Background

Age-related inflammation, or inflammaging, is a common phenomenon characterized by increased levels of pro-inflammatory factors in the blood [1]. Contrary to acute inflammation and its role in immunity, chronic inflammation is highly detrimental to general health. It is a central risk factor for numerous age-related morbidities such as cardiovascular disease, diabetes, cancer, and kidney disease [1]. Fundamentally, chronic inflammation is hallmarked by a complex disbalance between pro-and anti-inflammatory responses that, in the case of inflammaging, are predominantly controlled by macrophages [2, 3]. Besides their role in apoptotic cell clearance and innate immunity, macrophages are the most potent producers of TNFα, a pro-inflammatory cytokine involved in numerous inflammatory and auto-immune disorders [4]. Systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) are such diseases and have been found to affect females disproportionally [5]. Sex-based associations remain undetermined in inflammaging, but emerging reports suggest such differences [5, 6]. Thus far, the most important genetic factors for inflammaging are polymorphism in IL-6 [7], its receptor IL-6R [8], and IL-1RN (IL1Ra antagonist gene) [9].

Thousand-and-one (TAO) kinases belong to a subfamily of Ste20-like proteins that can act as MAP4Ks to phosphorylate MAP3Ks [10,11,12,13]. The TAO kinase family members have been shown to share kinase domain resemblance and substrate specificity with conventional MAP4K enzymes [14, 15]. The function of those enzymes is involved in immune and inflammatory disorders such as RA and SLE [16]. To our knowledge, it remains unknown whether TAOKs are involved in inflammatory disorders and, more broadly, if STE20-kinases are implicated with inflammaging. Among the three members of TAOs kinases, TAOK3 uniquely shows increased expression in the hematopoietic compartment, especially in macrophages [34].

Fig. 4
figure 4

Taok3 deletion or inhibition increases differentiation of macrophages and their production of pro-inflammatory response factors. a Schematic representation of the method employed to differentiate and polarize bone marrow derive- macrophages (BMDMs). Representative flow cytometry plot showing the purity of BMDM cultures, as assessed by the frequency of CD11 + , F4/80 + macrophages. b Heatmaps representing the LOG10 (Fold change) relative to Taok3.+/+ unpolarized M0 using the mean fluorescence intensity by flow cytometry. Statistical comparisons are made between genotypes between the same polarization regimens. c Quantification of the TNFα levels in BDMSs from b) using ELISA. d Gating strategy used to discern between GM-SCF grown BMDCs (CD11c +) and M-CSF differentiated BMDMs (F4/80 +). e Flow cytometric analysis of the mean expression (mean fluorescence intensity) of surface molecules and transcription factors in BMDCs and BMDMs. f Bar plot of the frequency of differentiated macrophages, depending on the phase at which the cells were treated with the inhibitor. Differentiation (Day 0–6) → Polarization (Day 6–7). g Heatmaps representing the LOG10 (Fold change) relative to DMSO → DMSO unpolarized M0 using the mean fluorescence intensity by flow cytometry. Statistical comparisons are made between genotypes between the same polarization regimens against the respective DMSO → DMSO group. h Quantification of the TNFα levels in BDMSs from g) using ELISA. Data representative of at least three independent experiments. Chart error bars represent mean ± SEM. Statistical analysis: a-h Multiple unpaired, bilateral T-tests. ns, non-significant. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001

Full size image

Given that bone marrow progenitor cells were shown to be skewed towards myelopoiesis, it remained possible that alternative transcriptional programs were already found in precursors of BMDMs before differentiation with M-CSF. We thus used a small molecule inhibitor of TAO- family kinases, compound-43 [35], to validate our previous findings and to make two demonstrations: 1. That the catalytic/kinase activity of Taok3 is important in controlling TNFα production and M1 pro-inflammatory responses and 2. that those changes are intrinsic to differentiating macrophages and not a result of differential gene expression in progenitors. Using 5 µM of compound-43, we tested whether the phenotype relied on inhibition during the differentiation phase (Day 0–6) or the polarization phase (Day 6–7). We used four different treatment conditions: DMSO → DMSO (D → D), DMSO → TAOKi (D → T), TAOKi → DMSO (T → D), and TAOKi → TAOKi (T → T) that represent the phase at which the drug/vehicle was added (Differentiation → Polarization). As expected, only the BMDMs treated during their differentiation phase (TAOKi → DMSO) had an increased frequency of BMDMs compared to the DMSO → DMSO control group (Fig. 4f). We next polarized those BMDMs to M1 macrophages using the same method as described previously to validate previous findings in a macrophage-intrinsic model. We observed that administration of TAOK inhibitor, regardless of condition, consistently increased CD11b expression (Fig. 4g). This increase was also accompanied by an increase in F4/80 in TAOKi → DMSO M0 and M1 macrophages.

