Abstract
Interleukin-1β (IL-1β) is dysregulated in chronic inflammatory diseases, yet the genetic factors influencing IL-1β production remain largely unknown. Myeloid-derived cells are the primary producers of IL-1β, which prompted a genome-wide CRISPR knockout screen in the human myeloid-derived U937 cells treated with lipopolysaccharide (LPS) to mimic inflammatory conditions and sorted for high and low intracellular IL-1β levels. A total of 295 genes are identified as regulators of IL-1β production, with 57 overlapping loci associated with inflammatory diseases, including the TNRC18 gene locus associated with multiple diseases in the Finnish population. U937 cells engineered with the Finnish-enriched rs748670681 risk allele demonstrate decreased expression of TNRC18 and an adjacent gene WIPI2, reduction in LPS-dependent gene activation and cytokine production, but elevation of interferon-responsive gene programs. Transcriptomic profiles for individual knockouts of TNRC18 and WIPI2 attribute the loss of LPS-dependent signaling primarily to TNRC18, which occurs through the modulation of H3K27 acetylation around inflammatory regulatory regions via TNRC18 and its protein interaction network. In contrast, the loss of WIPI2 is characterized by an exacerbation of interferon signaling. These findings delineate the global regulatory mechanisms of IL-1β production and provide molecular insights to the role of the rs748670681 variant in inflammatory diseases.
Similar content being viewed by others
Introduction
Precise regulatory control of the innate and adaptive immune responses is essential for proper tissue homeostasis and the prevention of inflammation-driven diseases. Cytokines play a critical role in the maintenance of this homeostasis. Sustained levels of inflammatory cytokines such as TNF-α, IL-6, and IL-1β contribute to the development of chronic diseases including type II diabetes1, atherosclerosis2, cancer3, neurodegeneration4, and autoimmunity5. These cytokines are typically produced by hyperactivated effector immune cells such as macrophages, natural killer cells, B cells, and dendritic cells. Establishing and maintaining immune cells during hematopoietic differentiation requires coordinated activities of pioneering and signal-dependent transcription factors6. For macrophages, PU.1 is an essential factor that binds to specific DNA sequences within regulatory regions, reorganizes the nucleosome structure, and interacts with members of the AP1 and C/EBP family of transcription factors to establish a permissive epigenetic landscape, including high levels of H3K27ac and H3K4me1/2/37,8. In response to pathogen-associated molecular patterns or damage-associated molecular patterns, which are recognized by pattern recognition receptors, signal-dependent transcription factors, including members of the NF-κB, STAT, and IRF families, bind to active regulatory regions established by PU.1. They recruit the RNA polymerase complex and initiate the rapid and robust activation of proinflammatory gene programs9. One example of this regulation is the proinflammatory cytokine IL-1β, essential for maintaining immune homeostasis, whose dysregulation is a hallmark of chronic inflammation. PU.1 constitutively binds to the IL1B promoter, facilitating the LPS-dependent recruitment of the NF-κB family to promote transcriptional activation10. Pro-IL-1β, the precursor for IL-1β, is encoded by the IL1B gene located in a cluster of IL1 genes on chromosome 2. Upon activation of PPRs, the mRNA level of IL-1β rapidly increases in myeloid cells. Pro-IL-1β is synthesized within the cell as a relatively inactive 31 kDa protein. It is then processed into mature IL-1β by caspase-dependent inflammasomes and secreted from the cell through various mechanisms, including the exocytosis of secretory lysosomes, shedding of plasma membranes, and pyroptosis11,12. IL-1β interacts with IL1 receptors on target cells to activate the NF-κB, JNK, and p38 signaling pathways and induce downstream inflammatory genes13.
Over the past two decades, hundreds of genome-wide association studies (GWAS) have established the link between genetic variation and human immune-mediated diseases14,15. Interestingly, shared genetic associations, often due to pleiotropy, have been observed for many of them16,17,18, supporting the idea of hubs and hot spots driving the origin of multiple immune disorders19. Recent studies have also demonstrated that therapeutic targets with genetic linkage to diseases have increased success rates in clinical trials and for reaching FDA approval20,21,22,23. Gene products with roles in multiple diseases are particularly attractive targets as they can be repurposed for numerous indications. For example, TNF-α, Janus kinases, TYK2, IL-12, IL-23, IL-4 and IL-13 have been successfully targeted and gained regulatory approvals for inflammatory diseases such as psoriasis24,25, inflammatory bowel disease26,27,28, atopic dermatitis29,30 and others. However, not all therapeutic agents demonstrate equally desirable efficacies for patients, making it essential to continuously search for and discover effective treatments tailored to specific patient populations through precision medicine. Therefore, prioritizing targets that exhibit substantial biological effects, sufficient cellular expression, and clear genetic links to human diseases is a cost-effective strategy. IL-1 blockade therapies using anakinra have been successful in the clinic for treating patients with psoriatic arthritis, ankylosing spondylitis, and rheumatoid arthritis31. However, they have been unsuccessful in treating systemic lupus erythematosus and Sjögren’s syndrome32. Upstream factors involved in IL-1β regulation might provide better therapeutic targets for additional autoimmune indications.
In this study, we aimed to utilize unbiased genome-wide CRISPR-Cas9 mediated loss-of-function screening to identify regulators of IL-1β production and processing in LPS-treated myeloid cells. We identify known and unreported regulators of intracellular IL-1β expression, with 57 showing genetic association with at least one immune indication. Furthermore, we provide evidence that one of the identified IL-1β regulators, Trinucleotide Repeat Containing 18 (TNRC18), harbors an intronic single-nucleotide polymorphism (SNP) associated with multiple immune-mediated diseases in the Finnish founder population33, which regulates the expression of TNRC18 and an adjacent gene WIPI2. Knockout of WIPI2 results in hyperactivation of interferon-responsive gene programs, while knockout of TNRC18 reduces LPS-dependent gene activation through modulating H3K27ac at inflammatory regulatory regions.
Results
Genome-wide CRISPR knockout screen identifies regulators of intracellular IL-1β expression
To assess the robustness of Toll-like receptor 4 (TLR4; the receptor for LPS)-dependent activation and secretion of IL-1β in the human monocytic cell line U937, we treated phorbol-12-myristate-13-acetate (PMA)-differentiated U937 cells with increasing doses of LPS and measured secreted levels of IL-1β. Secretion of IL-1β correlated with the LPS dose and reached maximum concentration at around 500 pg/ml. However, the addition of Nigericin, a potent activator of the NLRP3 inflammasome, which promotes the processing and release of IL-1β, robustly enhanced IL-1β secretion to 12,000 pg/ml starting at a dose of 0.0001 μg/ml LPS (Fig. 1a). In addition, we observed a 100-fold increase in intracellular levels of IL-1β after 24 h of LPS stimulation (Fig. 1b). Thus, exposure of PMA-treated U937 cells to LPS leads to robust production of IL-1β protein expression, which remains primarily intracellular until it is secreted upon inflammasome activation.
a Treatment of PMA-differentiated U937-LC3-Cas9+ cells with increasing doses of LPS induces expression and secretion of IL-1β. Addition of Nigericin augments the secretion of IL-1β. Values are represented as an average of three independent replicates -/+ standard deviation (SD). b Intracellular IL-1β level upon induction with LPS compared to vehicle treatment. Values are represented as an average of three independent replicates -/+ SD. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). c Overview of the genome-wide CRISPR screening strategy using the Brunello sgRNA library. The diagram was created with BioRender.com. Genes that decrease (d) or increase (e) IL-1β levels when knocked out in U937-LC3-Cas9+ cells. For each gene the x axis shows the enrichment of sgRNAs in log2 enrichment values, and the y axis shows statistical significance of enrichment as -log10P value. Statistically significant genes were identified by aggregating differential sgRNAs ranked consistently higher at gene level using the robust rank aggregation (RRA) method. Pathways and classes of known and unknown genes that modulate IL-1β levels identified in IL-1βlow (f) and IL-1βhigh (g) populations. LPS lipopolysaccharide, sgRNA single guide RNA, NGS next-generation sequencing.
To conduct genetic screens, U937 cells expressing constitutive Cas9 and a LC3 autophagy reporter (U937-LC3-Cas9+)34 were transduced with the pooled Brunello CRISPR library containing 76,441 single guide RNAs (sgRNAs) systematically targeting every protein-coding gene (n = 19,114) in the genome, each with four independent sgRNAs, and 1000 non-targeting controls35. Transduced cells were differentiated with PMA and treated with LPS for 24 h. Cell populations with the lowest 10% (IL-1βlow) or the highest 10% (IL-1βhigh) levels of intracellular IL-1β were collected using fluorescence-activated cell sorting (FACS), and sgRNA counts in each population were compared to those in the pre-sorted control population using next-generation sequencing (Fig. 1c). We performed this experiment in triplicate and considered genes to be significantly enriched if they had a Q-value < 0.35 in either the IL-1βlow or IL-1βhigh populations when compared to pre-sort samples in 2 or 3 replicates. The correlations (R2) between our three IL-1βlow populations were between 0.31 and 0.50, while the correlations between the three IL-1βhigh populations were between 0.39 and 0.45. This analysis identified 128 genes (Q-value < 0.35) (Fig. 1d, Supplementary Data 1), whose CRISPR-mediated knockout led to loss of IL-1β production, and 167 genes (Q-value < 0.35) that led to increased intracellular levels of IL-1β (Fig. 1e, Supplementary Data 2).
