Threonine phosphorylation of STAT1 restricts interferon signaling and promotes innate inflammatory responses

Significance

For decades, signal transducers and activators of transcription (STAT) proteins have been central to our understanding of immunity. STAT1, a central signaling mediator for interferons (IFN) and antiviral defense, can alter inflammatory responses, yet the scope and mechanisms remain elusive. Here, we unveil Thr⁷⁴⁸ as a conserved IFN-independent phosphorylation switch in Stat1, which restricts IFN signaling and promotes innate inflammatory responses. Together with the canonical tyrosine and serine, our identification of the Thr⁷⁴⁸ completes the phosphorylation catalog of the transactivation domain of STAT1 and more generally other STAT proteins. Our study indicates that Stat1 displays a phosphorylation-dependent modular functionality in innate immune responses, which may offer opportunities for better and more specific treatments for inflammatory diseases such as sepsis.

Abstract

Since its discovery over three decades ago, signal transducer and activator of transcription 1 (STAT1) has been extensively studied as a central mediator for interferons (IFNs) signaling and antiviral defense. Here, using genetic and biochemical assays, we unveil Thr⁷⁴⁸ as a conserved IFN-independent phosphorylation switch in Stat1, which restricts IFN signaling and promotes innate inflammatory responses following the recognition of the bacterial-derived toxin lipopolysaccharide (LPS). Genetically engineered mice expressing phospho-deficient threonine748-to-alanine (T748A) mutant Stat1 are resistant to LPS-induced lethality. Of note, T748A mice exhibited undisturbed IFN signaling, as well as total expression of Stat1. Further, the T748A point mutation of Stat1 recapitulates the safeguard effect of the genetic ablation of Stat1 following LPS-induced lethality, indicating that the Thr⁷⁴⁸ phosphorylation contributes inflammatory functionalities of Stat1. Mechanistically, LPS-induced Toll-like receptor 4 endocytosis activates a cell-intrinsic IκB kinase-mediated Thr⁷⁴⁸ phosphorylation of Stat1, which promotes macrophage inflammatory response while restricting the IFN and anti-inflammatory responses. Depletion of macrophages restores the sensitivity of the T748A mice to LPS-induced lethality. Together, our study indicates a phosphorylation-dependent modular functionality of Stat1 in innate immune responses: IFN phospho-tyrosine dependent and inflammatory phospho-threonine dependent. Better understanding of the Thr⁷⁴⁸ phosphorylation of Stat1 may uncover advanced pharmacologically targetable molecules and offer better treatment modalities for sepsis, a disease that claims millions of lives annually.

In multicellular organisms, cells recognize and respond to their environment by utilizing networks of signaling molecules that act as sensors, information processors, and effectors to mount an appropriate response (1). This is best illustrated in the context of host–pathogen interaction and innate immune signaling (2). Over three decades ago, experiments on the antiviral mechanisms of interferons (IFNs) uncovered the Janus kinase (JAK)-signal transducers and activators of transcription (STAT) pathway, which has since represented a paradigm for membrane-to-nucleus signaling. STAT proteins have been central to our understanding of the innate immune system and more generally immunity (3).

The canonical JAK-STAT signaling posits that STAT proteins remain latent in the cytoplasm until phosphorylated on a key tyrosine residue within their C-terminal transactivation domain (TAD), which is mediated by a receptor-associated JAK complex (4). In mammals, the STAT family comprises seven evolutionary conserved proteins: Stat1, 2, 3, 4, 5a, 5b, and 6. Stat1 is ubiquitously expressed and is composed of a full-length α isoform and its splice variant β isoform, which lacks the last 38 amino acids of its TAD, suggesting a possible contribution of these amino acids to the specificity of the transcriptional activities of Stat1 (5). The JAK-mediated Tyr⁷⁰¹ phosphorylation of Stat1 is central for all types of IFN signaling and the transcription of IFN-stimulated genes (ISGs) (6). Several posttranslational modifications such as serine phosphorylation on 708 and 727 contribute to the selectivity of the transcriptional activity of Stat1 following IFN stimulation (7, 8).

Innate immune responses rely on the recognition of evolutionarily conserved structures on pathogens, termed pathogen-associated molecular patterns, through pattern recognition receptors, of which the family of Toll-like receptors (TLRs) has been studied most extensively (9). TLRs can be classified according to their cellular localization into cell surface such as TLR1, 2, 4, 5, and 6 located at the plasma membrane and recognize microbial cell surface molecules or endosomal that detect nucleic acids such as TLR3, 7, 8, and 9. Such spatial regulation of TLRs is critical for the downstream signaling activation and eventually the expression of distinct downstream genes. Surface TLRs activate NF-κB signaling and inflammatory response, while endosomal TLRs activate IRFs and type I IFN response (10, 11).

TLR4 is unique among TLRs in its ability to activate two distinct signaling pathways. A cell surface MyD88 dependent pathway, which activates inflammatory cytokines expression. Then, TLR4 is endocytosed and activates endosomal TRIF signaling, which promotes type I IFN expression, thus amplifying the signal and culminating in a strong immune response against LPS (12). Despite their key role in the antiviral defense, the contribution of IFNs to the host defense against LPS is controversial, with emerging evidence supporting both pro- and anti-inflammatory properties. On the other hand, deficiency of Stat1 safeguards the host against the LPS-induced lethality, suggesting that the role of Stat1 in the innate immune response against LPS extends beyond the canonical IFN-JAK signaling (13–18). Supporting this premise, we have shown that LPS-induced TLR4 endocytosis activates a potential alternative inflammatory signaling pathway in human macrophages where the IKK-related kinase beta (IKKβ) mediates the Thr⁷⁴⁹ (equivalent to mouse Thr⁷⁴⁸) phosphorylation of STAT1, which promotes the production of the proinflammatory cytokines independently of the canonical IFN-JAK pathway. The Thr⁷⁴⁹ phosphorylation of STAT1 alters its target DNA motif and stimulates the expression of the genes encoding IL-12p40 and the protein ARID5A, which enhanced the stability of IL6 mRNA (19).

Sepsis is a major and rising health threat with around 48.9 million cases recorded and 11 million deaths worldwide in 2017 (20). In this report, we explored the physiological relevance of Thr⁷⁴⁸ STAT1 in regulating the host’s defense and inflammatory response following sepsis. We uncovered Thr⁷⁴⁸ as a conserved IFN-independent inflammatory phosphorylation switch in Stat1, which restricts IFN signaling and promotes innate inflammatory responses following LPS-induced sepsis.

Results

Disruption of the Thr⁷⁴⁸ Phosphorylation Recapitulates the Safeguard Effect of the Genetic Ablation of Stat1 Following LPS-Induced Lethality.

