (25 mM Tris-HCl pH 8.3, 192 mM glycine, 0.1% (w/v) SDS) and 10%, 12% or 15% polyacrylamide gels for around 90 min at 140 V.
After SDS–PAGE, proteins were transferred onto methanol-activated PVDF membranes (IPVH00010, Millipore) using wet-transfer at 400 mA for 1 h in Towbin buffer (25 mM Tris-HCl pH 8.3, 192 mM glycine, 20% (v/v) methanol). Membranes were blocked with 5% BSA in TBS-T (0.05% Tween-20) for >30 min at room temperature. Incubation with primary antibodies (Supplementary Table 5) was performed at 4 °C overnight (typical dilution 1:1,000). Blots were developed by washing with TBS-T, probing with 1:10,000 diluted HRP-conjugated secondary antibodies (Supplementary Table 6), washing again and imaging on a ChemiDoc MP imaging system (Bio-Rad) with Image Lab 6.1 (Bio-Rad) software using SuperSignal West Pico PLUS (34580, Thermo Fisher Scientific) or SuperSignal West Femto Maximum Sensitivity Chemiluminescent Substrate (34096, Thermo Fisher Scientific). For reprobing membranes with different primary antibodies, membranes were stripped using Restore PLUS Western Blot Stripping Buffer (46430, Thermo Fisher Scientific). Densitometry analysis used quantification of protein band intensities with ImageJ/Fiji68 and normalization to the loading control and/or holo-enzyme when assessing phosphorylated protein species. To determine the activity of TLR-responsive signalling pathways, obtained values were used to calculate the FC versus untreated cells and the total signalling pathway activity was determined by measuring the area under the curve as implemented in GraphPad/Prism.
Immunoprecipitations
Immunoprecipitation of endogenous MyD88 was performed as previously described69 using 0.5 μg of anti-MyD88 antibody (Supplementary Table 5) and 15 μl (bed volume) of protein G Sepharose beads (for endogenous MyD88; P3296, Millipore) or 15 μl (bed volume) of anti-Flag(M2) agarose beads (for MyD88–AGF; A2220, Millipore). Immunoprecipitation of HA-tagged TRAM species used 15 μl (bed volume) of anti-HA agarose beads (26181, Thermo Fisher Scientific). Immunoprecipitation was performed by incubation at 4 °C for 4 h on a tube rotator and proteins were eluted using either 200 ng ml−1 3×Flag peptide (F4799, Sigma-Aldrich) in lysis buffer, by acidification with 50 μl of 0.2 M glycine pH 2.0 followed by an immediate wash with 50 μl of lysis buffer or by resuspending in 1× loading dye and heating at 95 °C for 15 min. The input, unbound and elution fractions were kept during immunoprecipitations and analysed by immunoblotting.
Analysis of gene expression by RT–qPCR
RNA was extracted from 0.5 × 106 cells using the PureLink RNA Mini Kit (12183025, Thermo Fisher Scientific). During purification, on-column DNase treatment using PureLink DNase (12185010, Thermo Fisher Scientific) was performed to remove contaminating gDNA according to the manufacturer’s instructions. The purified RNA was reverse transcribed using high-capacity cDNA synthesis kit (4368813, Thermo Fisher Scientific). qPCR used iTaq Universal SYBR Green Supermix (1725121, Bio-Rad), 20–40 ng cDNA in a 10 μl reaction and primers at 1 μM final concentration (the sequences are provided in Supplementary Table 7). Data were recorded on a CFX384 Real-Time system (Bio-Rad) and analysed using CFX Manager v.3.1 (Bio-Rad). Primer specificity was ensured by designing primers to span exon–exon junctions, whenever possible. Recorded Ct values were normalized to the recorded Ct of the mouse Hprt1 housekeeping gene, and data were plotted as ∆Ct (relative expression) or as the FC compared with the untreated controls.
