Tissue Treg Secretomes and Transcription Factors Shared With Stem Cells Contribute to a Treg Niche to Maintain Treg-Ness With 80% Innate Immune Pathways, and Functions of Immunosuppression and Tissue Repair

Abstract

We used functional -omics angles and examined transcriptomic heterogeneity in CD4⁺Foxp3⁺ regulatory T cells (Treg) from spleen (s-Treg), lymph nodes (LN-Treg), intestine (int-Treg), and visceral adipose tissue (VAT-Treg), and made significant findings: 1) Five new shared Treg genes including NIBAN, TNFRSF1b, DUSP4,VAV2, and KLRG1, and 68 new signatures are identified. Among 27 signaling pathways shared in four tissue Treg, 22 pathways are innate immune pathways (81.5%); 2) s-Treg, LN-Treg, int-Treg, and VAT-Treg have zero, 49, 45, and 116 upregulated pathways, respectively; 3) 12, 7, and 15 out of 373 CD markers are identified as specific for LN-Treg, int-Treg, and VAT-Treg, respectively, which may initiate innate immune signaling; 4) 7, 49, 44, and 79 increased cytokines out of 1176 cytokines are identified for four Treg, respectively, suggesting that Treg have much more secretory proteins/cytokines than IL-10, TGF-β, and IL-35; 5) LN-Treg, int-Treg, and VAT-Treg have 13 additional secretory functions more than s-Treg, found by analyzing 1,706 secretomic genes; 6) 2, 20, 25, and 43 increased transcription factors (TFs) out of 1,496 TFs are identified four Treg, respectively; 7) LN-Treg and int-Treg have increased pyroptosis regulators but VAT-Treg have increased apoptosis regulators; 8) 1, 15, 19, and 31 increased kinases out of 661 kinome are identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively; 9) comparing with that of s-Treg, LN-Treg, int-Treg, and VAT-Treg increase activated cluster (clusters 1–3) markers; and decrease resting cluster (clusters 4–6) markers; and 10) Treg promote tissue repair by sharing secretomes and TFs AHR, ETV5, EGR1, and KLF4 with stem cells, which partially promote upregulation of all the groups of Treg genes. These results suggest that stem cell-shared master genes make tissue Treg as the first T cell type using a Treg niche to maintain their Treg-ness with 80% innate immune pathways, and triple functions of immunosuppression, tissue repair, and homeostasis maintenance. Our results have provided novel insights on the roles of innate immune pathways on Treg heterogeneity and new therapeutic targets for immunosuppression, tissue repair, cardiovascular diseases, chronic kidney disease, autoimmune diseases, transplantation, and cancers.

Introduction

Our and others’ recent reports showed that cardiovascular (CVD) stressors and risk factors such as hyperlipidemia (1, 2), hyperglycemia (3, 4), hyperhomocysteinemia (5, 6), and chronic kidney disease (7–11), promote atherosclerosis and vascular inflammation via several mechanisms. These mechanisms include endothelial cell (EC) activation (1, 12–15) and injury (16); caspase-1/inflammasome activation (7, 9, 16, 17), mitochondrial reactive oxygen species (ROS) (2, 18, 19); Ly6C^high mouse monocyte and CD40⁺ human monocyte differentiation (4, 6, 20, 21); impaired vascular repairability of bone marrow-derived progenitor cells (17, 22); downregulated histone modification enzymes (23) and increased histone 3 lysine 14 acetylation (18), and increased expressions of trained immunity pathway enzymes (24, 25). In addition, we also reported that decreased/transdifferentiated CD4⁺Foxp3⁺ regulatory T cells (26–29) (Treg) facilitate vascular inflammation.