To address the physiological relevance of the Thr⁷⁴⁸ phosphorylation of Stat1 and its potential role in promoting the host inflammatory response, we generated Stat1 knock-in mice bearing phospho-deficient Thr748-to-Ala (T748A) mutation using CRISPR/Cas9 (SI Appendix, Fig. S1). We next examined the survival of Wt/Wt (hereafter referred to as Wt), Wt/Stat1T748A (heterozygous), and Stat1T748A/Stat1T748A (hereafter referred to as T748A) littermates following LPS-induced lethality. Intriguingly, haplo- and complete-deficiency of Thr⁷⁴⁸ phosphorylation of Stat1 safeguards the host following LPS-induced lethality as shown by the survival of the T748A mice (83% survival, 20 out of 24 survived), heterozygous littermates (48% survival, 11 out of 23 survived) while Wt littermates succumbed to LPS-induced lethality (0 out of 21 survived) (Fig. 1A).

Fig. 1.

Disruption of the Thr⁷⁴⁸ phosphorylation recapitulates the safeguard effect of the genetic ablation of Stat1 following LPS-induced lethality. (A) Survival of Wt, Wt/T748A, and T748A littermates following LPS challenge (12.5 mg/kg) (n = 21 to 24 mice per genotype). (B) Survival of KO and T748A littermates and age- and sex-matched Wt mice following LPS challenge (12.5 mg/kg) (n = 21 to 23 mice for KO and T748A and n = 7 for Wt). *P < 0.05, **P < 0.01, and ***P < 0.005 as measured by the log-rank test (Mantel–Cox).

We next investigated the extent to which the Thr⁷⁴⁸ phosphorylation contributes to the entirety of the Stat1 function following LPS-induced lethality; we generated Stat1 knock-out (hereafter referred to as KO) and T748A littermates and examined their survival following LPS-induced lethality. As expected, genetic ablation of Stat1 safeguards the host against LPS-induced lethality as shown by the resistance of the KO mice to LPS-induced lethality (21 out of 21 survived). Strikingly, disruption of the Thr⁷⁴⁸ phosphorylation largely recapitulated the safeguard effect of the genetic ablation of Stat1 following LPS-induced lethality as shown by the survival of the T748A mice (74%, 17 out of 23 survived) compared to their KO littermates (Fig. 1B). Our in vivo analysis shows that the Thr⁷⁴⁸ phosphorylation is necessary for the Stat1-mediated inflammatory activities and the host defense against LPS.

T748A Mice Exhibit Intact Total Expression of Stat1 and IFN Phospho-Tyrosine Signaling at Naive and LPS-Challenged Conditions.

We next examined whether the Thr⁷⁴⁸ phosphorylation mediates Stat1 inflammatory activities by regulating its total protein expression or the canonical IFN-mediated phospho-tyrosine signaling. Of note, the T748A mutation does not affect the expression of total Stat1 protein as shown by the comparable expression of Stat1 in Wt, heterozygous, and T748A littermates (Fig. 2 A and B). Besides, the expression levels of other STAT proteins such as Stat2, 3, and 5 and inflammatory proteins IkBα and p42/44 MAPK were comparable between T748A mice and their littermates at the basal level (Fig. 2C).

Fig. 2.

T748A mice exhibit intact total expression of Stat1 and IFN phospho-tyrosine signaling at naive and LPS-challenged conditions. (A) Splenocytes from naive Wt, Wt/Stat1T748A, and T748A littermates and age- and sex-matched naive KO mice. Whole-cell lysates were harvested and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Data show three to four biological replicates per genotype. (B) Quantitation of Stat1/Actb mean band intensity of (A). (C) Splenocytes from naive Wt and T748A littermates and age- and sex-matched naive KO mice. Whole-cell lysates were harvested and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Data show two independent experiments. (D) Splenocytes from untreated or LPS-treated (7.5 mg/kg for 4 hrs) Wt and T748A littermates. Splenocytes from age- and sex-matched naive KO mice were used as control. Whole-cell lysates were harvested and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Data are representative of three independent experiments. ns is not significant as measured by one-way ANOVA with post hoc Tukey’s test in (B).

Notably, T748A mice exhibit comparable IFN signaling to their Wt littermates as shown by the expression of phospho-Stat1, Stat2, and Irf3 following LPS challenge (Fig. 2D). Besides, the expression of the negative regulator of IFN-Stat1 signaling, Socs1, was comparable between T748A and Wt littermates (Fig. 2D). Moreover, the expression of inflammatory signaling mediators p65 NF-κB, IkBα, and p38 MAPK was comparable between T748A and Wt littermates at basal and LPS-stimulated conditions (Fig. 2D). Our ex vivo analysis show that the Thr⁷⁴⁸ phosphorylation promotes Stat1-mediated inflammatory activities without affecting the IFN phospho-tyrosine signaling or the total expression of Stat1.

Single-Cell Analysis of the Immunologic Landscape of Wt and T748A Littermates.

We next investigated whether the Thr⁷⁴⁸ phosphorylation of Stat1 regulates the host’s inflammatory response following LPS challenge through modulating the host’s immunological landscape. Using mass cytometry (CyTOF), we performed single-cell proteomics analysis of splenocytes from Wt and T748A littermates at both basal and LPS-challenged conditions. We surveyed 24 simultaneous cellular parameters spanning diverse repertoire of immune cells including dendritic cells, macrophages, monocytes, neutrophils, natural killer cells, B cells, CD4, CD8, and gamma delta T cells (Fig. 3 A and B and SI Appendix, Fig. S2).

Fig. 3.

Single-cell analysis of the immunologic landscape of Wt and T748A littermates. (A) The heatmap shows aggregated median marker expressions per immune subset. (B) UMAP feature P of splenocytes assayed by mass cytometry. Cell types were categorized according to canonical marker expression patterns as shown in (A). (C) UMAP overlayed with cell density contour plots. Each condition is scaled individually. (D) Boxplots of the mean of pTyr Stat1 and pNFkB expression from each immune subset (n = 4 to 5 mice per genotype per group). *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001 as measured by one-way ANOVA with post hoc Tukey’s test in (D).

T748A and Wt littermates showed comparable immunologic landscape at both basal and LPS-challenged conditions as shown by the UMAP analysis (Fig. 3C), immune cell numbers and percentages (SI Appendix, Fig. S3), and diffcyt volcano plot analysis (SI Appendix, Fig. S4). Of note, the canonical IFN-mediated Tyr⁷⁰¹ phosphorylation of Stat1 and the phosphorylation of the canonical inflammatory marker, NF-κB, were comparable among different immune cell populations of the T748A mice and their Wt littermates at both basal and LPS-challenged conditions (Fig. 3D). Our single-cell analysis shows that the T748A mutation does not affect the host’s immunological landscape or IFN-mediated tyrosine and NF-κB activation in the examined immune cells.