Most importantly, inhibition of Taoks resulted in a significant increase in iNOS expression, which was more pronounced in T → D and T → T conditions (Fig. 4g). iNOS is a critical enzyme in the macrophage inflammatory response and is responsible for inducing bacterial killing via nitric oxide (NO) synthesis [33, 36]. We also quantified TNFα levels in the supernatants of those conditions using ELISA. We found that only M1 macrophages that had received TAOKi during their differentiation phase (T → D and T → T) had increased concentrations of TNFα (Fig. 4h). This result not only validates the previous findings in a cell-intrinsic manner but also provides evidence for the importance of the kinase activity of Taok3 during the differentiation of macrophages. Altogether, these findings demonstrate that Taok3 negatively modulates the differentiation of bone marrow-derived macrophages and their secretion of pro-inflammatory cytokines IL-6 and TNFα. In addition to this function, in vitro and in vivo results suggest that Taok3 has a role in limiting sex and age-related inflammation through its role in leukocytes.

Discussion

With the increase in life expectancy worldwide, the number of aged individuals is set to rise to historic highs in the following decades. With increasing age comes inflammaging, an incompletely understood condition that increases the risk of chronic diseases. Here, we uncovered that an unexpected gene, TAOK3, plays a role in limiting inflammaging by negatively modulating the polarization of macrophages towards pro-inflammatory responses. We identified that the kinase activity of the enzyme is necessary for limiting macrophage differentiation in-vitro. Our study supports the mounting evidence for the implication of innate immune cells and their receptors in chronic inflammation [37]. Other essential macrophage functions we did not cover include efferocytosis and debris clearance within the organism. The importance of those functions with inflammaging has already been demonstrated [2, 38]. Our findings support that genetic factors can induce susceptibility to pro-inflammatory cytokine and chemokine production in a sex-dependent manner. Such associations are important for understanding other diseases, such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), given that those diseases disproportionately affect women. These diseases, along with inflammatory skin conditions, strongly implicate the pro-inflammatory functions of TNF-α [39]. Our findings involving TAOK3, TNF-α, and ulcerative dermatitis in mice, further support this implication. TAOK3 single nucleotide polymorphisms (SNPs) have also been linked with an increased risk of SLE [40]. The rs428073 SNP encodes a C-T missense variant causing the S47N substitution. The functional role of this mutation remains unknown, but the variant is near the predicted ATP binding site. Thus, it may interfere with the catalytic activity of the kinase. Our study demonstrates that the Ser/Thr kinase activity of TAOK3 is important in limiting pro-inflammatory responses in macrophages. Several groups have shown that inhibiting the enzyme induces anti-cancer responses [35, 41, 42]. A demonstrated relationship between NF-κB mediated signaling and TAOK3 could also imply the presence of inflammatory responses in cancer cells [43]. TAOK3 has also been implicated in skeletal mineralization by modulating the differentiation of osteoblasts [44]. This brings essential implications, given the established link between age-related inflammation and bone disorders, such as osteoporosis [45]. TAOK3, and possibly other STE-20 kinases, finds itself at the center of an association between inflammation, sex-specific associations, and innate immunity. We are limited in our interpretation of our data, given that we do not possess a tissue-specific knockout model. Such a model would be paramount in dissecting other macrophage-specific functions, such as efferocytosis and phagocytosis. We are also limited in the statistical power of our study. We also uncovered an unexpected role for TAOK3 in hematopoiesis. It remains unclear, however, at which precise stage does Taok3 participates in hematopoiesis. The cell-autonomous effects of the kinase in this setting also remain unknown. Many different molecular mechanisms could explain the phenotypes observed here. TAOK3 has already been linked to Notch signaling in other immune cell types [20, 21]. However, it seems unlikely that this mode of action is central here, given the reported positive role of TAOK3 in Notch signaling and the importance of Notch signaling in the M1 polarization [46, 47]. Overall, our findings align with previous reports stating that other STE-20 enzymes negatively modulate macrophage function and pro-inflammatory responses [16, 48, 49]. Thus, like the other STE-20 kinases, TAOK3 could be involved in innate receptor signaling, such as TLR receptors, through interaction with TRAF proteins [16]. TAOK3 has been previously demonstrated to interact with TRAF2, although it was not in the context of TLR signaling [50]. TAOK1, which is 73.56% identical in amino acid sequence to TAOK3, was shown to interact with TRAF6 during IL-17-mediated inflammation [48]. These reports, combined with recent evidence showing that some TLRs are involved in sex-dependent inflammatory and auto-immune disorders, provide further directions for studying the mechanism by which TAO kinases operate [37, 51, 52].