As expected, members of the TLR4 pathway including IRAK1, IRAK4, MYD88 and TRAF6 were enriched in the IL-1βlow cell population (Fig. 1f). In addition, members of the JAK-STAT pathway (e.g., CSF2RA/B, JAK1, JAK2), the EGF pathway (e.g., KRAS, MAPK1, MAP2K1, SHOC2), and many transcription factors (e.g., AHR, GATA1, JUNB) were essential for LPS-dependent activation of IL-1β. Conversely, knockout of members of the Cul5 ubiquitin ligase (e.g., ARIH2, CUL5, RNF7, SENP8, UBE2F3) and the STAGA (e.g., CCDC101, KAT2A, SUPT7L, SUPT20H) complexes and genes involved in immune cell activation (e.g., BCL10, IRF3, TBK1, TNFAIP3) lead to elevated intracellular IL-1β (Fig. 1g). Collectively, these unbiased genome-wide screening results confirm known pathways of IL-1β production and identified several dozen regulators of IL-1β.
Arrayed screen validates regulators of IL-1β production and secretion
Our pooled CRISPR screen was designed to capture genes that modulate intracellular IL-1β protein levels measured through flow cytometry. However, the identified genes may modulate IL-1β levels at various stages of regulation including production, processing, or secretion. To confirm our genome-wide pooled screening results and provide additional mechanistic support, we selected 18 genes that produced the lowest and 18 genes that produced the highest levels of intracellular IL-1β for a multiplexed arrayed CRISPR screen to evaluate the effect of gene knockout on two phenotypic readouts: (1) intracellular IL-1β production through flow cytometry and (2) production and secretion of IL-1β using Meso Scale Discovery (MSD) immunoassays in supernatants. We confirmed all 18 screening hits identified as positive regulators of intracellular IL-1β levels, including expected regulators such as MAPK1, TRAF6, and MYD88, in addition to previously unknown regulators such as RIT1, CSK and CCDC134 (Fig. 2a). CRISPR knockout of these genes also reduced the secretion of IL-1β in the presence of LPS or LPS and Nigericin, supporting a role for these genes in IL-1β mRNA expression or protein production (Fig. 2c). Conversely, knockout of 12 of 18 genes that yielded the highest levels of IL-1β in the pooled CRISPR screen, showed enhanced intracellular IL-1β levels in the arrayed screen. However, the majority (13 of 18) of these screening hits reduced IL-1β secretion in supernatants compared to control samples. For example, knockout of several members of the CRL5 complex (CUL5, RNF7, SENP8) led to robustly increased intracellular IL-1β (Fig. 2b) but reduced the secretion of IL-1β even in the presence of Nigericin (Fig. 2d). Therefore, knockout of these genes leads to elevated IL-1β levels that become trapped inside the cell, even upon activation of the inflammasome.
Intracellular IL-1β production measured by flow cytometry following LPS treatment of U937-LC3-Cas9+ cells that were transduced with sgRNAs targeting genes identified in genome-wide CRISPR screens in either IL-1βlow (a) or IL-1βhigh (b) populations. Values are represented as an average of three independent replicates -/+ SD. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05) compared to NTC. Normalized levels of secreted IL-1β collected from supernatants of LPS or LPS and Nigericin treated U937-LC3-Cas9+ cells, following Cas9-RNP nucleofections targeting CRISPR screening hits in IL-1βlow (c) or IL-1βhigh (d) populations. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05) comparing NTC to CRISPR targeted cells. MFI median fluorescence identity, RNP ribonucleoprotein, NTC non-template control, LPS lipopolysaccharide.
Genetic association of CRISPR screening hits with immune-mediated diseases
Despite extensive efforts to fine-map disease-associated loci discovered by GWAS, identifying the underlying target genes remains challenging. Genes involved in the regulation of the expression and/or secretion of an essential cytokine such as IL-1β, may offer supporting evidence for causal genes found within GWAS loci associated with immune-mediated diseases. The Open Targets locus-to-gene (L2G) model ranks genes at each GWAS locus by integrating genetic, epigenetic, and functional genomics data36. To assess the genetic links between immune-mediated diseases and the genes identified in our CRISPR screen, we searched the Open Targets Genetics database for evidence supporting causal relationships between these genes and the diseases at GWAS loci37. We selected 93 published GWAS and studies conducted on immune-mediated diseases in FinnGen and UK Biobank in which loci containing top screening hits from our study showed genome-wide significant associations. Harmonization of identical (e.g., atopic dermatitis and eczema combined as atopic dermatitis) or related phenotype (e.g., ACPA-positive rheumatoid arthritis [RA] and unspecified [RA] combined as [RA]) definitions across the studies resulted in 28 unique or jointly evaluated phenotypes (Supplementary Data 3). Among the 295 screening hits, 57 genes (19%) were found to reside in loci associated with at least one immune disease and have been linked to GWAS signals with an average L2G score of 0.1 and above (Fig. 3a, b, Supplementary Data 3). Almost half (28/57) of these genes were linked to GWAS signals with L2G scores above the hard threshold36 of 0.5. Some loci, including those harboring genes such as SH2B3, IL12RB2, and TNFAIP3, show association with multiple diseases. TNFAIP3 encodes the AP20 enzyme, a known regulator of NF-κB-mediated immune activation38 and pro-IL-1β processing39,40. The association signals at the TNFAIP3 gene locus have previously been resolved to TNFAIP3 using both experimental41,42,43 and in silico44 methods for some of the immune-mediated diseases. Therefore, the CRISPR-mediated validation of the role of AP20 in IL-1β production supports the validity of this approach for inferring potential mechanistic links between GWAS signals and target genes.
a, b Top locus-to-gene (L2G) scores from Open Targets Genetics for 57 gene-disease pairs. 57 genes identified in the CRISPR screen were linked to various immune-mediated diseases in GWAS with varying levels of confidence as determined by the L2G scores. L2G scores above 0.1 are shown. IL-1βhigh genes are highlighted in red (a) and IL-1βlow genes are in blue (b). c Forest plot showing direction of effect of the rs748670681 variant in the TNRC18 gene locus, associated with immune-mediated diseases in the Finnish population. The minor allele T concomitantly increases risk for 7 diseases (OR > 1) and decreases risk (OR < 1) for 5 diseases at genome-wide significance (P value < 5 × 10-8). L2G scores for the TNRC18 gene in the tile plot in panel b are derived from FinnGen GWAS in data freeze (DF) 6 (n = 260,405), whereas the odds ratios (ORs) in the forest plot in panel c are based on data from DF9 GWAS (n = 377,277), where additional genome-wide significant associations with rs748670681 were detected with increased sample sizes (e.g., multiple sclerosis) that were not detected in DF6. In the forest plot, dots indicate the odds ratios of the variant and whiskers represent 95% confidence intervals. CID chronic inflammatory diseases, AS ankylosing spondylitis, CD Crohn’s disease, PSO psoriasis, PSC primary sclerosing cholangitis, SLE systemic lupus erythematosus, SS systemic sclerosis, UC ulcerative colitis.
CRISPR screening hits can facilitate prioritization of candidate genes in unmapped GWAS loci. For example, TENT4A can potentially be prioritized as a candidate gene for multiple sclerosis (MS), given the role of IL-1β in MS45. TENT4A stabilizes and increases the half-life of mRNA molecules by guanylation of their poly(A) tails46. Our screen shows that loss of TENT4A leads to reduced IL-1β, suggesting that it directly or indirectly contributes to IL-1β production. A common SNP (rs34681760) in the promoter of TENT4A is significantly associated with MS risk47 (P = 2.0 × 10-11, OR = 1.1), which has been linked to TENT4A with a high L2G score (0.73). Furthermore, rs34681760 has been identified as a strong expression quantitative trait locus (eQTL) for TENT4A in multiple studies, where the MS-associated risk allele C has been linked to higher expression of TENT4A in blood48, monocytes49,50,51 (β = 0.25, P = 6 × 10-30 in Quach et al.) and macrophage cells52, but with lower expression in dendritic cells53 (β = –0.34, P = 2 × 10-15), compared to the non-risk allele T. However, no experimental evidence currently exists to explain the biological mechanisms underlying the MS association. The role of TENT4A in mRNA modification and IL-1β regulation suggests that it may influence pathways leading to inflammation in MS. Therefore, these data shed light on potential mechanistic links of such genetic associations and can facilitate hypothesis generation for further investigation.
Genetic pleiotropy of the TNRC18 gene locus in the Finnish founder population
One of the statistically significant (Q-value < 0.35) genes identified in our CRISPR screen codes for a poorly characterized, intracellular protein TNRC18. The TNRC18 gene locus has been associated with several autoimmune and inflammatory diseases in the Finnish and Estonian populations33. The association signal is particularly pronounced for inflammatory bowel disease (IBD) (OR = 2.23, P = 1.47 × 10-117), followed by ankylosing spondylitis (AS) (OR = 2.33, P = 7.6 × 10-53). Furthermore, the lead SNP rs748670681 is 114-fold enriched in the Finnish population (allele frequency [AF] = 3.6%) and is rare outside Finland and Estonia (Supplementary Data 4). The association signal spans a 3-Mb genomic region that contains more than 30 protein-coding genes (Fig. 4). Because of the strong genetic association with multiple diseases and the fact that this gene is poorly characterized, we decided to investigate this finding further.