Thr⁷⁴⁸ Phosphorylation of Stat1 Promotes Macrophage Inflammatory Response while Restricting the IFN and Anti-Inflammatory Responses Following LPS Stimulation.

Macrophages are key players in shaping the inflammatory response following LPS challenge. The balance between pro- and anti-inflammatory functions of macrophages is critical for the host survival following sepsis (21). Using bone marrow–derived macrophages (BMDMs), we investigated the potential role of the Thr⁷⁴⁸ phosphorylation of Stat1 in shaping macrophage inflammatory response. BM monocytes, the precursors of BMDMs, were comparable between T748A and their Wt littermates at the basal condition as shown by the flow cytometer (FACS) analysis (SI Appendix, Fig. S5). To analyze the phosphorylation status of Stat1, we developed cross-reactive phospho-specific rabbit monoclonal antibodies against Thr⁷⁴⁸ in mouse and Thr⁷⁴⁹ in human STAT1 (SI Appendix, Fig. S6). Remarkably, LPS stimulation of BMDMs activates the Thr⁷⁴⁸ phosphorylation of Stat1 (Fig. 4A). Of note, the phosphorylation of Irf3 and p65 and the degradation of IkBα were comparable between LPS-stimulated T748A and Wt BMDMs (Fig. 4A). Strikingly, T748A BMDMs show enhanced expression of type I IFN signaling following LPS stimulation as manifested by the higher expression of the phosphorylation of Stat1 and Stat2 (Fig. 4A).

Fig. 4.

Thr⁷⁴⁸ phosphorylation of Stat1 promotes macrophage inflammatory response while restricting the IFN and anti-inflammatory responses following LPS stimulation. (A) BMDMs were stimulated with LPS (100 ng/mL). Whole-cell lysates were harvested at indicated time points and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. (B) BMDMs were stimulated with LPS (100 ng/mL). Total RNA was isolated at indicated time points. The indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). (C) PCA of gene expression profiles of unstimulated and LPS-stimulated Wt and T748A BMDMs. (D) Venn diagrams show overlaps of DEGs (FDR ≤ 0.05) between unstimulated and LPS-stimulated Wt and T748A BMDMs. (E) PANTHER GO-slim pathway analyses of up- and down-regulated genes in unstimulated and LPS-stimulated Wt and T748A BMDMs. (F) Heatmaps of expression levels of indicated genes in unstimulated and LPS-stimulated Wt and T748A BMDMs. (G) Healthy donor MDMs were stimulated with LPS (100 ng/mL). Whole-cell lysates were harvested at indicated time points and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. *P < 0.05, **P < 0.01, and ***P < 0.005 as measured by one-way ANOVA with post hoc Tukey’s test in (B).

Correspondingly, LPS-stimulated T748A BMDMs exhibited higher expression of Ifnb1 and its downstream ISGs, Rsad2 compared to their Wt counterparts (Fig. 4B). Further, T748A BMDMs showed higher expression of the anti-inflammatory cytokine, Il10, following LPS stimulation (Fig. 4B). By contrast, LPS-stimulated T748A BMDMs exhibited diminished expression of proinflammatory cytokines; Il6 and Il12b compared to their Wt counterparts, while the expression of the proinflammatory cytokine, Tnf was comparable (Fig. 4B).

Transcriptomics analysis of T748A and Wt BMDMs at basal and LPS-stimulated conditions showed robust changes of their gene expression as shown by the principal component analysis (PCA) (Fig. 4C), scatter plots, and heat maps (SI Appendix, Fig. S7). T748A BMDMs displayed upregulation of 364 and 1,127 differentially expressed genes (DEGs) at basal and LPS-stimulated conditions, respectively, and downregulation of 468 and 688 DEGs at basal and LPS-stimulated conditions, respectively (Fig. 4D). T748A BMDMs exhibited a boosted type I IFN signaling as shown by the gene ontology (GO) pathway analysis (Fig. 4E) and the higher expression of diverse ISGs such as Ifnb1, Rsad2, Ifit1, Ifit2, Ifit3, Irf7, Irf8, Mx2, and Iigp1 (SI Appendix, Fig. S8). Disruption of the Thr⁷⁴⁸ phosphorylation of Stat1 affects genes regulating diverse cellular processes critical for macrophage function such as cytokine expression, adhesion, chemotaxis, defense, and metabolism (Fig. 4F).

Furthermore, LPS stimulation induced the Thr⁷⁴⁹ phosphorylation as well as the canonical Tyr⁷⁰¹ and Ser⁷²⁷ phosphorylation of Stat1 in human macrophages (Fig. 4G), indicating that the threonine phosphorylation of Stat1 is a conserved inflammatory controller in mouse and human macrophages.

Thr⁷⁴⁸ Phosphorylation of Stat1 Is a General Mechanism Utilized by Macrophages to Promote the Inflammatory Response at the Expense of the IFN and Anti-Inflammatory Responses Following LPS Stimulation.

To investigate whether the Thr⁷⁴⁸ phosphorylation of Stat1 is a general mechanism for macrophages to promote the innate inflammatory response, we utilized peripheral blood monocyte–derived macrophages (PBMDM) and peritoneal cavity macrophages (PCMs). Remarkably, similar to their BMDMs counterparts, T748A PBMDMs exhibited similar pattern of gene expression with enhanced ISGs such as Ifit1 and Rsad2; enhanced anti-inflammatory Il10; diminished proinflammatory Il6 and Il12b; and comparable Tnf expression following LPS stimulation (Fig. 5A). In addition, T748A thioglycolate-elicited PCMs showed similar pattern of gene expression with enhanced Ifnb and Rsad2; enhanced Il10, albeit not statistically significant; diminished Il6 and Il12b; and comparable Tnf expression following LPS stimulation (Fig. 5B).

Fig. 5.

Thr⁷⁴⁸ phosphorylation of Stat1 is a general mechanism utilized by macrophages to promote the inflammatory response at the expense of the IFN and anti-inflammatory responses following LPS stimulation. (A) PBMDMs were unstimulated or stimulated with LPS (100 ng/mL) for 4 h. Total RNA was isolated, and the indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). (B) PCMs were stimulated with LPS (100 ng/mL). Total RNA was isolated at indicated time points. The indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). (C) PCA of gene expression profiles of unstimulated and LPS-stimulated Wt and T748A PCMs. (D) Venn diagrams show overlaps of DEGs (FDR ≤ 0.05) between unstimulated and LPS-stimulated Wt and T748A PCMs. (E) PANTHER GO-slim pathway analyses of up- and down-regulated genes in unstimulated and LPS-stimulated Wt and T748A PCMs. (F) Indicated RAW 264.7-Stat1 were unstimulated or stimulated with LPS (100 ng/mL) for 4 h. Total RNA was isolated, and the indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001 as measured by one-way ANOVA with post hoc Tukey’s test in (A, B, and F). ns, not significant.