Methods

Mice

Whole-body Taok3 KO mice were purchased from Jackson Laboratories MMRRC stock #43,790. These mice were generated by the KOMP team using a CRISPR-generated KO mutant of the Taok3 kinase by deleting exon 6, leading to a frameshift mutation causing an early stop codon [53]. This frameshift nullifies protein expression but not mRNA transcription. Taok3−/− mice were bred with background-compatible C57/Bl6 mating partners. Control animals for Taok3−/− mice were Taok3+/+ littermates from the heterozygous breedings mentioned above. Animals were housed and bred in specific-pathogen-free facilities at the Comparative Medicine and Animal Resources Centre (CMARC) at McGill University in accordance with the Canadian Council on Animal Care Guidelines.

Histology

Tissues were fixed in 10% Buffered formalin. From fixed tissue, representative sections were trimmed and processed for paraffin embedding. Paraffin blocks are cut at a width of approximately 4 µm onto glass slides and stained with hematoxylin and eosin.

Flow cytometry

Staining and washing were performed in round bottom polystyrene 5 mL tubes (Falcon, 14–959-6). Unless otherwise stated, intracellular staining was performed with the BD biosciences Fixation/Permeabilization solution (BD Biosciences, 554,714) according to the manufacturer’s protocol. Beferdin A (BD Biosciences, 555,029) solution was used as the protein transport inhibitor during intracellular stainings. A 1:1000 dilution was incubated directly in cell culture plates for 4 h before stainings. Data were acquired using the Cytek Aurora (19-color panel) and the BD 5-laser LSR FORTESSA. Data were analyzed using FlowJo. V10 software.

Antibodies: Bone marrow HSCs: Anti-CD3-Biotin(Biolegend,1:400),TER119-Biotin(Biolegend,1:400),CD11b-Biotin (Biolegend,1:400), CD45R-Biotin(Biolegend,1:400), GR-1-Biotin (Biolegend,1:400), C-kit-APC (Biolegend,1:400), Sca-1-PECY7 (Biolegend,1:400), CD48-BV605 (Biolegend,1:300), CD150-BV450 (Biolegend,1:300), FLT3-PE (Biolegend,1:300), CD34-FITC (Biolegend,1:300), D16/32-APCCY7 (Biolegend, 1:300), IL7RA-BV785 (Biolegend,1:300). Spleens: CD19-AF647 (Biolegend, 1:1600), B220-AF700 (Biolegend, 1:800), IgM -BV605 (Biolegend, 1:200), IgD-PerCPCy5.5 (Biolegend, 1:400), CD3-APC/Fire810 (Biolegend, 1:150), CD4-PE/Dazzle 594 (Biolegend, 1:800), CD8-Pacific Blue (Biolegend, 1:600), CD44-BV421 (Biolegend, 1:400), CD62L-BV786 (BD, 1:1200), NK1.1-FITC(Biolegend, 1:300), TCRb-APC (Biolegend, 1:200), Ter119-BV650(Biolegend, 1:400), CD71-PE(Biolegend, 1:800), Ly6G-PE/Cy7 (Biolegend, 1:400), Ly6C-APC/Cy7(Biolegend, 1:400), CD11b-SparkYG593 (Biolegend, 1:800), CD11c-BV510(Biolegend, 1:150), F4/80-PE/Cy5 (Biolegend, 1:600), MHCII Spark Blue 550(Biolegend, 1:800), F4/80-PECY5 (Biolegend, 1:600), CD11c -PE (Biolegend, 1:600), GR1 (bv650) biotin (Biolegend, 1:500), CD126-PECY7(Biolegend, 1:400), CD86 (BV786)(Biolegend, 1:300), PD-L1-APC(Biolegend,1:1:200),CD44-PECY7 (Biolegend, 1:1:300),MHC1-PERCPCY5.5(Biolegend,1:1:200), CCR7-BV421 (Biolegend, 1:1:300),CD11b-FITC (Biolegend, 1:500), IL4-Rα-PECY7 (Biolegend, 1:300),IL-6-PE(Biolegend, 1:100), IRF4-BV421 (Biolegend, 1:100), iNOS-APC (Biolegend, 1:300).