Negative log10 P values are shown for variants within a 2 Mb region centered at the reference SNP (rs748670681). The reference SNP is marked with a purple diamond, and pairwise LD (r2) between the reference SNP and other variants are indicated by color. The r2 values were estimated from high-coverage whole-genome sequences of 8557 Finns. Both directly genotyped and imputed SNPs are plotted. Genomic coordinates are shown according to the human genome build GRCh38/hg38. Two-sided P values are obtained from a likelihood ratio test in regression analysis and are not corrected for multiple comparisons. Mb megabase.
To identify the likely causal variant(s) driving the association signals, we first performed conditional analyzes, conditioning on the lead SNP rs748670681 in IBD and type I diabetes, where the minor allele shows opposite effects33. After the conditional analyzes rs748670681 remained the sole, disease-associated SNP with genome-wide significance, suggesting that this SNP is responsible for the association signals in both conditions. To identify other diseases and related phenotypes that might show association with this SNP, we queried GWAS summary statistics for 2266 endpoints conducted in 377,277 individuals in the FinnGen study (Supplementary Data 5). The rs748670681 variant showed genome-wide significant (P < 5 × 10-8) association with ten other immune-mediated diseases. Strikingly, the low-frequency, T allele of rs748670681 showed risk effects for some inflammatory disorders, whereas it was protective for autoimmune conditions such as Grave’s disease, hypothyroidism, and MS (Fig. 3b).
The risk allele of rs748670681 alters the expression levels of TNRC18 and WIPI2
The rs748670681 variant is located within the second intron of TNRC18 and overlaps with epigenetic marks for active transcription (H3K27Ac), enhancer activity (H3K4me1), and promoter activity (H3K4me3) in primary human macrophages (Fig. 5a). To test the regulatory activity of this sequence we engineered DNA fragments containing either the risk-associated (i.e., T) or non-risk associated (i.e., C) allele of the rs748670681 variant into a luciferase-based reporter vector. We independently transfected these constructs into Jurkat cells, an immortalized T lymphocyte line, or Caco-2 cells, an immortalized cell line of human colorectal adenocarcinoma cells. The vector harboring the non-risk allele C significantly increased luciferase expression levels compared to the empty vector. However, the vector carrying the risk allele T showed significantly reduced activity compared to the non-risk allele C in both cell types (Fig. 5b, c). This confirms the functional impact of the rs748670681 variant and demonstrates that the risk allele T has decreased transcriptional activity.
a UCSC genome browser view of the TNRC18 gene locus. Tracks from top to bottom show RefSeq protein coding genes, H3K4Me1, H3K4Me3, and H3K27Ac marks in human macrophage cells from the BLUEPRINT Project. The rs748670681 variant, located inside intron 2 of TNRC18, is marked with a red triangle. The minor, risk-associated T allele of rs748670681 diminishes luciferase reporter gene expression in Jurkat (b) and Caco-2 (c) cells compared to the common, non-risk allele C. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). d Target sequence for HDR-mediated CRISPR-Cas9 editing of the rs748670681 variant sequence. e Editing was performed in three independent experiments and quantified using next-generation sequencing with at least 50 K sequencing reads. Normalized expression values of the WIPI2 (f) and TNRC18 (g) transcripts in cells edited with CRISPR-Cas9 to be homozygous for the T allele of rs748670681. Each bar in graphs represents mean of replicates ± SD (n = 9). Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). H3K4Me1 histone H3 lysine 4 monomethylation, H3K4Me3 histone H3 lysine 4 trimethylation, H3K27Ac histone H3 lysine 27 acetylation, CPM counts per million, Luc luciferase, PAM protospacer adjacent motif, LPS lipopolysaccharide.
To gain a better understanding of the functional effect of the rs748670681 risk allele, we utilized CRISPR-Cas9 mediated homology-directed repair (HDR) and engineered U937 cells to harbor homozygous risk (TT) and non-risk (CC) genotypes. We specifically designed conversion of the C nucleotide to T in our sgRNA (GAGGGCTTGGCGCTGGAGTC) to disrupt the PAM (NGG) sequence and prevent successive Cas9 cutting of our correctly edited genomic sequence (Fig. 5d). After allowing the cells to recover for 72 h, we performed next-generation sequencing analysis of the bulk edited cell populations and identified ~60% genome editing efficiencies across three independent samples (Fig. 5e). We successfully isolated several homozygous clones for the non-risk (CC) and the risk (TT) genotypes, which were subjected to global transcriptomic analysis under PMA-differentiated -/+ LPS treatments.
The mRNA expression levels for both WIPI2 (vehicle: log2 fold-change LFC = –0.55, P < 1.1 × 10-15, LPS: LFC = –0.5, P < 1.4 × 10-11) and TNRC18 (vehicle: LFC = –0.76, P < 0.001, LPS: LCF = –0.87, P < 0.001) were reduced in cells engineered to carry the homozygous rs748670681 TT risk genotype compared to the non-risk CC genotype (Fig. 5f, g). The expression of two nearby genes, SLC29A4 and FBXL18, was below reliable detection by RNA-seq in U937 cells. Furthermore, the mRNA-Seq dataset shows signal surrounding the location of rs748670681 in non-risk (CC) cells, but not in the risk (TT) cell (Supplementary Fig. 1). This loss of expression eliminates the first exon of isoform XM_017012735.1 of TNRC18 in risk (TT) cells and contributes to the loss of TNRC18 mRNA observed in the risk (TT) cells. Therefore, the genetic variant rs748670681 appears to exhibit pleiotropic, regulatory effects on two genes in the TNRC18 locus: TNRC18 and WIPI2.
TNRC18 regulates expression and secretion of proinflammatory cytokines
The identification of TNRC18 as a genome-wide CRISPR screening hit of IL-1β expression and an abundance of genetic evidence linking it to multiple immune-mediated diseases prompted us to explore its role in myeloid cells in more detail. TNRC18 is expressed as 13 unique splice isoforms, with the longest isoform spanning 30 exons (2968 AAs) and the shortest 3 exons (99 AAs) (Supplementary Fig. 1). Although the exact function of TNRC18 in regulation of IL-1β is currently unknown, it has been reported to interact with several proteins involved in chromatin remodeling54,55,56,57,58, suggesting a potential role for TNRC18 in transcriptional regulation. Within our CRISPR screen, three TNRC18 guides were enriched in the IL-1βlow populations compared to either the presort samples or IL-1βhigh populations (Supplementary Fig. 2a). We generated TNRC18 knockout U937 cells using CRISPR-Cas9 with Inference of CRISPR Edits score of 80% (Supplementary Fig. 2b) and confirmed the LPS-dependent reduction of IL-1β in TNRC18 knockout cells relative to non-edited cells by flow cytometry (Supplementary Fig. 2c, d). TNRC18 was also required for maximal secretion of proinflammatory cytokines IL-1β, IL-6 and TNF-α in LPS-treated U937 cells (Supplementary Fig. 2e–g). Finally, Cas9 ribonucleoproteins (RNPs) targeting TNRC18 in primary macrophages from human donors resulted in reduced levels of secreted IL-1β, IL-6, and TNF-α in response to a range of LPS doses (Supplementary Fig. 3). Collectively, these results demonstrate the requirement for TNRC18 in LPS-dependent activation and secretion of proinflammatory cytokines in immortalized and primary human myeloid cells.
TNRC18 and WIPI2 regulate distinct sets of genes that overlap with genes altered by rs74867068
To assess whether U937 cells carrying the homozygous TT risk genotype of rs74867068 demonstrated diminished regulation of proinflammatory gene expression we generated global transcriptomic profiles using RNA-seq. We compared the expression of individual genes between cells homozygous for the CC non-risk genotype versus those engineered to carry the homozygous TT risk genotype with and without LPS stimulation. Compared to cells homozygous for the CC genotype, TT cells treated with LPS showed overexpression of 658 genes with LFC > 1 and adjusted P value < 0.05. Conversely, 433 genes were downregulated at the same threshold (Fig. 6a, Supplementary Data 6). Genes elevated in cells homozygous for the TT risk genotype are enriched for interferon signaling pathways and defense to viral responses, including IFIT3, OAS3, IFI6 (Fig. 6b), while genes exhibiting decreased expression include IL1B, CCL4, CXCL8 and genes participating in pathways responsible for regulating immune cell responses, cell migration, and proliferation (Fig. 6c). Selected genes among the up- and down-regulated genes are shown in Fig. 6d, e, highlighting increases in interferon-responsive genes (IFIT3, OAS3, and IFI6) and decreases in inflammatory cytokines (IL1B, CCL20, and CCL4).
a Differentially expressed genes in U937 cells edited with CRISPR-Cas9 to be homozygous for the T allele of rs748670681 and treated with LPS compared with LPS-treated rs748670681-CC cells. Differential expression was compared using the test in limma R package (v3.54.2). P values were adjusted using the Benjamini–Hochberg method. Selected pathways for enhanced (b) or decreased (c) genes ranked by -log(q-value) in pathway enrichment analysis using gene ontology terms for biological processes. Risk allele T of rs748670681 increases expression of the IFIT3, OAS3, and IFIH6 genes (d), but decreases expression of IL1B, CCL4 and CCL20 genes (e). Each bar in graphs represents mean of replicates ± SD (n = 9). Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). Effect of rs748670681 on gene expression was assessed with and without LPS stimulation. Number of overlapping genes up-regulated (f) and down-regulated (g) in WIPI2 KO, TNRC18 KO cells, and cells homozygous for the T allele of rs748670681 (LFC > 1, adjusted P-value < 0.05). LFC log2(fold-change), CPM count per million, IFN interferon. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05).