Transcriptomics analysis of T748A and Wt PCMs at basal and LPS-stimulated conditions shows robust changes of their gene expression as shown by the PCA (Fig. 5C), scatter plots, and heat maps (SI Appendix, Fig. S9). T748A PCMs displayed upregulation of 82 and 182 DEGs at basal and LPS-stimulated conditions, respectively, and downregulation of 40 and 145 DEGs at basal and LPS-stimulated conditions, respectively (Fig. 5D). Similar to their BMDMs counterparts, T748A PCMs exhibited a boosted type I IFN signaling as shown by the GO pathway analysis (Fig. 5E).

We next examined whether the observed effect of the Thr⁷⁴⁸ phosphorylation of Stat1 on macrophage immune responses is attributed to a phenotypic change of macrophages toward an anti-inflammatory type. Of note, T748A did not affect the macrophage phenotypic polarization as shown by the comparable expression of different phenotypic markers such as Cd68, AdgreI, Cd80, Cd86, Arg1, Mrc1, and Cd163 between Wt and T748A macrophages (SI Appendix, Fig. S10).

Phosphorylation sites in proteins are suggested to be originated from acidic amino acids such as glutamic acid (Glu), and a tandem Glu.Glu has been reported to mimic a phosphorylation site better than a single Glu mutant (22). Mouse Stat1 has been shown to be phosphorylated at Ser⁷⁴⁷ (8). Therefore, we generated a Ser747Glu.Thr748Glu (hereafter referred to as S747E/T748E) double mutant to mimic the Thr⁷⁴⁸ phosphorylation of Stat1 (SI Appendix, Fig. S11A). To examine the effect of the phosphomimic S747E/T748E Stat1 mutant, we generated Stat1 KO murine macrophages, RAW 264.7 (hereafter referred to as RAW 264.7 KO) using CRISPR/Cas9, followed by lentiviral-reconstitution with Stat1 Wt, T748A (phospho-deficient), or S747E/T748E (phosphomimic) mutants, hereafter referred to as RAW 264.7-Stat1 (SI Appendix, Fig. S11 B and C). As expected, LPS-stimulated RAW 264.7-Stat1T748A exhibited lower expression of Il6 while showing higher expression of Rsad2 and Il10 compared to RAW 264.7-Stat1Wt (Fig. 5F). In contrast, LPS-stimulated RAW 264.7-Stat1S747E/T748E exhibited higher expression of Il6 while showing lower expression of Rsad2 and Il10 compared to RAW 264.7-Stat1Wt (Fig. 5F). Collectively, our observations show that LPS induces the Thr⁷⁴⁸ phosphorylation of Stat1, which promotes macrophage inflammatory response at the expense of IFN and anti-inflammatory responses.

Dual Activation of Stat1 Downstream of LPS-Induced TLR4 Endocytosis: Autocrine/Paracrine IFN Tyrosine Dependent and Cell-Intrinsic Inflammatory Threonine Dependent.

Surface LPS-bound TLR4 activates MYD88 signaling and inflammatory cytokines expression, followed by LPS-induced TLR4 endocytosis and the activation of endosomal TRIF signaling and type I IFN expression (23). We have shown that in human macrophages, STAT1, potentially through its Thr⁷⁴⁹ phosphorylation, contributes to the production of IL-6 and IL12-p40 downstream of LPS-induced TLR4 endocytosis, independently of IFN and NF-κB signaling (19). Therefore, we examined whether TLR4 endocytosis activates the Thr⁷⁴⁸ phosphorylation of Stat1 following LPS stimulation in murine macrophages. As expected, endocytosis inhibitor MitMAB abolished the LPS-induced IFN-mediated Tyr⁷⁰¹ phosphorylation of Stat1 in LPS-stimulated BMDMs (Fig. 6A). Remarkably, MitMAB ablates the expression of the Thr⁷⁴⁸ phosphorylation of Stat1 in LPS-stimulated BMDMs (Fig. 6A). Correspondingly, MitMAB did not affect the MYD88-driven Tnf expression (Fig. 6B). By contrast, MitMAB diminished the expression of Il6 and Il12b and abolished the expression of Ifit1, Rsad2, and Il10 resulting in comparable expression of the examined cytokines between Wt and T748A BMDMs (Fig. 6B).

Fig. 6.

Dual activation of Stat1 downstream of LPS-induced TLR4 endocytosis: autocrine/paracrine IFN tyrosine dependent and cell-intrinsic inflammatory threonine dependent. (A) Wt BMDMs were untreated or treated with MitMAB, then stimulated with LPS (100 ng/mL) for 4 h. Whole-cell lysates were harvested and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. (B) Wt and T748A BMDMs were untreated or treated with MitMAB and then stimulated with LPS (100 ng/mL). Total RNA was isolated at indicated time points. The indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). (C) Wt BMDMs were untreated or treated with tofacitinib and then stimulated with LPS (100 ng/mL) for 4 h. Whole-cell lysates were harvested and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. (D) IFNAR2^−/− THP-1 were stimulated with LPS (100 ng/mL). Whole-cell lysates were harvested at indicated time points and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. (E) Wt and T748A BMDMs were untreated or treated with tofacitinib and then stimulated with LPS (100 ng/mL). Total RNA was isolated at indicated time points. The indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). (F) Wt, T748A, and KO BMDMs were stimulated with LPS (100 ng/mL). Total RNA was isolated at indicated time points. The indicated transcripts were quantified by qRT-PCR. Data are presented as medians (n = 3 biological replicates). *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001 as measured by two-way ANOVA with post hoc Tukey’s test in (B and D) and one-way ANOVA with post hoc Tukey’s test in (F). ns, not significant.

Cooperative transcriptional activation between NF-κB and IFN-mediated phospho-tyrosine Stat1 has been reported to regulate the expression of downstream genes, such as Il6 (24). In addition, type I IFN activates the expression of Il10 in LPS-stimulated macrophages (25). In the light of these, we hypothesized that the expression of ISGs and Il10 requires primarily the IFN-mediated Tyr⁷⁰¹ phosphorylation, whereas the expression of Il6 and Il12b requires both Tyr⁷⁰¹ and Thr⁷⁴⁸ phosphorylation of Stat1 following LPS-induced TLR4 endocytosis. We first examined whether the Thr⁷⁴⁸ phosphorylation is IFN dependent. Notably, the pan-JAK inhibitor, tofacitinib, ablates the IFN-induced Tyr⁷⁰¹ phosphorylation, while the Thr⁷⁴⁸ phosphorylation of Stat1 was unabated following LPS stimulation of Wt BMDMs (Fig. 6C). Additionally, LPS stimulation of the IFNAR2-deficient human macrophage cell line, THP-1, induced the expression the Thr⁷⁴⁹ phosphorylation, while the IFN-mediated Tyr⁷⁰¹ phosphorylation of STAT1 was abated (Fig. 6D). These observations show that the Thr⁷⁴⁸ phosphorylation of Stat1 is IFN independent.