Current understanding on T helper cell (Th) differentiation is that in response to stimulation by several different inducing cytokines such as interferon-γ (IFN-γ), interleukin-12 (IL-12), and IL-4, and also anatomical locations (30), naïve CD4+ T cells can be differentiated/polarized into at least nine terminally differentiated Th cell subsets. These subsets include Th1, Th2, Th9, follicular T (Tfh) (30), Tfh-13 (31), Th17, Treg, Th22 (28, 32), Th25 (33), CD4⁺ cytotoxic T cells (CD4+ CTL) (34), tissue-resident memory T cells (Trm), circulating effector memory T cells (Tem), central memory T cells (Tcm) (35), and CD28^null T cells (36), suggesting that antigen epitopes-independent innate immune inducing cytokine environments play critical roles for naïve Th0 polarization/differentiation into Treg and other Th subsets. Foxp3 is the major transcription factor (TF), and co-expression of lineage-specifying transcription factors alters the potential function and flexibility of subsets of CD4⁺ T cell; this, in turn, favors the autoimmune pathology (37). Tregs are specialized in the suppression of immuno-pathological reactions in the host immune system against antigens and dangers (28). In addition to inhibition of adaptive immune response, Tregs also play a critical role in controlling various innate immune responses involved in cancers (38), inflammatory diseases including cardiovascular diseases and atherosclerosis (36, 39). Additionally, Tregs play a highly broad spectrum of versatile anti-pathophysiological roles. For example, Tregs facilitate blood flow recovery after ischemia (40), control adipose tissue inflammation, promote muscle repair (41) and maintain tissue/organ homeostasis (42). Tregs’ roles in maintaining self-tolerance and prevention of autoimmune responses and chronic inflammation are mediated by various mechanisms including: a) Treg killing of target cells (38); b) modulation of target cells via cell-cell contact; c) inhibition of target cells by exosome-carried microRNAs (28); and d) secretion of anti-inflammatory/immunosuppressive cytokines (13) including interleukin-10 (IL-10), IL-35 (1, 43, 44), and transforming growth factor-β (TGF-β). Therefore, cellular therapies using regulatory T (T_reg) cells are currently undergoing clinical trials for the treatment of autoimmune diseases, transplant rejection and graft-versus-host disease (45).

We previously reported that Treg cell death pathways (26–28, 46–52), Treg generated IL-35 (1, 43, 44), and epigenetic pathways (23, 53) may be novel therapeutic targets for maintaining Treg survival, preventing immunosuppressive Tregs from becoming pathological Tregs (28), plastic Tregs and even antigen-presenting Tregs (29), and suppressing inflammation (39). Recently, we proposed a novel concept, which suggests that pathological conditions/environments, via antigen epitopes-dependent or independent cellular interactions, re-shape physiological Tregs into pathological Tregs that have weakened immuno-suppressive functions and increased plasticity (28). The following supporting evidence published by other investigators validate our proposed model: First, Th1-like Treg phenotype (54, 55), and pro-inflammatory IL-17A cytokine secreting Treg (56); Second, immunosuppression-compromised Treg after myocardial infarction (57); Third, four different types of “lymphoma Treg” (38); Fourth, self-reactive T cells, termed anti-Treg (58); and Fifth, FOXO3-expressed in tolerogenic dendritic cells (DCs) in modulating Treg and activating anti-Treg (59). It is accepted that Tregs undergo phenotypic and functional plastic changes into other Th subsets under pathological conditions (32, 60), which are modulated by co-inhibitory receptors (61).

One of Tregs’ functional modes is the secretion of anti-inflammatory/immunosuppressive cytokines (13), including IL-10, IL-35 (1, 43, 44), and TGF-β. However, two important questions remain how many cytokines and secretomes are generated in Treg in various tissues and pathological conditions; and whether those secretomes create a stem cell niche-like microenvironment for Treg maintenance. The secretome, defined as a portion of total proteins secreted by cells to the extracellular space, secures a proper micro-environmental niche, thus maintaining tissue homeostasis (62, 63). Secreted molecules are key mediators in cell-cell interactions, via autocrine, and paracrine manners, and influence the cross-talk with the surrounding tissues in addition to their endocrine functions in long-distance by hormones, growth factors, cytokines, adipokines, myokines, cardiokines (64), and chemokines (65). There is strong evidence supporting that crucial cellular functions such as proliferation, differentiation, communication, and migration are regulated strictly by the cell secretome (66). The major differences between our current study and previous reports on the roles of cytokines and chemokines in Treg are that secretome analyses provide a panoramic view on all the secreted genes in Treg, as opposed to focusing on only one or a few cytokines/chemokines (10).