The observation that the rs74867068 variant modulates the expression levels of both the TNRC18 and WIPI2 mRNAs prompted us to examine whether we could assign the gene regulation patterns defined in U937 cells carrying the TT genotype of rs74867068 to knockout of either TNRC18 or WIPI2. Knockout of WIPI2 leads to the increased expression of 191 genes with the overlap of 29 genes with rs74867068-TT U937 cells TT (P = 5.77 × 10-81) (Fig. 6f, Supplementary Data 7). In contrast, knockout of TNRC18 leads to decreased expression of 203 genes with an overlap of 55 downregulated genes with rs74867068-TT U937 cells (Fig. 6g, Supplementary Data 8) (P = 0.0003). These results demonstrate that knockout of WIPI2 leads to enhanced expression of interferon-responsive genes, while TNRC18 controls the expression of proinflammatory cytokines.
TNRC18 complexes modulate H3K27 acetylation at loci regulating inflammation
To uncover potential mechanisms through which TNRC18 regulates inflammatory gene expression, we analyzed protein-protein interaction databases57 and leveraged recently published TNRC18 BioID experiments58 to identify connections between TNRC18-associated protein complexes and functional transcriptional outcomes. These analyzes identified a core set of 205 proteins that interact with TNRC18 and enriched for cellular process involving epigenetic regulation of gene expression (i.e., HDAC3, ZCCHC8), the CCT complex (i.e., CCT2, CCT4, CCT8), chromosomal organization (i.e., CCAR2, TRIM28), and regulation of cytokine signaling (i.e., ABL2, KRAS) (Fig. 7a, b). Several TNRC18 interactors were also significant hits in our genome-wide CRISPR screen for increased (i.e., CEBPA, FAM60A, YLPM1) or decreased (i.e., CCAR2, HDAC3, RBBP7) IL-1β production (Fig. 7c). Interestingly, HDAC3 and RBBP7 are components of the Sin3-NCoR-HDAC complex which has recently been shown to positively regulate inflammatory gene expression59,60. Loss of NCoR59 or HDAC361 results in changes in H3K27ac at specific promoters and enhancers of inflammatory genes. This prompted us to evaluate the global role of TNRC18 in H3K27 acetylation levels in U937 cells. Cleavage Under Targets and Tagmentation (CUT & Tag) assays62 for H3K27ac were performed in WT or TNRC18 knockout U937 cells treated with LPS for 0 or 16 h. We identified a total of 41,674 H3K27ac peaks. Of these, 2119 peaks exhibited enhanced signals in the TNRC18 knockout cells, while 6313 peaks showed decreased H3K27ac (Fig. 7d). Motifs for the transcription factors AP1 and PU.1 were enriched in both the enhanced and decreased peaks (Fig. 7e, f). However, motifs for ZNF669, NF-κB, and CEBP were specifically enriched in the enhanced peaks, while motifs for RUNX2, MITF, and CTCF were enriched in the decreased H3K27ac peaks. Furthermore, knockout of TNRC18 results in altered regulation of inflammatory gene regulation (Fig. 6f, g) and hierarchical clustering results in distinct regulation patterns (Fig. 7g). Loss of TNRC18 decreased the basal expression genes within clusters B, C and E, while increasing the expression of genes in clusters A and D. Notably, knockout of TNRC18 blocked the LPS-induction of genes within cluster B, but not genes within clusters A and E. CCL24 and CCL4 are two cluster B genes that are significantly reduced by TNRC18 KO in LPS treated U937 cells (Fig. 7h, i). Consistent with reduced mRNA production, both CCL24 and CCL4 have lower LPS-dependent deposition of H3K27ac across their gene bodies in TNRC18 KO cells (Fig. 7k, l). To determine whether loss of TNRC18 results in decreased H3K27ac near cluster B peaks compared to cluster E peaks, we assigned LPS-dependent H3K27ac peaks to the nearest gene and assessed the signal across WT and TNRC18 knockout cells. Peaks assigned to Cluster E genes were induced to similar magnitudes by LPS in WT and TNRC18 KO cells, while peaks assigned to Cluster B genes showed significant reduction in LPS-dependent H3K27ac in TNRC18 KO cells compared to WT cells (Fig. 7j). Collectively these results identify TNRC18 associated complexes as drivers of IL-1β production and define a role for TNRC18 in modulating the levels of H3K27ac at genomic regions controlling inflammatory gene expression.
a Overlap of proteins known to interact with TNRC18 in the BioGRID database and in published BioID experiments. b Complex enrichment analysis for the proteins identified to interact with TNRC18. Gene enrichment and pathway analyzes were performed using Metascape. Significance was determined using a hypergeometric test and adjusted for multiple testing using the Benjamini–Hochberg procedure. c Volcano plot of the performance for TNRC18 interactors within the genome-wide CRISPR screen for IL-1β production with the low population normalized to the high population. Statistically significant genes were identified by aggregating differential sgRNAs ranked consistently higher at gene level using the robust rank aggregation (RRA) method. d Heatmap of H3K27ac CUT&Tag in WT and TNRC18 KO cells treated with LPS for 0 or 16 h. e, f Motif enrichment for H3K27ac peaks where TNRC18 KO alters signal. Motif enrichment and significance is calculated by HOMER using a modified version of the cumulative hypergeometric distribution. g Heatmap of hierarchical clustering of mRNA-Seq from WT and TNRC18 KO U937 cells treated with LPS for 0 and 24 h. mRNA-Seq expression for CCL24 (h) or CCL4 (i) in WT and TNRC18 KO U937 cells treated with LPS for 0 or 24 h. Each bar in graphs represents mean of replicates ±SD (n = 3). Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). j Boxplot of H3K27ac CUT&Tag signals at H3K27ac peaks assigned to LPS responsive genes within clusters E (left) and B (right) identified in f. The box represents the middle 50% of the data with the lower and upper quartiles defining the box boundaries. The median is represented by the line inside the box. The whiskers represent the minima and maxima values. Statistical significance was determined using two-sided, unpaired Student’s t test (*P < 0.05). UCSC genome browser imagines showing H3K27ac27 CUT&Tag signal within the CCL24 (k) or CCL4 (l) genomic loci in WT and TNRC18 KO WT and TNRC18 KO U937 cells treated with LPS for 0 or 16 h. m Model for TNRC18 regulation of inflammatory gene expression through deposition of H3K27ac at gene promoters and enhancers. Proteins highlighted in color are genome-wide CRISPR screening hits. The diagram was created with BioRender.com.
Discussion
Therapeutic targets with causal relationships to human diseases usually demonstrate increased success rate in clinical development. Genetics provides some of the strongest evidence for causality in non-communicable human diseases. Since 2002, thousands of GWAS have uncovered genetic variation influencing risk for human complex diseases, including many with alterations in the immune system63,64. Most GWAS disease-associated SNPs localize to noncoding regions, suggesting that they influence disease risk through dysregulation of gene expression. Following GWAS, various computational and experimental approaches are employed to link statistically significant signals in non-coding regions to causal variants and genes, such as through colocalization with eQTLs, or analyzes of chromatin accessibility and interactions. However, even after successfully identifying a GWAS risk gene, the molecular and cellular mechanisms by which these causal genes contribute to disease risk often remain unclear, particularly for genes without known biological functions. Linkage disequilibrium can further complicate the identification of true causal alleles, since it is not uncommon to observe eQTL SNPs affecting more than one gene. Unbiased CRISPR loss-of-function screens in disease-relevant cell types and physiological states offer a complementary, orthogonal approach for identifying disease-associated genes under GWAS peaks, while also providing mechanistic insights into pathogenic processes. The IL-1β cytokine and macrophages are among the most critical drivers of inflammation underlying immune diseases. Here, we describe a genome-wide CRISPR knockout screen followed by arrayed validation to identify regulators of intracellular IL-1β in the human monocytic cell line U937 under stimulation conditions. Our screen validated many known modulators of IL-1β production including the IL1B gene itself. The screen also identified previously unknown modulators, providing a list of potential target candidates for further investigation. Most importantly, 57 (19%) of 295 screening hits reside in genomic loci that have been associated with various inflammatory or autoimmune diseases.
We further characterized one of the most significant GWAS risk loci in the Finnish population. The 7p22.1 chromosomal locus was previously found to modify the risk for several autoimmune diseases in the Finnish and Estonian populations33. The derived allele of the noncoding SNP rs748670681 concomitantly increases the risk for certain diseases and decreases the risk for others. Statistical fine mapping of the risk haplotype suggested that the intronic lead SNP rs748670681 is causal and the TNRC18 gene, which harbors this SNP, was identified as a significant hit in our CRISPR knockout screen, suggesting that TNRC18 could be the effector gene of this GWAS signal. To elucidate the functional role of this variant and to identify the causal genes at the locus, we generated a CRISPR-Cas9 edited cell model carrying the homozygous risk genotype and showed the effect of the risk genotype on the expression of TNRC18 and, also surprisingly on the adjacent WIPI2 gene. WIPI2 codes for a component of the autophagy pathway65. Genetic variation in autophagy genes, as well as perturbation of the autophagy pathway in general, has been implicated in susceptibility to immune disease66,67. The risk allele of rs748670681 diminishes cytokine expression while concurrently enhancing the expression of interferon signaling pathway genes. The latter finding is noteworthy, since the risk allele demonstrates protective effects for multiple sclerosis, a condition traditionally treated with interferon injections68. In validation studies, reduced expression of cytokine genes and induction of the interferon signaling genes appear to result from loss of TNRC18 and WIPI2, respectively. The risk allele decreases the expression of both genes. The opposing effects of the rs748670681 variant on disease risk may be attributed to combined perturbations of both WIPI2 and TNRC18 expression or due to their separate effects. Further studies are necessary to gain a deeper understanding of the pleiotropic effect of the rs748670681 variant, which contributes to increased risk for some immune diseases, simultaneously lowering the risk for others.