In accordance with our hypothesis, tofacitinib recapitulated the MitMAB effect and abolished the expression of Ifit1, Rsad2, and Il10 in LPS-stimulated Wt and T748A BMDMs (Fig. 6E). Strikingly, tofacitinib could not recapitulate the MitMAB effect on the expression of Il6 and Il12b, and LPS-stimulated T748A BMDMs exhibited attenuated expression of Il6 and Il12b compared to their Wt counterparts with or without tofacitinib treatment (Fig. 6E). Correspondingly, T748A mutation largely, but not fully, recapitulates the effect of Stat1 deficiency on inflammatory cytokines expression in LPS-stimulated macrophages as shown by the diminished expression of Il6 and Il12b in KO macrophages compared to their Wt and T748A counterparts (Fig. 6F). In sharp contrast to their KO counterparts, T748A macrophages show higher expression of Ifit1, Rsad2, and il10 following LPS stimulation (Fig. 6F). Our findings indicate that Stat1 displays a dual activation downstream of LPS-induced TLR4 endocytosis: an autocrine/paracrine IFN phospho-tyrosine and a cell-intrinsic inflammatory phospho-threonine, which is IFN independent.

IKKβ Phosphorylates STAT1 at Thr⁷⁴⁸ in Mice and Thr⁷⁴⁹ in Humans.

We next examined whether Ikkβ phosphorylates Stat1 at the Thr⁷⁴⁸ and whether it is conserved between the mouse and human. Of note, Ikkβ is required for STAT1 phosphorylation at the Thr⁷⁴⁸ in the mouse and the Thr⁷⁴⁹ in the human as shown by the exogenous expression of mouse and human IKKβ and Wt, T748A, or T749A mutant constructs of STAT1 in HEK293T cells (Fig. 7A). To determine whether IKKβ directly phosphorylates Stat1 at Thr^748/9, we performed biochemical cell-free kinase assays using purified GST-tagged IKKβ and Flag-tagged STAT1 Wt, T748A, or T749A mutant proteins from transfected HEK293T cells expressing the relevant proteins (SI Appendix, Fig. S12). Indeed, IKKβ directly phosphorylates Stat1 at Thr^748/9, and T748A and T749A mutation abolished the IKKβ-mediated phosphorylation of mouse and human STAT1, respectively (Fig. 7B).

Fig. 7.

IKKβ phosphorylates STAT1 at Thr⁷⁴⁸ in mice and Thr⁷⁴⁹ in humans. (A) HEK293T cells were transiently cotransfected with the indicated expression plasmids. Twenty-four hours later, whole-cell lysates were harvested and separated by SDS-PAGE. The indicated proteins were detected by western blotting analysis. The data are representative of three independent experiments. (B) Biochemical cell-free kinase assays of indicated purified IKKβ and STAT1 proteins, followed by separation by SDS-PAGE. The indicated proteins were detected by western blotting analysis. The data are representative of three independent experiments. (C) U3A cells were transiently cotransfected with the indicated mouse expression plasmids. Twenty-four hours later, cells were stimulated with IFNγ for 30 min, and whole-cell lysates were harvested and separated by SDS-PAGE. The indicated proteins were detected by western blotting analysis. The data are representative of three independent experiments. (D) U3A cells were transiently cotransfected with the indicated human expression plasmids. Twenty-four hours later, cells were stimulated with IFNγ for 30 min, and whole-cell lysates were harvested and separated by SDS-PAGE. The indicated proteins were detected by western blotting analysis. The data are representative of three independent experiments. (E) Wt BMDMs were untreated or treated with BMS-345541 and then stimulated with LPS (100 ng/mL). Whole-cell lysates were harvested at indicated time points and separated by SDS-PAGE. The indicated endogenous proteins were detected by immunoblot. Blots are representative of three independent experiments. (F) Schematic diagram of a dual activation of Stat1 downstream of the LPS-induced TLR4 endocytosis in macrophages.

The IFN-STAT1 signaling is highly conserved in vertebrates, and the STAT1-deficient human fibrosarcoma cells, U3A, have been utilized for STAT1 rescue and IFN response experiments using zebrafish, murine, and human STAT1 constructs (26). To examine the effect the IKKβ-mediated Thr^748/9 phosphorylation on the canonical Tyr⁷⁰¹ phosphorylation of STAT1, we coexpressed Wt, phosphotyrosine-deficient (Y701F and Y701A), or phosphothreonine-deficient (T748A and T749A) Stat1 constructs with IKKβ in U3A cells, followed by IFNγ stimulation. Notably, Ikkβ-mediated Thr⁷⁴⁸ phosphorylation restricts the IFN phospho-tyrosine activation of mouse Stat1 (Fig. 7C). Intriguingly, IKKβ-mediated Thr⁷⁴⁹ phosphorylation restricts the IFN phospho-tyrosine activation of human Stat1 (Fig. 7D). In contrast to the T748A mouse Stat1 mutant, we observed that T749A human Stat1 exhibited Tyr⁷⁰¹ phosphorylation without IFNγ stimulation (Fig. 7D), suggesting possible difference in the extent of the effect of the Thr^748/9 phosphorylation on the Tyr⁷⁰¹ phosphorylation between mouse and human STAT1 at basal and IFN-stimulated conditions. Of note, the IFN signaling was dispensable for the Thr⁷⁴⁸ and Thr⁷⁴⁹ phosphorylation of mouse and human Stat1, respectively (Fig. 7 C and D).

To corroborate these findings in macrophages, we examined the effect of chemical inhibition of Ikkβ on Stat1 phosphorylation following LPS stimulation of BMDMs from Wt mice. As expected, the Ikkβ inhibitor BMS-345541 diminished the phosphorylation of the p65 NF-κB and the degradation of IkBα (Fig. 7E). Remarkably, BMS-345541 treatment abolished the Thr⁷⁴⁸ phosphorylation and enhanced the Tyr⁷⁰¹ phosphorylation of Stat1 following LPS stimulation of Wt BMDMs (Fig. 7E). These observations suggest that LPS-induced TLR4 endocytosis activates a conserved cell-intrinsic IFN-independent Ikkβ-mediated Thr⁷⁴⁸ phosphorylation of STAT1, which restricts the basic IFN activation and promotes the innate inflammatory response (Fig. 7F).