Tregs have functions in various tissue repair (67) including promoting muscle repair (68) and repair after cardiac injury (69), controlling neutrophil recruitment (70), facilitating skin epithelial stem cell differentiation (71) and wound healing (72), enhancing satellite cell expansion in muscle but blocking satellite cell differentiation (73), facilitating lung resolution (74), promoting lung epithelial cell proliferation (75) and lung injury repair via generating the growth factor amphiregulin (68), promoting myelin regeneration in central nerve system (76), and protecting kidney injury (77). However, molecular mechanisms underlying Treg promotion of tissue repair remained poorly defined.

In order to broaden our understanding of transcriptomic heterogeneity in Treg from s-Treg, LN-Treg, int-Treg, and VAT-Treg, we hypothesized that tissue Treg heterogeneity could be characterized by examining signature genes, upregulated signal pathways, clusters of differentiation (CD) markers, cytokines and secretomes, TFs, cell death regulatomes, activation and resting status, and Treg similarity of secretomes and TFs to that of stem cells. We conducted comprehensive data analyses on numerous microarray datasets from the NIH-NCBI-GEO database (https://www.ncbi.nlm.nih.gov/gds/). We made the following findings: 1) Five new core Treg genes and 68 new signature genes are identified; 2) s-Treg, LN-Treg, int-Treg, and VAT-Treg have zero, 49, 45, and 116 upregulated pathways, respectively; 3) LN-Treg, int-Treg, and VAT-Treg have 12, 7, and 15 specific CD markers out of 373 CD markers, respectively; 4) analyses of 1,176 cytokines and 1,706 secretomic genes suggest that LN-Treg, int-Treg, and VAT-Treg have 13 functions more than s-Treg; 5) 2, 20, 25, and 43 increased transcription factors (TF) out of 1,496 TFs are identified four Treg, respectively; 6) LN-Treg and int-Treg have increased pyroptosis regulators but VAT-Treg have increased apoptosis regulators, judging by 305 regulatomes in 13 cell death forms; 7) IL-2 receptor β (IL2Rβ) plays an essential role in promoting all tissue Treg shared functions and specific functions; 8) LN-Treg, int-Treg, and VAT-Treg increase activated cluster (clusters 1–3) markers; and decrease resting cluster (clusters 4–6) markers; and 9) Four Treg promote tissue repair by generating secretomes similar to that of stem cells; and sharing TFs aryl hydrocarbon receptor (AHR), ETS variant transcription factor 5 (ETV5), early growth response 1 (EGR1), and Kruppel like factor 4 (KLF4) with stem cells. Our results have provided novel insights on tissue Treg heterogeneity and new therapeutic targets for immunosuppression, tissue repair, cardiovascular diseases, chronic kidney disease, autoimmune diseases, transplantation, and cancers.

Materials and Methods

Expression Profiles of Splenic Regulatory T Cells, Lymph Nodes Regulatory T Cells, Intestine (Lamina Propria) Regulatory T Cells, and Visceral Adipose Tissue Regulatory T Cells

Microarray datasets were collected from National Institutes of Health (NIH)-National Center for Biotechnology Information (NCBI)-Gene Expression Omnibus (GEO) databases (https://www.ncbi.nlm.nih.gov/gds/) and analyzed with an online software GEO2R (https://www.ncbi.nlm.nih.gov/geo/geo2r/).