Cytokine ELISAs and Cell Titer Glow assays
ELISAs were performed to measure TNF, IL-6, IL12-p40, IFNβ or IP-10 secretion. Cell culture supernatants were cleared of cell debris by centrifugation at 400g for 3 min. Cleared supernatants were diluted between 1:2 and 1:100 with ELISA diluent to be within the assay range. The concentrations of cytokines were measured according to the manufacturer’s protocols using uncoated mouse IL-6 (88-7064-88), uncoated mouse IL-12p40 (88-7120-88), uncoated mouse TNF (88-7324-88, all from Thermo Fisher Scientific), LumiKine Xpress mIFN-β (luex-mifnbv3, Invivogen) or mouse IP-10 (DY466, R&D Systems) ELISA kits.
Cell viability was quantified using the CellTiter-Glo Luminescent Cell Viability Assay (G7571; Promega) according to the manufacturer’s instructions. The background signal from medium-only wells was subtracted, and viability was expressed relative to the indicated control. ELISA absorbance and luminescent readings were measured with a Tecan Spark 10 M Multi-Mode Plate Reader.
Fixed immunofluorescence imaging
For imaging, 0.25 × 106 cells were seeded onto coverslips in 24-well plates. Cells were fixed with 4% methanol-free formaldehyde (50-980-487, EMS) added directly to the cell culture medium followed by incubation for 15 min at room temperature. Fixed cells were washed with PBS and kept at 4 °C overnight. Fixed specimens were permeabilized with PermQuench buffer (0.2% (w/v) BSA and 0.02% (w/v) saponin in PBS) for 30 min at room temperature and stained with primary antibodies (Supplementary Table 5) for 1 h at room temperature. After three washes with PBS, cells were incubated with dye-conjugated secondary antibodies (Supplementary Table 6) and 1 μg ml−1 DAPI (D1306, Thermo Fisher Scientific) diluted in PermQuench buffer for 1 h at room temperature. The coverslips were extensively washed with PBS, once with double-distilled H2O and mounted using 5 μl ProLong Glass Antifade Mountant (P36982, Thermo Fisher Scientific).
Imaging of prepared specimens was performed using a LSM 880 point-scanning confocal microscope with fast Airyscan controlled with Zeiss Zen Blue software (Zeiss Instruments). Imaging used a Plan Apochromat ×40/1.3 NA or a Plan Apochromat ×63/1.4 NA oil-immersion objective (Zeiss Instruments). In a typical experiment, 9–25 fields of view were imaged using multiposition acquisition and the following settings: bidirectional scanning, 2–4× frame averaging, 16 bit bit-depth, ROI 2,000 × 2,000 px and optical zoom 1 resulting in a pixel size of 65 nm px−1 and pinhole: 89.3, resulting in 0.5 µm sections. Acquired raw microscopy data were processed using ImageJ/Fiji to produce representative images and analysed for quantification of myddosome formation and localization.
Live-cell confocal microscopy
Live-cell imaging was performed using a LSM 880 point scanning confocal microscope (Zeiss Instruments) with an environment control chamber. Imaging used a Plan Apochromat ×63/1.4 oil immersion objective (Zeiss Instruments) and the following settings: bidirectional scanning, 2× line averaging, 16 bit bit-depth, ROI 2,000 × 2,000 px and optical zoom 1 resulting in a pixel size of 65 nm px−1, pinhole: 89.3, resulting in 0.5 µm sections and sampling every 60 s with minimal laser power to minimize acquisition bleaching. In a typical experiment, 10–20 fields of view were imaged simultaneously, and z focus was controlled every five acquisitions using the Definite Focus 2 system (Zeiss Instruments).
To prepare samples for live-cell imaging, 2 × 106 MyD88–AGF iBMDMs were seeded onto 35 mm dishes with an integrated coverslip (P35G-1.5-14-C, Mattek). If more than one cell line was imaged, 0.2 × 106 cells were seeded per well of an 8 high-well µ-Slide (80806, ibidi) instead. Then, 1 h before imaging, the culture medium was replaced with 2 ml prewarmed FluoroBrite DMEM supplemented with 20 mM HEPES. Simultaneously, the environmental control chamber on the confocal microscope was set to pre-heat to 37 °C and left to equilibrate. To image myddosome formation, the prepared cells were placed on the microscope and left to equilibrate for around 15 min. Myddosome formation was induced by PAMP-treating cells using a microfluidics setup followed by live imaging for up to 12 h. Similarly, MPP was added during live imaging to disrupt myddosomes. Raw imaging data were further processed as described below to quantify myddosome formation and localization.