Depletion of Macrophages Restores the Sensitivity of the T748A Mice to LPS-Induced Lethality.

To this end, our data show that the Ikkβ-mediated Thr⁷⁴⁸ phosphorylation of Stat1 promotes inflammatory responses while restricting IFN and anti-inflammatory responses in macrophages following LPS stimulation. To corroborate our findings in vivo, we performed clodronate liposome-mediated macrophage depletion (27) and assessed the effect on the survival of T748A mice following LPS-induced lethality (Fig. 8A). Injection of clodronate liposomes resulted in >90% depletion of macrophages (Fig. 8 B–D). Remarkably, depletion of macrophages restores the sensitivity of the T748A mice to LPS-induced lethality as shown by the survival of different treated groups: T748A+ control liposomes (91%, 10 out of 11 survived), T748A+ clodronate liposomes (36%, 4 out of 11 survived), and Wt+ control liposomes (20%, 2 out of 10 survived) (Fig. 8E). Taken together, our study identifies a cell-intrinsic Ikkβ-mediated Thr⁷⁴⁸ phosphorylation that promotes macrophage inflammatory response at the expense of IFN and anti-inflammatory responses following LPS-induced TLR4 endocytosis (Fig. 8F).

Fig. 8.

Depletion of macrophages restores the sensitivity of the T748A mice to LPS-induced lethality. (A) Experimental diagram of the clodronate-mediated macrophage depletion. (B) Flow cytometry analysis of macrophages from spleens of Wt and T748A littermates 24 h following control or clodronate liposomes injection. Data are representative of three independent experiments. (C) Analysis of macrophage percentage following clodronate-mediated depletion. Data are presented as medians (n = 3 biological replicates). (D) Analysis of macrophage number following clodronate-mediated depletion. Data are presented as medians (n = 3 biological replicates). (E) Survival of Wt and T748A littermates injected with control or clodronate liposomes and challenged with LPS (12.5 mg/kg) (n = 10 to 11 mice per genotype per group). (F) Schematic diagram of the role of the Thr⁷⁴⁸ phosphorylation of Stat1 in macrophage inflammatory response to LPS. **P < 0.01, ***P < 0.005, and ****P < 0.001 as measured by one-way ANOVA with post hoc Tukey’s test in (C and D) and the log-rank test (Mantel–Cox) in (E. ns, not significant.

Discussion

Protein phosphorylation is a widespread and powerful tool utilized by signaling cassettes to mount rapid modifications of cellular responses toward diverse environmental stimuli. Protein phosphorylation increases the organism’s proteome complexity far beyond the diversity conferred by the genome (28–31). Decades of extensive research on STAT1 have uncovered tyrosine and serine phosphorylation and other posttranslational modifications, yet they provide little insight into the potential STAT1 functionality beyond IFN and the phospho-tyrosine signaling (32). To fill this gap in our understanding, the term unphosphorylated STAT1 (U-STAT1) has been adopted to explain STAT1 functions beyond the JAK phospho-tyrosine signaling. However, the U-STAT1 expression itself relies on the IFN-JAK signaling and enhances the expression of ISGs (33–35). Thus, the U-STAT1 signaling is likely a continuation or an amplification mechanism for IFN signaling, leaving the long-standing biological conundrum of how STAT1 functions beyond IFN-JAK dogma to be answered.

Our complementary genetic and biochemical analysis provides clear evidence on the Thr⁷⁴⁸ phosphorylation of Stat1. Together with the canonical tyrosine and serine, our identification of the Thr⁷⁴⁸ completes the phosphorylation catalog of the TAD of STAT1 and more generally other STAT proteins. In addition to its ubiquitous antiviral defense function, STAT1 can also alter inflammation in a context-specific manner, yet the scope and mechanisms remain incompletely characterized (32). Our identification of the Thr⁷⁴⁸ phosphorylation as an inflammatory controller of Stat1 functionalities is supported by the protective phenotype of the T748A mice, which recapitulates the safeguard effect of the genetic ablation of Stat1 following LPS-induced sepsis. Remarkably, the Thr⁷⁴⁸ phosphorylation fulfills the two basic principles as an inflammatory signaling (36). First, modifying the core functionality of a given cell type. In macrophages, it promotes the inflammatory immune response following LPS stimulation. Second, restricting the basic signaling to modulate high-priority cell-specific core functionality. Two complementary experimental observations support this. Exogenously, disruption of the Thr phosphorylation enhances the IFN-induced Tyr phosphorylation of Stat1. Endogenously, T748A mutant macrophages exhibit enhanced IFN signaling while showing attenuated inflammatory response.

Mechanistically, we show that the Thr⁷⁴⁸ phosphorylation per se is IFN independent. Three independent lines of evidence support this. First, exogenously, IFN stimulation or disruption of the Tyr phosphorylation site in Stat1 is dispensable for the Thr phosphorylation. Second, inhibition of the JAK signaling in macrophages did not affect the Thr phosphorylation following LPS stimulation. Third, IFNAR2-deficient macrophages activate the Thr phosphorylation of Stat1 following LPS stimulation. Our findings suggest that Stat1 displays a phosphorylation-dependent signaling modularity in innate immune responses: autocrine/paracrine IFN phospho-tyrosine dependent and cell-intrinsic Ikkβ-mediated inflammatory phospho-threonine dependent. Our study suggests that the two modules are nonmutually exclusive, and combinatorial synergetic and antagonistic interactions between them dictate the downstream overall functionality of Stat1. Of note, the phospho-tyrosine module not only controls the expression of ISGs but also Il10. The required complementary activation between both modules to promote Il6 expression is notable and suggests critical roles of such pleiotropic cytokine in amplifying the inflammatory response that need to be tightly controlled. Notably, IL-6 and IL-10 are highly expressed in septic patients and are associated with sepsis severity (37, 38). Future research is needed to illustrate the detailed mechanisms governing Stat1-dependent regulation of Il-6 and Il-10 expression and whether these regulatory mechanisms are restricted to macrophages or represent a general feature for other cells.