Statistical Analysis of Microarray Data

We applied a statistical method similar to our previously reported meta-analysis (10, 24, 78). We designed a robust housekeeping gene list (Supplementary Table 1 of housekeeping genes) with help from Eisenberg and Levanon’s (79) excellent work, including ACTB, GAPDH, PGK1, PPIA, B2M, YWHAZ, SDHA, HMBS, and TBP. Briefly, the mean log fold change (LogFC) of housekeeping genes between treatment and control groups vary from −1.27 to 1.28. The target genes with expression changes more than 2-folds were defined as the upregulated genes, while genes with their expression decreased more than 2-folds were defined as downregulated genes |logFC|>1).

Ingenuity Pathway Analysis

We utilized Ingenuity Pathway Analysis (IPA, Qiagen, https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis/) to characterize clinical relevance and molecular and cellular functions related to the identified genes in our microarray analysis. Differentially expressed genes were identified and uploaded into IPA for analysis. The core and pathways analysis was used to identify molecular and cellular pathways, as we have previously reported (10, 78, 80).

Results

1. Transcriptomic differences between Treg and Tconv are small, ranging between 0.29 and 4.28% (1–14.8 folds) among s-Treg, LN-Treg, int-Treg, and VAT-Treg; five new Treg core genes including NIBAN, TNFRSF1b, DUSP4, Vav2, and Klrg1 have been identified; and non-lymphoid tissue niches and lymph node niches contribute more than splenic niches to Treg transcriptomic differences from that of CD4⁺Foxp3⁻ conventional T cell (Tconv).

Our recent report showed that Treg-specific transcription factor (TF) FOXP3 was expressed in trachea, thymus, spleen, mammary gland, lymph node, lung, eye, and blood (29). Additionally, various Treg populations in these tissues have been identified (42). Given Tregs can arise in an antigen epitopes-dependent manner, is this a driving factor in Treg differentiation, or does tissue environment shape Treg transcriptomes? To test this, we collected seven Treg microarray datasets from NIH NCBI-Geo Datasets database (https://www.ncbi.nlm.nih.gov/gds/). These covered four tissue Treg datasets from spleen (SP, s-Treg), lymph nodes (LN, LN-Treg), intestine (small intestinal lamina propria, int-Treg) and visceral adipose tissue (VAT-Treg), one IL-2 receptor β deficient (IL2rβ-/-) Treg, hepatocellular carcinoma Treg, co-stimulatory antibody-treated Treg (Table 1). Of note, these datasets were obtained from high-quality experiments since the expression variations of nine housekeeping genes were in limited ranges (Table 2 and Supplementary Table 1 of Housekeeping Genes). Surprisingly, when Treg transcriptomes from four tissues were compared to that of Tconv shown in Table 3, s-Treg had 31 genes (0.2%) upregulated, 13 genes (0.09%) downregulated from Tconv; LN-Treg had 325 genes (1.5%) upregulated, 72 genes (0.33%) downregulated from Tconv; int-Treg had 371 genes (1.7%) upregulated, 385 genes (1.77%) downregulated from Tconv; and VAT-Treg had 641 genes (2.97%) upregulated, 283 genes (1.31%) downregulated from Tconv. Our results correlated well with a previous report of 200 differentially expressed genes between human type 1 T helper cells (Th1) and and type 2 T helper cells (Th2) (81). These results have demonstrated that first, comparing to int-Treg and VAT-Treg, s-Treg and LN-Treg were 10 times less different from that of Tconv in transcriptomes, suggesting that non-lymphoid tissue environments play a significant role in re-shape Treg transcriptomes; second, the ratios of upregulated/downregulated genes in LN-Treg and VAT-Treg were a few folds bigger than that in s-Treg and int-Treg, suggesting that functional selection plays critical roles in re-shape the transcriptomes; and third, significant expansions of transcriptomic differences in Treg versus Tconv in VAT and intestine in comparison to that Treg in lymphoid tissues such as spleen and lymph nodes indicate that adipose tissues environments and intestine have more types and high strengths of stimuli such as danger/pathogen-associated molecular patterns (DAMPs/PAMPs); and transcriptomic regulatory signals from tissue environment cues are much bigger than that antigen epitopes-dependent Treg differentiation and polarization in healthy conditions.