Quantification of myddosome plasticity was performed and analysed using FRAP as described previously1,70. Myddosome formation was induced by LPS treatment for 1 h with subsequent addition of 10 µM MPP or DMSO control. The experiment was performed 5 min after MPP addition.
TIRF microscopy
TIRF microscopy was performed using a Nikon Ti inverted microscope equipped with an Andor Zyla 4.2 Plus sCMOS camera, an environment control chamber, and Nikon TIRF Lun-f Illumination for 488 nm excitation of MyD88–AGF. TIRF microscopy was performed using an Apo TIRF ×100/1.49 DIC N2 oil-immersion objective (Nikon) and the following settings: no pixel binning, 12-bit & Gain 4 bit depth, 150 ms exposure, 1% laser power, TIRF angles between 1955–2010, ROI 1,024 × 1,024 px resulting in a pixel size of 65 nm px−1 and sampling every 15 s. Raw imaging data were further processed as described below to quantify proto-myddosome formation.
Imaging analysis
Myddosome detection and quantification was performed as described previously1. Other object detection used Cellpose-SAM as implemented in Cellpose (v.4.0.8)71 to detect cells, nuclei were detected using the ImageJ/Fiji plugin for StarDist72,73. Proto-myddosome density from TIRF imaging was calculated by dividing the number of detected myddosomes by the cell-covered area. The average myddosome position over time was extracted by determining the centroid of each myddosome label within single cells. Membrane distance was calculated as minimal Euclidian distance to the plasma membrane, that is, the outline of the cell label containing the respective myddosome. Distances of less than 2 µm were considered as membrane proximal/associated.
Label images of nuclei and cells were used for tracking of objects using the ImageJ/Fiji plugin Trackmate74 and the simple LAP tracker. Tracking was performed using the following empirically determined parameters: max. linking distance, 8 µm; max. gap closing and max. distance, both 10 µm; and max. frame gap, 2. Detected nuclei and cell labels were used to analyse NF-κB activation in PAMP-treated iBMDMs. Peak analysis as implemented in GraphPad/PRISM was used to quantify the NF-κB activity peaks and oscillations.
MS analysis of endogenous mouse TIRAP
Immunoprecipitation of TIRAP from iBMDM WT was performed as described above but starting with 1.00 × 108 cells using 2 μg of primary antibody against TIRAP (Supplementary Table 5) and 40 μl (bed volume) of protein G Sepharose beads. After immunoprecipitation, beads were washed three times with lysis buffer and twice with PBS and retained proteins eluted by boiling in 50 μl 1× loading dye for 15 min. The entire eluate was run on a 12% SDS–PAGE gel and the gel stained with Coomassie Instant Blue (ab119211, Abcam). The band corresponding to the expected size of mouse TIRAP was excised and submitted for MS analysis by the HMS Taplin Biological Mass Spectrometry Facility.
In brief, excised gel bands were cut into approximately 1 mm3 pieces. Gel pieces were then subjected to a modified in-gel trypsin digestion procedure75. Gel pieces were washed and dehydrated with acetonitrile for 10 min followed by removal of acetonitrile. The pieces were then completely dried in a speed-vac. Rehydration of the gel pieces was with 50 mM ammonium bicarbonate solution containing 12.5 ng μl−1 modified sequencing-grade trypsin (Promega) at 4 °C. After 45 min, the excess trypsin solution was removed and replaced with 50 mM ammonium bicarbonate solution to just cover the gel pieces. The samples were then placed in a 37 °C room overnight. Peptides were later extracted by removing the ammonium bicarbonate solution, followed by one wash with a solution containing 50% acetonitrile and 1% formic acid. The extracts were then dried in a speed-vac (for around 1 h) and stored at 4 °C until analysis.
On the day of analysis, the 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 2.6 µm C18 spherical silica beads into a fused silica capillary (100 µm inner diameter × approximately 30 cm length) with a flame-drawn tip76. After equilibrating the column each sample was loaded through a Thermo EASY-LC (Thermo Fisher Scientific). A gradient was formed and peptides were eluted with increasing concentrations of solvent B (90% acetonitrile, 0.1% formic acid).