The precise function of the TNRC18 protein remains uncharacterized; however, TNRC18 was identified as a top positive hit that confers resistance against mycobacterium infection in CRISPR knockout and knockdown screens in human monocytic THP-1 cells (Lai et al.)69. Interestingly, some of the same pathways (e.g., type I interferon signaling) and top hits (JAK1, SHOC2, JUNB, AHR, ARNT) were common between our and the Lai et al. study. Our data and some recent studies have provided evidence that TNRC18-associated complexes modulate the H3K27ac levels at regulatory regions of inflammatory genes. In high-throughput proteomics and BioID assays, TNRC18 was found to interact with several proteins known to regulate chromatin structure (HDAC1, HDAC3, FAM60A, SIN3A, SIN3B, SAP30, PRDM5, HIST1H1E)54,55,56,57,58,70,71,72,73,74 and transcription factors (AR, ESR2, SOX2)75,76,77. Moreover, TNRC18 contains a bromo-adjacent homology (BAH) domain found in a variety of proteins playing roles in chromatin remodeling and transcriptional silencing78, as well as a Tudor domain that recognizes and binds to methylated lysine and arginine residues allowing them to function as readers of the epigenetic landscape79. In fact, this BAH domain was recently identified as an epigenetic reader for H3K9me3 which mediates the silencing of retrotransposons58. Interestingly, several proteins described to interact with TNRC18 were also found to be significant hits in our genome-wide CRISPR screen for IL-1β production (Fig. 7c), including repressors of IL-1β levels (ex. CEBPA, TKT, and FAM60A) and activators of IL-1β levels (ex. CCAR2, GATAD1, HDAC3). This supports the hypothesis that particular TNRC18 interaction complexes may regulate distinct gene sets (Fig. 7m). For instance, knockout of TNRC18 enhances the H3K27ac signal around 2100 locations with enrichment of AP1, PU.1, NF-κB, and CEBP motifs (Fig. 7e, f). TNRC18 has been shown to interact with CEBPα80, a gene that when knocked down, leads to increased IL-1β production (Fig. 7c), suggesting that TNRC18 interacts with CEBPα to decrease H3K27ac and repress transcriptional activity (Fig. 7m). In contrast, we identified around 6300 regions where TNRC18 appears to have a role in deposition and/or maintenance of H3K27ac which are enriched for AP1, PU.1, RUNX2, MITF, and CTCF motifs. These are regulatory regions likely to contain TNRC18 in complex with the underlying transcription factors for these motifs in addition to proteins that regulate IL-1β production such as CCAR2, HDAC3, GATAD1. Many of these proteins that interact with TNRC18 have not been described to regulate inflammation and warrant further investigation into their regulation mechanisms. Collectively, these studies identify potential mechanisms through which TNRC18 can regulate distinct sets of inflammatory genes through alternative protein complex formation and subsequent modulation of surrounding H3K27ac levels.
In conclusion, our study defines global regulation strategies for IL-1β production in macrophage cells and identifies TNRC18 as a previously unrecognized modifier of LPS-dependent transcription. Our findings suggest that the rs748670681 variant on chromosome 7p22.1 may influence various immune diseases by modulating expression of the TNRC18 and WIPI2 genes.
Methods
Cell culture
U937 cells were maintained in RPMI 1640 media (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco) and 1% Pen/Strep (Gibco). Cells were differentiated by addition of 10 ng/ml phorbol 12-myristate 13-acetate (PMA, Sigma, P1585) for 24 h and treated with 100 μg/ml lipopolysaccharide (LPS) (Sigma, L26375MG) or Nigercin (Invivogen, tlrl-nig-5) for designated times. Primary human CD14+ monocytes (BioIVT) were maintained in RPMI 1640 (Gibco) with sodium pyruvate (ThermoFisher, 11360070), non-essential amino acids, HEPES (ThermoFisher, 15630080), 1% Pen/Strep (ThermoFisher, 15070063), beta-mercaptoethanol (Sigma, M6250), GlutaMAX (ThermoFisher, 35050079), 10% FBS (Gibco), and 100 ng/mL recombinant human M-CSF (Peprotech, 3002510UG) at 37 °C, 5% CO2. Jurkat cells were grown in RPMI 1640 medium with 10% heat-inactivated FBS, and Caco-2 cells were cultured in DMEM media, supplemented with 10% heat-inactivated FBS (Gibco, Life Technologies) in a 37 °C humidified 5% CO2 incubator.
Genome-wide CRISPR screening
120 million U937-LC3-Cas9+ cells, U937 cells that stably express Cas9 and a LC3 autophagy reporter, were infected with the Brunello sgRNA library at multiplicity of infection (MOI) = 0.3 in three biological replicates. U937-LC3-Cas9+ was a pre-engineered cell line that robustly expresses the Cas9 encoding gene with >90% Cas9 activity. The LC3 reporter is neither required for the assays presented here nor does it affect the readout. The Brunello CRISPR-Cas9 knockout library contains 76,441 sgRNAs targeting 19,114 genes with 1000 non-targeting controls35. Following puromycin selection, cells were differentiated with PMA and treated with LPS for 24 h. 40 million cells were collected with trypsin as presort for genomic DNA extraction for each replicate. Greater than 120 million cells per replicate were then processed and stained for intracellular IL-1β at 10 million cells/ml using the protocol mentioned in the Flow Cytometry section. Cell populations expressing the top and bottom 10% IL-1β levels were enriched using a FACS Aria Fusion cell sorter (BD Biosciences). At least four million IL-1βlow and IL-1βhigh cells were collected for genomic DNA extraction using Quick-DNA FFPE Kit (Zymo Research). PCR reactions containing up to 10 μg genomic DNA in each reaction were performed using ExTaq DNA Polymerase with primers to amplify the sgRNAs. Samples were then purified with SPRIselect beads, mixed and sequenced on a NextSeq 500 by 75-bp single-end sequencing.
Screen analysis
Sequencing data were demultiplexed and reads with sgRNA sequences were quantified using custom Perl scripts. Given that the position of the sgRNA could vary per read, the primer sequences flanking the sgRNA were considered. Sequences in between the flanking primers were extracted and then compared to sequences in the sgRNA library. Only sequences with no mismatches were used in the calculation of guide-level read counts. Samples with a minimum of 80% reads mapping to a sgRNAs were considered for further analysis. Through the edgeR Bioconductor package81, reads were normalized between samples using the trimmed mean of M values (TMM) method82. Differential analysis of guide level counts was performed using the limma R Bioconductor package83. Gene-level effects were considered as the median log fold change of the four guides associated with each gene. Statistically significant genes were identified by aggregating differential sgRNAs ranked consistently higher at gene level using the robust rank aggregation (RRA) method84.
Flow cytometry
Cells were harvested and washed twice with FACS buffer (PBS + 1% FBS), incubated with IC Fix buffer (eBioscience, Cat #88-8824-00) in dark for 30 min at room temperature, washed twice with FACS buffer, resuspended in 1X Permeabilization buffer (eBioscience, Cat #88-8824-00) for 10 min and washed twice with 1X perm buffer. Cells were then resuspended in 100 mL/well of FC-block (BioLegend, 422302) (1:100 in Perm buffer) for 15 min at room temperature in the dark. Without washing, directly add 100 μL/well of IL-1β (BioLegend, 508208) at 2X (1:50 in perm buffer) to the FC-block currently in well (1:100 final dilution) and incubate at RT for 15 min in the dark. Cells were then washed twice in 1X Permeabilization Buffer, resuspended in 200 ml of FACS buffer and fluorescence was measured using a BD LSR Fortessa Cell Analyzer. Data was analyzed using the Flow-Jo (v.10) analysis software. The gating strategy for our flow cytometry experiments is shown in Supplementary Fig. 4.
Arrayed screening and cytokine detection
Library of arrayed guide lentivirus was obtained from Broad Institute targeting screening hits with 2–3 guides/gene. Guide RNA sequences are available in Supplementary Data 9. U937-Cas9 clonal cell line (U937-AA7F9) was seeded at 0.05 M cells/100 μl/well in a 96-well plate in presence of 8 μg/ml polybrene and 10 μg/ml blasticidin using Multidrop Combi. The cells were then transduced with guide lentivirus using Bravo automated liquid handler from Agilent in three different replicates and incubated at 37 °C. Two wells of each screening microplate were transduced with positive and negative control guide lentivirus. After overnight incubation transduction media was changed into selection media containing 10 μg/ml blasticidin and 2 μg/ml puromycin and returned to 37 °C incubation. After 3 days of puromycin selection, lentivirus transduced cells were thoroughly mixed and equally divided into 4 different 96 well plates – to be used downstream for IL-1β flow assay and cytokine secretion assay run in duplicate. For IL-1β flow assay, cells were further treated with 100 ng/ml PMA for 24 h, rested for 24 h in PMA-free media followed by 24 h treatment with 100 ng/ml LPS, after which cells were harvested for IL-1β flow assay. For cytokine secretion assays, cells were treated with 25 ng/ml PMA for 3 days, rested in PMA-free media for 4 days and then treated with 1.5 mg/ml LPS with and without 1 μM Nigericin for 24 h. After LPS treatment cell supernatants were collected and analyzed for IL-1β secretion by MSD assay. Puromycin selection was continued throughout PMA and LPS treatments, and all media changes and compound treatments were performed using a Bravo and Multidrop Combi, respectively.