Tissue diversification of Treg transcriptomic differentiation from Tconv emphasized the tissue environmental effects and signals on transcriptomic remodeling. We hypothesized that regardless of tissue re-modeling, Treg from four different tissues share Treg signature genes in addition to the differences. To test this hypothesis, we used the Venn Diagram to analyze Treg genes from four tissues. As shown in Figure 1A-A, the results on Treg upregulated genes showed that: four tissue Treg shared 11 upregulated genes including Foxp3 (38, 51), Fam129a (NIBAN, Niban apoptosis regulator 1) (82), tumor necrosis factor receptor (TNFR) superfamily member 4 (TNFRSF4, CD134, OX40) (83, 84), IKAROS family zinc finger 2 (Ikzf2) (85), TNFR type II (TNFRSF1b, CD120b) (86), Dual Specificity Phosphatase 4 (DUSP4) (87), Vav guanine nucleotide exchange factor 2 (Vav2) (88), capping actin protein, gelsolin like (Capg) (84), Ctla4 (84), TNFR superfamily member 18 (TNFRSF18, GITR, CD357) (84) and killer cell lectin like receptor G1 (Klrg1) (89); and four tissue Treg shared five downregulated genes (Figures 1A, B) such as heparan sulfate-glucosamine 3-sulfotransferase 3B1 (Hs3st3b1), insulin like growth factor binding protein 4 (Igfbp4), transmembrane inner ear expressed protein (Tmie), transforming growth factor Beta receptor 3 (Tgfbr3), and ATPase Na+/K+ transporting subunit Beta 1 (Atp1b1). Of note, it has been reported that nine cytokines and 88 transcription factors are differentially expressed among tissue Treg but not shared genes (90). Comparing with the 12 core Treg genes expressed in all Treg regardless of their areas recently reported using single cell RNA-Seq (84) including Foxp3, Il2ra, Il2rb, Tnfrsf4, Ikzf2, Ctla4, Capg, Izumotr, Chchd10, Tnfrsf18, Dapl1, and Igfbp4, our results in Figure 1B-A and Figure 3A-A have identified five new upregulated Treg core genes regardless of tissues including Fam129a (NIBAN, regulating p53-mediated apoptosis), TNFRSF1b (CD120b, mediating metabolic effects of TNFa), DUSP4 (dual specificity phosphatase 4, dephosphorylating mitogen-activated protein kinases (MAPKs) extracellular signal-regulated kinases 1/2, ERK1/2), Vav2, EFNA1-induced RAC1 GTPase activation and vascular endothelial cell migration and assembly), and Klrg1 [killer cell lectin like receptor G1, playing an inhibitory role on natural killer (NK) cells and T-cell functions upon binding to their non-MHC ligands] and five downregulated genes Hs3stb1, Tmie, Tgfbr3, Igfbp4 and Atp1b1 (Figure 1A-B).