As peptides eluted, they were subjected to electrospray ionization and then entered into an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific). Peptides were detected, isolated and fragmented to produce a tandem mass spectrum of specific fragment ions for each peptide. Peptide sequences (and therefore protein identity) were determined by matching protein databases with the acquired fragmentation pattern by the software program, Sequest (Thermo Fisher Scientific)77. All databases include a reversed version of all sequences, and the data were filtered to a 1–2% peptide FDR.
Recombinant protein production and purification
Mouse MyD88TIR (residues 129–296), mouse TIRAPTIR (residues 74–249), and mouse TRAMTIR (63–232) were subcloned into a pET28a expression vector using Gibson Assembly. The TIRAPTIR(P125H) mutant was generated using site-directed mutagenesis with Q5 DNA polymerase. All constructs were designed with an N-terminal His6 tag, MBP, and a TEV recognition sequence, and verified by whole-plasmid sequencing. Clones were transformed into E. coli BL21 (DE3) cells, grown at 37 °C in LB broth (Miller) to an optical density at 600 nm (OD600) of 0.4–0.6, and induced with 1 mM IPTG (I56000, Research Product International) for 4 h at 30 °C. Bacterial cells were collected by centrifugation at 5,000g for 5 min, flash-frozen in liquid nitrogen and pellets stored at −80 °C until purification. Bacterial cells were lysed by sonication in buffer containing 50 mM HEPES pH 8.0, 150 mM NaCl, 10 mM β-mercaptoethanol and 1 mM PMSF. The soluble supernatant fraction was extracted by ultracentrifugation at 50,000g for 1 h, filtered through a glass fibre filter and loaded over a Ni-NTA affinity column (cOmplete His-Tag Purification Resin; 5893682001, Roche) pre-equilibrated with lysis buffer. The column was washed in 10 column volumes (CV) of wash buffer A (50 mM HEPES pH 8.0, 300 mM NaCl, 10 mM β-mercaptoethanol), followed by 20 CV of wash buffer B (50 mM HEPES pH 8.0, 500 mM NaCl, 10 mM β-mercaptoethanol, 5 mM imidazole), and eluted in 5 CV of elution buffer (50 mM HEPES pH 8.0, 150 mM NaCl, 10 mM β-mercaptoethanol, 150 mM imidazole). The resulting eluate was supplemented with 1% (w/w) TEV protease and dialysed overnight at 4 °C in 10 kDa cut-off SnakeSkin Dialysis Tubing (68100, Thermo Fisher Scientific) in 1 l buffer (50 mM HEPES pH 8.0, 150 mM NaCl, 10 mM β-mercaptoethanol). The nickel resin column was re-equilibrated in 20 CV dialysis buffer and passed with the dialysate three times. The TEV-cleaved TIR domain containing flow-through was further concentrated using Amicon Ultra Centrifugal Filters (3/10 kDa MWCO; UFC900324, Millipore Sigma) and loop-loaded over a single Superdex 200 Increase 10/300 GL (Cytiva) column in 500 µl volumes. The fractions corresponding to the protein of interest were verified by SDS–PAGE and Coomassie Instant Blue staining, pooled and concentrated to about 4 mg ml−1, flash-frozen in liquid nitrogen and aliquots stored at −80 °C until further use. Protein was stored in a final buffer composition of 20 mM HEPES pH 7.0, 50 mM NaCl and 5 mM DTT.
Turbidity assays
Protein aliquots were thawed on ice, centrifuged at 15,000g for 5 min to remove precipitate and the protein concentration of the supernatant was measured. To assess TIR domain assembly formation, 0–200 µM concentration ranges of MyD88TIR, TRAMTIR and TIRAPTIR, alone, and in combination, were investigated. The assay was performed in a UV-STAR 96-well half area plate (Greiner Bio-One) using a TECAN plate reader, shaken continuously at 180 rpm for 5 h at 37 °C and OD350 measurements recorded at 5 min intervals. Reactions were performed in triplicate conditions in 30 µl volumes in assay buffer (20 mM HEPES pH 7.0, 50 mM NaCl, and 5 mM DTT), and their average used as one repeat. Assay data shown contain data from at least two independent purifications.