LPS-induced TLR4 endocytosis utilizes a transfer machinery of CD14 and MD-2-TLR4 dimerization, which allows TLR4 signaling at picomolar concentrations of LPS (39). So far, the TLR4 endocytosis–induced signaling has been studied mainly in the context of IFN induction with little insight into its role in regulating the inflammatory signaling. Our data show that TLR4 endocytosis–mediated Thr⁷⁴⁸ phosphorylation of Stat1 promotes innate immune inflammatory response. Our observations provide a plausible mechanistic explanation for the TLR4 endosomal inflammatory signaling. The induction of the IFN by TLR4 endocytosis plays a dual role to modulate the macrophage immune response. On the one hand, the IFN signaling potentiates the expression of the inflammatory cytokines as shown by the attenuated expression of Il6 and Il12b following LPS stimulation in tofacitinib-treated macrophages. Il-6 and Il-12p40 play important roles in innate and adaptive immunity (40–42), which may provide another layer of the signal amplification through recruiting adaptive immune cells that potentiate the macrophage inflammatory response, and aid in clearing invading pathogens. On the other hand, IFN likely acts as a regulatory negative feedback loop that suppresses the inflammatory response through the induction of Il10 expression, a potent anti-inflammatory mediator. The combinatorial signaling output of the tyrosine and threonine phosphorylation therefore likely dictates the final output of Stat1 functionalities in shaping the innate immune response of macrophages toward pro- or anti-inflammatory. Supporting this, depletion of macrophages of T748A mice restored their sensitivity to LPS-induced lethality.

Our study identifies a connection between IKK and STAT pathways, two master signaling cassettes for innate immune responses. IKK proteins are master regulators for NF-κB (43, 44), and cooperative transcriptional activation between NF-κB and Stat1 has been reported to regulate the expression of downstream genes, most notably Il6 (24). In addition, IKK complex controls the stability of Il6 mRNA by phosphorylating regnase-1 that degrades Il6 mRNA (45). Further, IKK and STAT pathways are ubiquitous and regulate multiple cellular processes beyond immunity such as barrier integrity, metabolism, growth, and cancer (32, 46). Future research is required to investigate the spectrum of the physiological importance of this connection between these pathways in shaping downstream gene expression and cellular responses. It will be of particular interest to explore how interaction and counteraction between Thr and Tyr phosphorylation regulates cell-specific and context-dependent Stat1 functionality in immune and nonimmune biological processes.

Our observations show that the genetic manipulation of the Thr^748/9 phosphorylation alters Stat1 inflammatory functionality. In addition, our findings point to the steric accessibility of the Thr^748/9 phosphorylation site, which may provide the starting point for the development of precise biologics targeting the inflammatory function of STAT1 without disturbing its antiviral functions. Further, our observations suggest that the Thr^748/9 phosphorylation of Stat1 downstream of endocytosed TLR4 contributes to proinflammatory cytokine production, which may have implications for the use of vaccine adjuvants targeting this receptor.

In summary, our study provides a plausible explanation for the long-standing biological conundrum of how STAT1 functions beyond the IFN-JAK dogma. Our findings show that Stat1 displays a phosphorylation-dependent modular functionality in innate immune responses: IFN phospho-tyrosine dependent and inflammatory phospho-threonine dependent, which provides deeper insight into the combinatorial signaling logic of Stat1 in shaping cell-specific and context-dependent responses. Further, our study stirs future research on the exploration of other possible functional threonine phosphorylation sites in the TAD of other STAT proteins, which may provide deeper understanding of other STAT proteins’ functionalities beyond the canonical JAK signaling. Better understanding of the Thr⁷⁴⁸ phosphorylation of Stat1 may offer opportunities for developing better and more specific treatment modalities for various inflammatory diseases such as sepsis without the risk of adverse effects associated with targeting the JAK signaling or total Stat1 protein.

Materials and Methods

Mice.

Stat1T748A mutant mice on a C57BL/6 background were generated by using the CRISPR/Cas9 method. Stat1KO/Stat1T748A mice were generated by backcrossing Stat1KO on a C57BL/6 background with Stat1T748A mutant mice. The breeding followed the Mendelian inheritance pattern. Mice were kept and bred in pathogen-free conditions. Age- and sex-matched littermates were used for the experiments unless indicated otherwise in the text. All animal experiments were conducted in accordance with the guidelines of the Animal Care and Use Committee of Osaka University. Details for mice generation, genotyping, DNA sequencing, and generation of BMDMs, PBMDMs, and PCMs are provided in SI Appendix, Supporting Materials and Methods.

Cell Lines and Chemicals.

HEK293T, U3A, and THP-1 IFNAR2 knock-out cell lines were cultured as before (19). RAW 264.7 cells were purchased from ATCC (TIB-71) and cultured according to the manufacturer’s instructions. LPS, MitMAB, tofacitinib citrate, and BMS-345541 were used as before (19). Recombinant human IFN-γ was purchased from PeproTech (300-02).

LPS-Induced Septic Shock.

Ten- to twelve-wk-old male littermate mice were injected intraperitoneally with LPS (L2880, Sigma) in PBS. For LPS-induced lethality, mice were injected with 12.5 mg/kg body weight. Mice were assessed twice a day for survival. For experimental analysis, mice were injected with a sublethal dose of 7.5 mg/kg body weight of LPS for the indicated time points. At the indicated times after LPS injection, the animals were killed by CO2 overdose and rapidly dissected. Spleens were isolated and processed for single-cell preparation for the downstream analysis.

In Vivo Macrophage Depletion.

A volume of 100 μL of clodronate or control liposomes or liposomes (FormuMax Scientific, Clophosome-A, Combo Kit) was injected intraperitoneally 24 h prior to the LPS challenge. Macrophage depletion was confirmed using flow cytometry as explained in the text.

Generation of Phospho-Specific Rabbit Monoclonal Antibodies against Thr⁷⁴⁸ (Mouse) and Thr⁷⁴⁹ (Human) Stat1 (pThr Stat1 mAbs) Using Immunospot Array Assay on a Chip (ISAAC).

Twelve- to thirteen-wk-old New Zealand White rabbits were immunized subcutaneously with 500 μg of KLH conjugates of mouse phospho-Stat1 peptide in complete Freund’s adjuvant (Millipore), followed by isolation of peripheral blood leukocytes. Phospho-specific monoclonal antibodies were generated using a microwell array chip and the ISAAC method. Details are provided in SI Appendix, Supporting Materials and Methods.

Generation of Human MDMs.

This study followed the principles of the Declaration of Helsinki and was approved by the Institutional Review Board of Osaka University Hospital (permit nos. 16109 and 885 [Osaka University Critical Care Consortium Novel Omix Project; Occonomix Project]). Informed consent was obtained from the healthy volunteers for the collection of all blood samples MDMs were generated from isolated PBMCs as described before (19). Details are provided in SI Appendix, Supporting Materials and Methods.

Plasmid Construction and Mutagenesis.

Recombinant overexpressing vectors encoding human Flag-STAT1WT, Flag-STAT1T749A, and HA-IKKβ were used as previously described (19). Mouse 3xFlag-Stat1, 3xFlag-Ikkβ, and GST-Ikkβ constructs were prepared using the In-Fusion HD Cloning Kit (Clontech) according to the manufacturer’s protocol. Stat1 mutant constructs were generated using the KOD-Plus-Mutagenesis Kit (TOYOBO) according to the manufacturer’s protocol. Details are provided in SI Appendix, Supporting Materials and Methods.