Figure 1

To further determine the signaling pathways that our 16 Treg signature genes (Figure 1B), we used the Metascape database (http://metascape.org/gp/index.html#/main/step1http://metascape.org/) that is suitable for analyzing small numbers of genes in comparison to that of Ingenuity Pathway Analysis (IPA) database. As shown in Figure 1B-B, two new Treg signaling pathways were i) RUNX1 and Foxp3 control Treg development (49), and ii) TNFs bind their physiological receptors. Two additional pathways included: 1) negative regulation of protein modification process and 2) negative regulation of cellular component organization. Of note, a recent report identified a new all-trans retinoic acid-induced CD161+ Treg in the intestine, which have a transcription network including BTB Domain And CNC Homolog 2 (BACH2), RAR related orphan receptor C (RORγT), FOS like 2, AP-1 transcription factor subunit (FOSL2), AP-1 (c-Jun and c-Fos), and RUNX Family Transcription Factor 1 (RUNX1) (91). It was reported that RUNX1 is required for the optimal regulation of Foxp3 expression in human T cells (92); and differentiating Treg will have recognized their cognate antigens and received T cell antigen receptor (TCR) signals before initiating Foxp3 transcription, which is triggered by TCR-induced transcription factors including Nuclear Factor Of Activated T Cells 2 (NFAT2), AP-1 (Jun and Fos) and NF-κB. Once expressed, Foxp3 seizes TCR signal-induced transcriptional and epigenetic mechanisms through interacting with AML1/Runx1 and NFAT (93). Thus, Foxp3 modifies gene expression dynamics of TCR-induced genes, which constitute cardinal mechanisms for Treg-mediated immune suppression. Our results have demonstrated that RUNX1 may also be a shared Treg transcription factor. Taken together, our results demonstrated that transcriptomic differences between Treg and Tconv are small, ranging between 0.29 and 4.28% (1–14.8 folds) among s-Treg, LN-Treg, int-Treg and VAT-Treg; five new Treg core genes including NIBAN, TNFRSF1b, DUSP4, Vav2, and Klrg1 have been identified; and non-lymphoid tissue niches and lymph node niches contribute more than splenic niches to Treg transcriptomic differences from that of CD4⁺Foxp3^- conventional T cell (Tconv).

2. A list of 68 new Treg signature genes have been generated from three partially overlapped gene groups; and among 27 tissue Treg shared pathways, 22 pathways (81.5%) are inante immune pathways.

We then hypothesized that Treg signature genes are induced by Treg specific transcription factor Foxp3. An excellent report identified 50 genes induced by Foxp3 and 37 genes suppressed by Foxp3 (94). Based on the expression levels based on the results from Nanostring of the 50 Foxp3-induced Treg genes and 37 Foxp3-suppressed Treg genes by a series of wild-type Foxp3 and Foxp3 mutants, we performed multiple linear regression analyses to determine the genes statistically significantly induced by Foxp3 and suppressed by Foxp3 (Figure 2A). The original Nanostring data were obtained from the Supplementary Table 4 of a paper from Dr. Benoist’s team (94). As shown in Figure 2B, nine out of reported 50 Foxp3-induced genes, and 11 out of reported 37 Foxp3-suppressed genes were identified with statistical significance (p < 0.05). The nine statistically significant Foxp3-induced genes were newly related to Foxp3 function.

Figure 2

As shown in Figure 2C, the Metascape analysis identified 14 signaling pathways out of the reported 50 Foxp3-induced genes. As shown in Figure 2D, after our statistical re-selection, the Metascape analysis only identified two pathways; regulation of mononuclear cell proliferation and cytokine mediated signaling pathway. As shown in Figure 2E, using protein-protein interaction (PPI) network database (https://www.sciencedirect.com/topics/medicine-and-dentistry/protein-protein-interaction), we found two hubs when using the original 50 reported Foxp3-induced genes, one TNF centered, one Foxp3 centered. Reanalylsis with the nine significant genes out of 50 resulted in only the Foxp3 hub. These results demonstrated that our re-selection of statistically significant Foxp3-induced genes was appropriate to zoom in on the Foxp3 regulated core genes.