Cytokine secretion assays
Cells were treated as previously described and supernatants were collected from three independent samples. Cytokines were assessed using the Proinflammatory Panel 1 (human) kit (Meso Scale Diagnostics, K15049G) according to the manufacturer’s instructions. Cytokine levels were measured using the MESO QuickPlex SQ120 and concentrations were calculated using a standard calibration curve.
Electroporation and dual luciferase reporter assay
gBlocks with selected SNP-containing fragments (IDT) were cloned into pGL4.10 plasmid (Promega). For the luciferase reporter assay, 150 K Caco2, Jurkat cells were nucleofected with 50 ng of Renilla vector along with 170 ng of luciferase vectors using the Lonza Nucleofector according to the manufacturer’s protocol. Twenty-four hours after transfection, cells were washed once with cold 1× PBS and the luciferase activities were measured with Perkin Elmer Envision using Promega Dual Luciferase Assay kit (Promega, E1960). The firefly luciferase activity was normalized to Renilla luciferase activity for each well. All the luciferase activity measurements were performed in 7 replicates for each condition. The Student’s t-test was applied to estimate the statistical significance of the difference in luciferase activities between the two alleles.
Primary macrophage CRISPR-Cas9 mediated genome editing
CD14+ monocyte cells were thawed and plated in growth media. On day 5, adherent cells were collected over ice using a cell lifter. The cells were washed once with Phosphate-buffered saline (PBS) and resuspended in P3 Primary Cell solution (Lonza). RNP complexes were formed by mixing Cas9 protein (IDT, 1081058) with the indicated sgRNA and incubating at room temperature for 10 min. RNP complexes were then added to the adherent cell solution and electroporated using the Lonza 4D-Nucleofector and protocol CA-137. The cells were then transferred into media and incubated at 37 °C, 5% CO2. After 3 days, cells were treated with LPS for times indicated.
RNA isolation
Total RNA was isolated using the Direct-zol RNA Miniprep Plus kit (R2072, Zymo Research) followed by cDNA synthesis with the SuperScript III First Strand Synthesis kit (11752-050, Thermo Fisher) according to the manufacturer’s instructions.
RNA sequencing
Five hundred nanograms of total RNA were used for NEBNext® Ultra II Directional RNA Library Prep Kit (NEB # E7760L). Poly-A selection (NEBNext Poly(A) Magnetic Isolation Module; NEB # E7490) and cDNA synthesis were performed according to NEB protocol. The NEBNext UMI RNA Adapters were diluted 50-fold in ice-cold UMI Adapter dilution buffer. Ligation reaction was purified using Ampure XP beads. (Beckman Coulter, # A63881) and the adapter ligated cDNA was amplified with 12 cycles of PCR with NEBNext Primer Mix (96 Unique Dual Index UMI Adapters RNA set1, (NEB #E7416). Equimolar amounts of the samples were then pooled and sequenced using NextSeq 500/550 70 base read lengths in paired-end mode.
RNA-Seq data analysis
An average of 23.7 million paired-end reads were generated. FastQ files were aligned to human genome reference using STAR85 (v2.7.10b) with GRCh38 primary assembly (v2.7.4a) gene annotations. Gene expression levels were counted using featureCounts86 (v2.4.3). Differential expression was compared using the test in limma R package83 (v3.54.2). P values were adjusted using the Benjamini and Hochberg method. The significance of the overlap of differentially expressed genes was calculated using the hypergeometric distribution with the phyper function in the R statistical package (v4.2.2). Gene enrichment and pathway analyzes were performed using Metascape87 (http://metascape.org).
Design and synthesis of gRNA and HDR repair template
To generate U937 cells carrying the TNRC18 risk allele, we designed CRISPR/Cas9 gRNA (GAGGGCTTGGCGCTGGAGTC) targeting upstream of the mutated allele in IBD on the reverse complementary strand so that the middle G of the PAM corresponds to the C to be edited in the leading strand (Fig. 5d). This prevented re-editing by Cas9 after the initial repair because the PAM was destroyed. The HDR template contained a T instead of a C flanked by the DNA sequence around the allele to be edited. The gRNA was synthesized as a 100-nucleotide containing the tracrRNA and 20 nucleotide spacer RNA (IDT). The HDR repair template (IDT) GTGTCCGGCCCCCAACACTCCAGTCCAGCCAGGCCCAGCCCCCTCACTGGACTCCAGCGCCAAGCCCTCGGGCCCCGGGGCCTCGGTGGCCAGTCAGTTCATCCATG was synthesized as single-stranded DNA with phosphorothioate modifications at 5 and 3’ ends to protect it from exonuclease degradation.
Nucleofection of Cas9/gRNA RNP complex and HDR repair template
RNP complex containing gRNA and Cas9 protein (IDT, Alt-R® S.p. Cas9 Nuclease V3, Cat #: 1081058) were prepared by mixing 150 pmol gRNA and 125 pmol of the Cas9 enzyme in 1.5 µL PBS to a total 5 µL reaction volume per well of a 96-well plate. The reaction was incubated at room temperature for 20 min. U937 cells (100,000 cells/well) were washed twice with PBS and resuspended in a solution containing a 4.5:5 mixture of nucleofection buffer (Buffer SE, Lonza Lot #: S-09279) and supplement (Lonza, Lot #: 09588). The Cell suspension was mixed with the 5 µL of RNP complex, 100 µM HDR template, 100 µM Alt-Cas9 electroporation enhancer (IDT cat# 10007805), and PBS to a final volume of 30 µL. The nucleofection mixture was transferred to a 96-well Nucleocuvette (25 µL/well) and electroporated using Lonza 96-well Shuttle following the FP-100 program. After electroporation, the cells were transferred to a 96-well tissue culture plate containing 1.0 µM Alt-R HDR Enhancer V2 (IDT, Cat #: 10007921) in 200 µL of cell culture media. The cells were incubated at 37 °C in a tissue culture incubator for 24 h. After 24 h, the media was changed to regular U937 culture media. Cell culture media was changed every 2–3 days until the cells were confluent.
Characterization of isogenic cell lines
Genomic DNA was isolated using the Quick DNA Miniprep kit (D3025, Zymo Research) according to the manufacturer’s instructions. PCR was performed with the Phusion High-Fidelity PCR kit (E0553L, New England BioLabs) with the following cycling conditions: 98 °C for 3 min, then 35 cycles of 98 °C for 45 s, 71 °C for 30 s, 72 °C for 90 s, then concluding with 72 °C for 10 min. The PCR primer sequences are TNRC18 For1: GATCCCAGTCCAGCCTCTGA and TNRC18 Rev1: GAGGCACTGAGAACGGCACT. PCR product was purified using AMPureXP beads (A63880, Beckman). Nested PCR was performed on the purified PCR amplicon with the following cycling conditions: 98 °C for 3 min, then 35 cycles of 98 °C for 30 s, 69 °C for 30 s, 72 °C for 60 s, then 72 °C for 10 min. The nested PCR primer sequences are: TNRC18 For3: CTTTGGGGTACATGTGGGGA, and TNRC18 Rev3: TTGGGCCGCATGGATGAAC. Amplicons were purified by AMPureXP beads (A63880, Beckman) and submitted to Azenta Life Science for Sanger sequencing on individual clones, CRISPR-Cas9 nucleofected samples, or NGS on bulk edited populations. A successful biallelic edit was indicated by a clean T instead of a C at the specified locus, as visualized by Geneious Prime software. Three different clones were used in the subsequent experiments.
Cleavage under targets and tagmentation (CUT&Tag)
CUT&Tag was performed on two biological replicates per condition using the CUT&Tag-IT Assay Kit (Active Motif, 53172) following manufacturers recommendation. Briefly, 450,000 cells per replicate per condition were bound to the prepared Concanavalin A beads (+spike-in control, Active Motif, 53173) and incubated overnight at 4 °C in a rotator with primary antibodies [1 µg of mouse anti-H3K27Ac (1:50, Active Motif, 91193), 1 µg of positive control mouse anti-H3K4me3 (1:50, Active Motif, 91263)]. The samples were then incubated with Rabbit Anti-Mouse secondary antibody (1:100, Active Motif, 53160) for 1 h at room temperature on a rotator, followed by washes in Dig-Wash buffer. The beads were resuspended with CUT&Tag-IT Assembled pA-Tn5 Transposons and incubated at room temperature on a rotator, followed by washes with Dig-300 buffer. The beads were finally resuspended in tagmentation buffer, and the reactions incubated at 37 °C for 60 min. The DNA was extracted as detailed in the manufacturer’s instructions. Libraries were PCR-amplified using i5 and i7 indexed primers (14 cycles) and primer dimers were removed with AmpureXP bead (Beckman Coulter, A63881) purification. The final libraries were evaluated on Agilent 4200 Tapestation (quality) and Thermo Qubit (quantity) and sequenced on the Illumina NextSeq550 platform (75-bp paired-end reads). Data analysis was performed using HOMER and DeepTools. Raw FASTQ read pairs were pre-processed where adapter, low-quality sequences, duplicates were removed using cutadapt, rmdup. Preprocessed reads were aligned to human genome (hg38) using Bowtie2 and Drosophila genome (spike in control) to generate the normalization factor. Bedtools were used to sort and index, HOMER to generate TagDirectory, peak calls and bedgraph for visualization.