Generation of BMDMs and BMDCs

Bone marrow from both femurs and tibiae was harvested in Dulbecco's High Glucose Modified Eagles Medium (DMEM) (Cytiva, SH30022FS) supplemented with 10% heat-inactivated FBS (Life technologies, 12,483,020). Marrows were filtered through a 70 μm Nylon cell strainer (Corning, 07–201-431) to remove solid fragments. The filtrate was centrifuged at 300 × g for 5 min at 4 °C. Cells were subsequently seeded in 2 ml DMEM supplemented with 10% hi-FBS, 1% non-essential amino acids (Gibco, 11,140,050), 1 mM sodium pyruvate (Gibco, 11,360,070), 2 mM L-Glutamine (Glutamax) (Gibco, 35,050,061), 100U/ml penicillin–streptomycin (Gibco, 15,140,122) in 6-well plates at a density of 1 × 105 cells/mL. Recombinant murine M-CSF (Preprotech, 315–02) was added directly to the wells on days 0, 3, and 6 at a final concentration of 30 ng/mL. The same method was employed to generate BMDCs, but with 20 ng/mL of recombinant murine GM-CSF (Preprotech, 315–03) and 5 ng/mL of recombinant murine IL-4 (Preprotech, 214–14). BMDCs were in their immature “unactivated” state during that study. Cells were harvested by gentle scra** and washing with PBS. Washed cells were then used for downstream applications.

For the polarization, 100 ng/mL of LPS (LPS from E. coli O111:B4, InVivogen, tlrl-3pelps) and 20 ng/mL of recombinant murine IFNγ (Preprotech, 315–05) were used to induce M1 macrophages for 24 h. For M2 macrophages, 20 ng/mL of recombinant murine IL-4 (Preprotech, 214–14) was added to the culture wells for 24 h.

The TAO kinases inhibitor, compound 43, was diluted in DMSO (water insoluble) and purchased from Cayman chemicals (Ref: 25632).

Enzyme-linked immunosorbent assays (ELISA) and cytokine/chemokine multiplexing

Supernatant from BMDM cultures was collected on day 7 ( 24 h post-polarization) and kept at -80C until downstream analysis. Mouse IL-6 (Biolegend, 431,301) and TNFα (Biolegend, 430,904) were used as per the manufacturer's protocol.

For cytokine multiplexing, serum was collected through cardiac puncture using 1 mL syringes and 25G × 1 needles (BD biosciences). Blood was transferred to S-Monovette® Serum Gel, 1.1 mL collection vessels (Sarstedt, 06.1667.001) and was left to cloth for 30 min at room temperature. Vessels were then centrifuged at 2000 × g for 10 min. Supernatants (serum) were collected and stored in new tubes at -80C until downstream analysis. For cytokine multiplexing, serum was diluted onefold before being sent to EVE Technologies, Calgary, Alberta, Canada, for cytokine multiplexing with a Bio-Plex 200.

Statistical analysis

GraphPad Prism software v9.0a (GraphPad Software) was used to prepare graphs and perform statistical analysis. We used a two-tailed student T-test for pairwise comparisons to determine statistical differences between distributions. ns, not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.000.