A recent single-cell RNA-Seq analysis reported the identification of 20 Treg core genes (95). To consolidate Treg signature genes, we collected 11 Treg genes shared in four tissues, 50 reported Foxp3-induced genes, and 20 thymic Treg cluster genes and performed the Venn Diagram analysis as shown in Figures 3A, B. The results identified 68 upregulated Treg genes and 40 downregulated genes (Table 4). When those upregulated Treg signature genes were searched in tissue Treg transcriptomic data, 15 out of 68, 39 out of 68, 30 out of 68, and 32 out of 68 upregulated Treg signature genes were upregulated in s-Treg, LN-Treg, int-Treg and VAT-Treg, respectively. We also found 2, and 5 out of 68 upregulated Treg signature genes were downregulated in int-Treg and VAT-Treg, respectively. Similarly, we found that 0 out of 40, 4 out of 40, 2 out of 40, and 12 out of 40 downregulated Treg signature genes were upregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively. We also found 5 out of 40, 9 out of 40, 15 out of 40, and 13 out of 40 downregulated Treg signature genes were downregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively (Table 5). To determine the distribution of 68 upregulated Treg signature genes in Treg in four tissues, we performed the Venn Diagram analysis. As shown in Figure 3B, the results showed that 1) s-Treg shared the same 15 upregulated Treg signature genes with LN-Treg and int-Treg; 2) s-Treg shared 11 upregulated Treg signature with LN-Treg, int-Treg, and VAT-Treg; 3) LN-Treg shared 27 and 23 upregulated Treg signature genes with int-Treg and VAT-Treg, respectively; and 4) int-Treg shared 20 upregulated Treg signature genes with VAT-Treg.

Figure 3

Our results have demonstrated that i) These 11 shared Treg signature genes were the same as 11 genes shared by Treg from four tissues shown in Figure 1A-A, suggesting that the 11 genes are the Treg essential suppressive function-required regardless of tissue environments; ii) nine statistically significant Foxp3-induced genes are different from these 11 core Treg signature genes except Foxp3, suggesting that nine out of 11 core Treg signature genes are not induced by Foxp3 but contribute to Treg suppressive function; iii) 20 thymic Treg cluster genes share only four genes with 11 Treg signature genes shared by Treg from four peripheral tissues and reported 50 Foxp3-induced genes such as Foxp3, TNFRSF4 (CD134, OX40), CTLA4 and TNFRSF18 (GITR, CD357); iv) seven out of 11 Treg core genes shared by Treg from four tissues and 46 out of reported 50 Foxp3-induced genes do not share with 20 thymic Treg cluster genes, suggesting that those 53 genes (93%) out of 57 genes (Figure 3A-A) are the peripheral Treg genes expressed from Foxp3 induction in peripheral tissues or signaling pathways from peripheral tissue environments but not expressed in thymic Treg; and v) 16 out of 20 (80%) thymic Treg cluster genes are not shared with tissue Treg, suggesting that thymic Treg lose 80% of thymic cluster genes when they adapt tissue environments.

We hypothesized that tissue Treg have their own signaling pathways associated with their upregulated genes and downregulated genes. To test this hypothesis, we performed IPA with the tissue Treg genes shown in Table 3. IPA did not yield any significant pathways since s-Treg had 31 genes upregulated and 13 downregulated in comparison to that of Tconv (Table 3). As shown in Table 6 and Supplementary Figure 1, the results showed LN-Treg had 47 significant pathways upregulated and one pathway downregulated; int-Treg had 45 pathways upregulated and 63 pathways downregulated; and VAT-Treg had 116 pathways upregulated and 48 pathways downregulated. As shown in Figure 3C, the results showed that 1) LN-Treg shared upregulated two pathways with int-Treg in T Helper Cells, and shared 12 upregulated pathways with VAT-Treg. Of note, these 14 pathways indicated lymphoid Treg to non-lymphoid Treg transition; 2) int-Treg shared 15 upregulated pathways with VAT-Treg, which indicated non-lymphoid Treg shared pathways; 3) three tissue Treg shared 27 upregulated pathways; and 4) based on activation z scores, we compiled all the top signaling pathways from LN-Treg, int-Treg, and VAT-Treg in Figure 3C. We found that the top 30 pathways in Figure 3C overlapped with the 27 Treg shared pathways in Figure 3C.The differences between those two lists included the following few pathways: Glioma Invasiveness Signaling was on the 27 shared pathways whereas four pathways on the 30 top pathways but not on the 27 Treg shared pathways included dendritic cell maturation, osteoarthritis pathway, NF-kB signaling, and B cell receptor signaling.