Open targets L2G scores
The Open Targets project has performed systematic fine-mapping and gene prioritization at over 130,000 published GWAS loci from the GWAS catalog, the FinnGen study and the UK Biobank36. The L2G algorithm trained on a set of unequivocally mapped causative genes using a gradient boosting model has enabled the ranking of genes within each GWAS locus by integrating statistical fine-mapping and various epigenetic signatures as predictive features. Multiple tables that contain these data from the October 2022 (22.10) release were downloaded in bulk from Google BigQuery88. We wrangled the data by merging various tables and harmonizing the phenotype descriptions across different GWAS (Supplementary Data 3). We then queried the L2G analysis output for the top 295 genes identified in our genome-wide screen CRIPSR. When multiple studies linked the same gene to the GWAS signals, we chose the study with the highest L2G score for plotting. Top L2G scores between genes and phenotypes above 0.1 were plotted using the ggplot2 R package (v3.3.6).
FinnGen GWAS
FinnGen is a large public-private partnership aiming to collect and analyze genome and health data from 500,000 Finnish biobank participants33. Endpoint definitions derived from electronic health records and curated by clinical expert groups are available in https://www.finngen.fi/en/researchers/clinical-endpoints (DF9). Tools, data QC and association analyzes are described in-depth in https://finngen.gitbook.io/documentation (DF9). In brief, 20,175,454 imputed variants were tested between cases and controls using the REGENIE89 method and software, adjusting for sex, age, 10 principal components and genotyping batch. Summary statistics of GWAS were obtained from FinnGen’s Google Cloud storage. Regional association plot was generated with the stand-alone version of LocusZoom90 (v1.3) using the Finnish population-specific LD structure estimated in 8557 whole-genome sequences of Finns. Forest plot was made using the R package forestplotter (v.0.1.7). Finnish enrichment was calculated as AF(FIN)/AF(NFSEE) in the gnomAD database, where NFSEE stands for non-Finnish-non-Swedish-non-Estonian European populations.
Conditional analysis
Approximate conditional analysis was performed using the COJO approach91 (--cojo-cond command) implemented in the GCTA software tool (v1.93.2)92. For this analysis, we used summary statistics output from the REGENIE GWAS and imputed genotypes from the entire FinnGen cohort for LD estimation within 3 Mb to each direction from the conditioning SNP.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The raw datasets generated in this study have been deposited in the Gene Expression Omnibus (GEO): RNA-seq (under accession code GSE299089), CRISPR screening (GSE299544) and CUT&Tag (GSE299545). Summary statistics from FinnGen GWAS used in this study are available from https://www.finngen.fi/en/access_results (DF9 - May 11, 2023). The screening data are provided in the Supplementary Data Files.
References
Spranger, J. et al. Inflammatory cytokines and the risk to develop type 2 diabetes: results of the prospective population-based European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam Study. Diabetes 52, 812–817 (2003).
Roy, P., Orecchioni, M. & Ley, K. How the immune system shapes atherosclerosis: roles of innate and adaptive immunity. Nat. Rev. Immunol. 22, 251–265 (2022).
Bouras, E. et al. Circulating inflammatory cytokines and risk of five cancers: a Mendelian randomization analysis. BMC Med. 20, 3 (2022).
Glass, C. K., Saijo, K., Winner, B., Marchetto, M. C. & Gage, F. H. Mechanisms underlying inflammation in neurodegeneration. Cell 140, 918–934 (2010).
Neurath, M. F. Cytokines in inflammatory bowel disease. Nat. Rev. Immunol. 14, 329–342 (2014).
Monticelli, S. & Natoli, G. Transcriptional determination and functional specificity of myeloid cells: making sense of diversity. Nat. Rev. Immunol. 17, 595–607 (2017).
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).
Glass, C. K. & Natoli, G. Molecular control of activation and priming in macrophages. Nat. Immunol. 17, 26–33 (2016).
Stender, J. D. & Glass, C. K. Epigenomic control of the innate immune response. Curr. Opin. Pharmacol. 13, 582–587 (2013).
Kominato, Y., Galson, D., Waterman, W. R., Webb, A. C. & Auron, P. E. Monocyte expression of the human prointerleukin 1 beta gene (IL1B) is dependent on promoter sequences which bind the hematopoietic transcription factor Spi-1/PU.1. Mol. Cell. Biol. 15, 58–68 (1995).
Lopez-Castejon, G. & Brough, D. Understanding the mechanism of IL-1beta secretion. Cytokine Growth Factor Rev. 22, 189–195 (2011).
Weber, A., Wasiliew, P. & Kracht, M. Interleukin-1beta (IL-1beta) processing pathway. Sci. Signal. 3, cm2 (2010).
Sims, J. E. & Smith, D. E. The IL-1 family: regulators of immunity. Nat. Rev. Immunol. 10, 89–102 (2010).
Caliskan, M., Brown, C. D. & Maranville, J. C. A catalog of GWAS fine-mapping efforts in autoimmune disease. Am. J. Hum. Genet. 108, 549–563 (2021).
Gerussi, A., Soskic, B., Asselta, R., Invernizzi, P. & Gershwin, M. E. GWAS and autoimmunity: what have we learned and what next. J. Autoimmun. 133, 102922 (2022).
Cotsapas, C. et al. Pervasive sharing of genetic effects in autoimmune disease. PLoS Genet. 7, e1002254 (2011).
Ellinghaus, D. et al. Analysis of five chronic inflammatory diseases identifies 27 new associations and highlights disease-specific patterns at shared loci. Nat. Genet. 48, 510–518 (2016).
Harroud, A. & Hafler, D. A. Common genetic factors among autoimmune diseases. Science 380, 485–490 (2023).
Schett, G., McInnes, I. B. & Neurath, M. F. Reframing immune-mediated inflammatory diseases through signature cytokine hubs. N. Engl. J. Med. 385, 628–639 (2021).
Nelson, M. R. et al. The support of human genetic evidence for approved drug indications. Nat. Genet. 47, 856–860 (2015).
King, E. A., Davis, J. W. & Degner, J. F. Are drug targets with genetic support twice as likely to be approved? Revised estimates of the impact of genetic support for drug mechanisms on the probability of drug approval. PLoS Genet. 15, e1008489 (2019).
Ochoa, D. et al. Human genetics evidence supports two-thirds of the 2021 FDA-approved drugs. Nat. Rev. Drug Discov. 21, 551 (2022).
Minikel, E. V., Painter, J. L., Dong, C. C. & Nelson, M. R. Refining the impact of genetic evidence on clinical success. Nature 629, 624–629 (2024).
Papp, K. et al. Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis. N. Engl. J. Med. 379, 1313–1321 (2018).
Leonardi, C. L. et al. Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with psoriasis: 76-week results from a randomised, double-blind, placebo-controlled trial (PHOENIX 1). Lancet 371, 1665–1674 (2008).
Sands, B. E. et al. Infliximab maintenance therapy for fistulizing Crohn’s disease. N. Engl. J. Med. 350, 876–885 (2004).
Sandborn, W. J. et al. Tofacitinib as induction and maintenance therapy for ulcerative colitis. N. Engl. J. Med. 376, 1723–1736 (2017).
Feagan, B. G. et al. Ustekinumab as induction and maintenance therapy for Crohn’s disease. N. Engl. J. Med. 375, 1946–1960 (2016).
Simpson, E. L. et al. Two phase 3 trials of dupilumab versus placebo in atopic dermatitis. N. Engl. J. Med. 375, 2335–2348 (2016).
Papp, K. et al. Long-term safety and disease control with ruxolitinib cream in atopic dermatitis: results from two phase 3 studies. J. Am. Acad. Dermatol. 88, 1008–1016 (2023).
Dinarello, C. A. Overview of the IL-1 family in innate inflammation and acquired immunity. Immunol. Rev. 281, 8–27 (2018).
Kaneko, N., Kurata, M., Yamamoto, T., Morikawa, S. & Masumoto, J. The role of interleukin-1 in general pathology. Inflamm. Regen. 39, 12 (2019).
Kurki, M. I. et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature 613, 508–518 (2023).
Kaizuka, T. et al. An autophagic flux probe that releases an internal control. Mol. Cell 64, 835–849 (2016).
Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nat. Biotechnol. 34, 184–191 (2016).
Mountjoy, E. et al. An open approach to systematically prioritize causal variants and genes at all published human GWAS trait-associated loci. Nat. Genet. 53, 1527–1533 (2021).
Ochoa, D. et al. The next-generation Open Targets Platform: reimagined, redesigned, rebuilt. Nucleic Acids Res. 51, D1353–D1359 (2023).
Lee, E. G. et al. Failure to regulate TNF-induced NF-kappaB and cell death responses in A20-deficient mice. Science 289, 2350–2354 (2000).
Duong, B. H. et al. A20 restricts ubiquitination of pro-interleukin-1beta protein complexes and suppresses NLRP3 inflammasome activity. Immunity 42, 55–67 (2015).
Vande Walle, L. et al. Negative regulation of the NLRP3 inflammasome by A20 protects against arthritis. Nature 512, 69–73 (2014).