BS3 cross-linking and immunoblot analysis
MyD88TIR, TRAMTIR and TIRAPTIR were mixed in 30 µl duplicate volumes in assay buffer, with one maintained at 4 °C and the other incubated at 37 °C for 5 h to mimic turbidity assay start and end point conditions, respectively. Bis(sulfosuccinimidyl)suberate (BS3) cross-linker (21586, Thermo Fisher Scientific) was reconstituted to 100 mM in turbidity assay buffer at room temperature, immediately before use. Each reaction received 1.5 µl reconstituted BS3 (final concentration, 5 mM) and was incubated at room temperature for 30 min for complete cross-linking. This was followed by a 15 min quench reaction at room temperature, by adding Tris HCl pH 7.5 to a final concentration of 50 mM. Cross-linked samples were analysed by SDS–PAGE and immunoblotting as described above. For complete transfer of the larger protein assemblies and TCPs, protein transfer onto PVDF membranes was performed at 45 V and 4 °C overnight.
In vivo myddosome inhibition experiments
For all in vivo studies, 12-week old female C57BL/6N mice purchased from Envigo were used. Mice were allowed at least a 72 h acclimatization period on arrival at Boston Children’s Hospital (BCH) and were maintained under specific-pathogen-free conditions with continuous access to food and water. All animals were maintained under a 12 h–12 h light–dark cycle in a temperature- and humidity-controlled vivarium, with lights on from 07:00 to 19:00, and the mice were examined before the initiation of the studies to ensure that they were healthy and acclimated to the laboratory environment. All of the animal studies were approved by the BCH Institutional Animal Care and Use Committee (IACUC) and conducted under the supervision of the department of Animal Resources at Children’s Hospital (ARCH). Mice were i.p. injected with 1 mg per kg LTA or 1 mg per kg LPS in 200 μl PBS or with PBS only. After 30 min mice were i.p. injected with 150 mg per kg TSI-13-57 formulated in 200 μl 5% N-methyl-2-pyrrolidone (NMP; 328634, Millipore-Sigma), 5% Solutol HS 15 (HY-Y1893, MedChem Express), 90% normal saline or vehicle control. Then, 90 min after the PAMP injection, the mice were euthanized with CO2 and spleens surgically removed. Spleens were kept in wash buffer (PBS + 1 mM EDTA + 2% FCS) on ice and mechanically dissociated by gentle homogenization through a 70 µm nylon mesh strainer using a syringe plunger. While on ice, the splenocyte suspension was washed with 20 ml wash buffer and cells were centrifuged at 500g for 5 min and 4 °C. Red blood cells were lysed by resuspending in 1 ml ACK buffer, incubating for 1 min at room temperature and quenching with 10 vol wash buffer. The cell suspension was filtered through a 40 µm strainer and the cells washed twice with wash buffer using centrifugation at 500g for 5 min at 4 °C. Splenocytes were resuspended in 1 ml PBS + phosphatase inhibitors and counted. Immunoprecipitation of MyD88 and associated proteins was performed as described above using lysates from 5.0 × 107 splenocytes.
To analyse serum cytokines and splenic IFN responses, mice were injected as described above. Then, 90 min after the PAMP injection, mice were bled retro-orbitally. After 4.5 h, a submandibular bleed was performed and mice were immediately euthanized with CO2, after which spleens were surgically removed. Blood was collected in EDTA-coated containers (BD Microtainer K2EDTA, 365974) and centrifuged at 2,000g for 10 min at 4 °C. The obtained serum was sampled for TNF, IL-6 and IL-12p40 by ELISA. Spleens were processed as described above and splenocyte lysates were assessed for an IFN response by immunoblotting and RT–qPCR.
Data handling and statistics
Data were plotted using Prism v.10.6.1 (GraphPad) and are presented as mean values of multiple experiments, with error bars showing the s.d. or s.e.m. unless otherwise noted. Statistical significance of results was determined using nonparametric one-way ANOVA, unpaired t-tests or two-way ANOVA as indicated in the figure legends. Benjamini, Krieger and Yekutieli false-discovery-rate-based (Q = 5%) correction for multiple comparisons as implemented in Prism was used when making more than three comparisons. Uncropped immunoblot images and source data are provided in Supplementary Data 1 and the Source Data.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