Generation of RAW 264.7-Stat1 Cells.

RAW 264.7 cells were transduced with lentivirus encoding CRISPR/Cas9 targeting Stat1. Lentiviral-transduced cells were cultured in puromycin (Invivogen) containing medium at a final concentration of 5 μg/mL. Stat1 knock-out was confirmed by immunoblotting, the generated cells hereafter referred to as RAW 264.7 Stat1KO. Then, RAW 264.7 Stat1KO were reconstituted with lentivirus encoding Stat1 Wt or T748A and S747E/T748E mutants. Lentiviral-transduced cells were cultured in blasticidin (Invivogen) containing medium at a final concentration of 5 μg/mL. Reconstitution of Stat1 was confirmed using immunoblotting. Details are provided in SI Appendix, Supporting Materials and Methods.

Purification of IKKβ and STAT1 Proteins.

A total of 3-4 × 10⁶ HEK293T cells were plated on a 10-cm dish (Corning). Next day, cells were transfected with 10 μg of overexpression plasmids as indicated in the text. Forty-eight hours later, cells were lysed in RIPA Lysis and Extraction Buffer (Thermo Fisher). To purify Flag-STAT1 proteins, cell extracts were precleared and incubated with anti-Flag antibody (MBL, 1:100) overnight at 4 C. Then, Pierce Protein A/G Magnetic Agarose beads (Thermo Fisher) were added to the cell extracts for 2 h at 4 °C to pull-down Flag-Stat1 proteins. To purify GST-IKKβ proteins, cell extracts were precleared and incubated with Pierce Glutathione Magnetic Agarose beads (Thermo Fisher) for 4 h at 4 °C. Tubes were placed on magnet, and beads were washed 5 times with RIPA Lysis and Extraction Buffer. GST-IKKβ proteins were eluted with 20 mM reduced glutathione in tris-buffered saline, and Flag-STAT1 proteins were eluted with Pierce IgG Elution Buffer (Thermo Fisher).

Biochemical Cell-Free In Vitro Kinase Assay.

Around 1 μg of purified mouse or human IKKβ and STAT1 proteins as indicated in the text was incubated at 37 °C in 50 μL buffer containing 100 mM ATP (ThermoFisher), 50 mM Tris-HCl pH 7.6 (ThermoFisher), 50 mM NaCl (Nacalai Tesque), and 10 mM CaCl2 (Nacalai Tesque) for 2 h with gentle shaking (300 rpm) in Eppendorf ThermoMixer C. Then, SDS-PAGE loading dye was added to the reaction and incubated at 96 °C for 5 min, followed by gel run, electrophoresis, transfer, and immunoblotting.

Immunoblotting.

Whole-cell lysates were prepared using RIPA Lysis and Extraction Buffer (Thermo Fisher). Protein concentration was quantified using the Pierce Detergent Compatible Bradford Assay Kit (Thermo Fisher). Lysates were boiled in SDS sample buffer (Nacalai Tesque) for 5 min at 96 °C, and 10 to 20 μg of total protein was loaded and resolved by 5 to 20% SDS-PAGE gels (Nacalai Tesque). Proteins were then transferred onto 0.45 µM polyvinylidene difluoride membrane (GE Healthcare) by the wet transfer system using Criterion Blotter (BioRad), washed, and incubated with primary antibodies. Then, the membrane was washed and incubated with anti-rabbit IgG horseradish peroxidase conjugated, blots were developed, and images were taken by an Image Quant LAS800 (GE Healthcare). Details for immunoblotting and antibodies used are provided in SI Appendix, Supporting Materials and Methods and Table S1.

Mass Cytometry.

Wt and T748A littermates were untreated or injected a sublethal dose of 7.5 mg/kg body weight of LPS for 24 h. The animals were killed by CO2 overdose and rapidly dissected. Spleens were isolated and processed for single-cell preparation, and mass cytometry analysis was performed. Details for mass cytometry and antibodies used are provided in SI Appendix, Supporting Materials and Methods and Tables S2 and S3.

Flow Cytometry.

Single-cell suspension was blocked with anti-mouse CD16/32 antibody (BioLegend) in FACS buffer (2 mM EDTA and 2% FCS in PBS). Then, cells were incubated in a cocktail of surface antibodies prepared in FACS buffer for 30 min at 4 °C as indicated in the text. Samples were acquired with LSR Fortessa (BD Biosciences) for analysis or Aria II (BD Biosciences) for cell sorting. Data analysis was performed using FlowJo v.10. The surface staining antibodies used in flow cytometry are listed in SI Appendix, Table S4.

RNA sequencing analysis.

A total of 2 × 10⁵ BMDMs and 1 × 10⁶ PCMs with or without indicated treatment were lysed in 0.5 mL of TRIzol Reagent (Thermo Fisher). RNA-seq analysis was performed by the Genome Information Research Center (GIRC, Osaka University). Details are provided in SI Appendix, Supporting Materials and Methods.

RT-qPCR.

A total of 5 to 10 × 10⁵ indicated cells with or without indicated treatment were lysed in 0.5 mL of TRIzol Reagent (Thermo Fisher). Total RNA was extracted using Direct-zol RNA MicroPrep (Zymo Research). Then, cDNA was prepared using SMART MMLV RT (Takara), Advantage UltraPure dNTP (Takara), and Oligo d(T)23 VN (New England Biolabs) according to the manufacturer’s instructions. Real-time qPCR was performed on the generated cDNA (5 to 10 ng/rxn) using predesigned TaqMan primers (Thermo Fisher) listed in SI Appendix, Table S5. The RT-qPCR was performed using TaqMan Fast Advanced Master Mix (Thermo Fisher) in a 96-well plate using the QuantStudio 3 Real-Time PCR System (Thermo Fisher). Relative expression was calculated from the Ct values.

Statistical Analysis and Illustrations.

Survival was analyzed using Kaplan–Meier curves and the log-rank (Mantel–Cox) test. P values were calculated using one-way or two-way ANOVA with post hoc Tukey’s test for multiple comparison and unpaired two-tailed Student’s t test. P value <0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism 10 software. Immunoblot band intensity was measured by ImageJ (v1.54i). Figures were prepared by Adobe Illustrator, and schematic diagrams were drawn using BioRender.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All data needed to evaluate the conclusions in the paper are present in the paper and SI Appendix. Mass cytometry data have been uploaded to the Zenodo repository (https://doi.org/10.5281/zenodo.10792053) (47). RNA-seq data have been deposited to the NCBI Gene Expression Omnibus database (GSE261824) (48).