Adrianto, I. et al. Association of a functional variant downstream of TNFAIP3 with systemic lupus erythematosus. Nat. Genet. 43, 253–258 (2011).
Wang, S., Wen, F., Wiley, G. B., Kinter, M. T. & Gaffney, P. M. An enhancer element harboring variants associated with systemic lupus erythematosus engages the TNFAIP3 promoter to influence A20 expression. PLoS Genet. 9, e1003750 (2013).
Pasula, S. et al. Systemic lupus erythematosus variants modulate the function of an enhancer upstream of TNFAIP3. Front. Genet. 13, 1011965 (2022).
Ray, J. P. et al. Prioritizing disease and trait causal variants at the TNFAIP3 locus using functional and genomic features. Nat. Commun. 11, 1237 (2020).
Lin, C. C. & Edelson, B. T. New insights into the role of IL-1beta in experimental autoimmune encephalomyelitis and multiple sclerosis. J. Immunol. 198, 4553–4560 (2017).
Lim, J. et al. Mixed tailing by TENT4A and TENT4B shields mRNA from rapid deadenylation. Science 361, 701–704 (2018).
International Multiple Sclerosis Genetics Consortium. Multiple sclerosis genomic map implicates peripheral immune cells and microglia in susceptibility. Science 365, eaav7188 (2019).
Vosa, U. et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat. Genet. 53, 1300–1310 (2021).
Quach, H. et al. Genetic adaptation and neandertal admixture shaped the immune system of human populations. Cell 167, 643–656.e617 (2016).
Schmiedel, B. J. et al. Impact of genetic polymorphisms on human immune cell gene expression. Cell 175, 1701–1715.e1716 (2018).
Fairfax, B. P. et al. Innate immune activity conditions the effect of regulatory variants upon monocyte gene expression. Science 343, 1246949 (2014).
Alasoo, K. et al. Shared genetic effects on chromatin and gene expression indicate a role for enhancer priming in immune response. Nat. Genet. 50, 424–431 (2018).
Ota, M. et al. Dynamic landscape of immune cell-specific gene regulation in immune-mediated diseases. Cell 184, 3006–3021.e3017 (2021).
Smith, K. T. et al. Human family with sequence similarity 60 member A (FAM60A) protein: a new subunit of the Sin3 deacetylase complex. Mol. Cell. Proteom. 11, 1815–1828 (2012).
Sardiu, M. E. et al. Suberoylanilide hydroxamic acid (SAHA)-induced dynamics of a human histone deacetylase protein interaction network. Mol. Cell. Proteom. 13, 3114–3125 (2014).
Adams, M. K. et al. Differential complex formation via paralogs in the human Sin3 protein interaction network. Mol. Cell. Proteom. 19, 1468–1484 (2020).
Oughtred, R. et al. The BioGRID database: a comprehensive biomedical resource of curated protein, genetic, and chemical interactions. Protein Sci. 30, 187–200 (2021).
Zhao, S. et al. TNRC18 engages H3K9me3 to mediate silencing of endogenous retrotransposons. Nature 623, 633–642 (2023).
Abe, Y. et al. RANK ligand converts the NCoR/HDAC3 co-repressor to a PGC1beta- and RNA-dependent co-activator of osteoclast gene expression. Mol. Cell 83, 3421–3437 e3411 (2023).
Abe, Y. et al. A TLR4/TRAF6-dependent signaling pathway mediates NCoR coactivator complex formation for inflammatory gene activation. Proc. Natl Acad. Sci. USA 121, e2316104121 (2024).
Chen, X. et al. Requirement for the histone deacetylase Hdac3 for the inflammatory gene expression program in macrophages. Proc. Natl Acad. Sci. USA 109, E2865–E2874 (2012).
Kaya-Okur, H. S. et al. CUT&Tag for efficient epigenomic profiling of small samples and single cells. Nat. Commun. 10, 1930 (2019).
Abdellaoui, A., Yengo, L., Verweij, K. J. H. & Visscher, P. M. 15 years of GWAS discovery: realizing the promise. Am. J. Hum. Genet. 110, 179–194 (2023).
Inshaw, J. R. J., Cutler, A. J., Burren, O. S., Stefana, M. I. & Todd, J. A. Approaches and advances in the genetic causes of autoimmune disease and their implications. Nat. Immunol. 19, 674–684 (2018).
Polson, H. E. et al. Mammalian Atg18 (WIPI2) localizes to omegasome-anchored phagophores and positively regulates LC3 lipidation. Autophagy 6, 506–522 (2010).
Yamamoto, H., Zhang, S. & Mizushima, N. Autophagy genes in biology and disease. Nat. Rev. Genet. 24, 1–19 (2023).
Levine, B., Mizushima, N. & Virgin, H. W. Autophagy in immunity and inflammation. Nature 469, 323–335 (2011).
Filipi, M. & Jack, S. Interferons in the treatment of multiple sclerosis: a clinical efficacy, safety, and tolerability update. Int J. MS Care 22, 165–172 (2020).
Lai, Y. et al. Illuminating host-mycobacterial interactions with genome-wide CRISPR knockout and CRISPRi screens. Cell Syst. 11, 239–251.e237 (2020).
Joshi, P. et al. The functional interactome landscape of the human histone deacetylase family. Mol. Syst. Biol. 9, 672 (2013).
Munoz, I. M. et al. Family with sequence similarity 60A (FAM60A) protein is a cell cycle-fluctuating regulator of the SIN3-HDAC1 histone deacetylase complex. J. Biol. Chem. 287, 32346–32353 (2012).
Huttlin, E. L. et al. Dual proteome-scale networks reveal cell-specific remodeling of the human interactome. Cell 184, 3022–3040.e3028 (2021).
Hein, M. Y. et al. A human interactome in three quantitative dimensions organized by stoichiometries and abundances. Cell 163, 712–723 (2015).
Fasci, D., van Ingen, H., Scheltema, R. A. & Heck, A. J. R. Histone interaction landscapes visualized by crosslinking mass spectrometry in intact cell nuclei. Mol. Cell. Proteom. 17, 2018–2033 (2018).
Cox, J. L. et al. The SOX2-interactome in brain cancer cells identifies the requirement of MSI2 and USP9X for the growth of brain tumor cells. PLoS ONE 8, e62857 (2013).
Giurato, G. et al. Quantitative mapping of RNA-mediated nuclear estrogen receptor beta interactome in human breast cancer cells. Sci. Data 5, 180031 (2018).
Velot, L. et al. Proximity-dependent mapping of the androgen receptor identifies Kruppel-like factor 4 as a functional partner. Mol. Cell. Proteom. 20, 100064 (2021).
Bierne, H. et al. Human BAHD1 promotes heterochromatic gene silencing. Proc. Natl Acad. Sci. USA 106, 13826–13831 (2009).
Faure, G. & Callebaut, I. Identification of hidden relationships from the coupling of hydrophobic cluster analysis and domain architecture information. Bioinformatics 29, 1726–1733 (2013).
Ramberger, E. et al. PRISMA and BioID disclose a motifs-based interactome of the intrinsically disordered transcription factor C/EBPalpha. iScience 24, 102686 (2021).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010).
Robinson, M. D. & Oshlack, A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 11, R25 (2010).
Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Kolde, R., Laur, S., Adler, P. & Vilo, J. Robust rank aggregation for gene list integration and meta-analysis. Bioinformatics 28, 573–580 (2012).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523 (2019).
Ghoussaini, M. et al. Open Targets Genetics: systematic identification of trait-associated genes using large-scale genetics and functional genomics. Nucleic Acids Res. 49, D1311–D1320 (2021).
Mbatchou, J. et al. Computationally efficient whole-genome regression for quantitative and binary traits. Nat. Genet. 53, 1097–1103 (2021).
Pruim, R. J. et al. LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics 26, 2336–2337 (2010).
Yang, J. et al. Conditional and joint multiple-SNP analysis of GWAS summary statistics identifies additional variants influencing complex traits. Nat. Genet. 44, 369–375 (2012). S361-363.
Yang, J., Lee, S. H., Goddard, M. E. & Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82 (2011).
Acknowledgements
We would like to thank the Genomic Technologies lab of the AbbVie Genomics Research Center for preparing and sequencing NGS libraries. We also thank AbbVie employees Nizar Smaoui, Archana Iyer, and Kathleen Smith for critical review of the manuscript.
Author information
Authors and Affiliations
Contributions
F.R. and J.D.S. conceived, designed and led the study and drafted the manuscript. S.G., S.P., M.S., E.N., J.T., D.V., S.S., C.A.E., A.K., T.A., performed experiments, M.S., V.A., A.M., C.L., J.W., M.M., N.A.M., analyzed the data, N.C., M.J.F., K.V.K., A.B., A.I.D., J.F.W. provided resources and critical feedback. All authors reviewed the manuscript, provided feedback, and approved the final version of the manuscript.
Corresponding authors
Ethics declarations
Competing interests
All authors were or are employees of AbbVie at the time of the study. The design, study conduct, and financial support for this research were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the publication. No honoraria or payments were made for authorship.
Peer review
Peer review information
Nature Communications thanks anonymous reviewer(s) for their contribution to the peer review of this work. [A peer review file is available].
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Rahimov, F., Ghosh, S., Petiwala, S. et al. A genome-wide CRISPR screen identifies the TNRC18 gene locus as a regulator of inflammatory signaling. Nat Commun 16, 10346 (2025). https://doi.org/10.1038/s41467-025-65277-y
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-65277-y