We then hypothesized that Treg from various tissues have shared upstream master regulators. To test this hypothesis, we performed IPA. As shown in Figure 4A, the results showed that Treg upregulated genes in each tissue can all be targeted by upstream regulator IL-2. In addition, Treg downregulated genes in the intestine and VAT can also be targeted by IL-2 (Figure 4B). Of note, IL-2 as the common upstream regulator was not on the IPA pathways on the 27 shared Treg pathways in Figure 3C, suggesting the IPA pathways in both tissue Treg specific pathways and shared Treg pathways are mostly downstream signaling pathways rather than the upstream regulator(s). Moreover, the results in Figure 4C showed that upregulated genes in Treg from LN, intestine and VAT but not spleen can all be targeted by universal upstream regulator NF-kB, which were well correlated with the top transition pathways from lymphoid Treg to non-lymphoid Treg that LN-Treg shared with int-Treg and VAT-Treg in Figure 3C. From the top pathway comparison data from four tissue Treg, NF-KB z scores were high in LN-Treg and int-Treg but lower in VAT-Treg.

Figure 4

Taken together, these results have demonstrated that first, the 27 shared signaling pathways are the key signaling pathways for Treg suppressive functions regardless of tissue differences; second, LN-Treg have eight specific signaling pathways; int-Treg have only one specific pathway; third, the 14 shared pathways between LN-Treg and int-Treg and LN-Treg and VAT-Treg focused on NF-kB, ICOS co-stimulation, DC maturation, and metabolisms; fourth, the pathways shared by int-Treg and VAT-Treg focused on autoimmune lupus, phagocytosis in macrophages and monocytes, neuroendocrine signaling, cytokine/chemokine and hormone secretion and senescence; fifth, all three tissue Treg shared pathways cover a broad spectrum of functions. The results have clearly demonstrated that Treg are functional not only in suppressing immune and autoimmune responses, inflammations, but also promoting muscle repairing, tissue regeneration, and tissue homeostasis, etc; and sixth, IL-2 and NF-KB were the universal upstream regulators for tissue Treg. Our previous report showed that in addition to kinase-phosphatase regulatory mode, NF-KB canonical and non-canonical pathways can be regulated in pre-translational mode (96). Our findings here are correlated well with recent reports that NF-kB canonical pathway components c-Rel is critical for thymic Treg development while p65 is essential for mature Treg identity and maintenance of immune tolerance by promoting the formation of a Foxp3-specific enhanceosome (97, 98); and NF-kB alternative (non-canonical) pathway components p100 (nfkb2) is essential for Treg suppressive function and inhibition of RelB (99–101).

3. S-Treg have eight CD markers but no specific ones; 12-, 7-, and 15- CDs have been identified as tissue-specific effector Treg markers for LN-, intestine-, and VAT-Treg, respectively; interactions between CDs and their receptors mediate tissue Treg signaling; and S-Treg are the most naïve peripheral lymphoid Treg.

Identification of clusters of differentiation (CDs) using monoclonal antibodies has revolutionized immunology. Originally CD4+CD25+ or CD4+CD25^high was the markers for characterization of Treg (26, 27, 102), later on Treg specific transcription factor Foxp3 was introduced as a reliable marker for Treg (29, 49, 51, 53). As we recently reviewed, at least six Treg subsets can be identified including: i) Fxop3⁺CD28⁺GranzymeB⁺Helios⁺TGFβ-insensitive thymic (tTreg), ii) CD62L^highCCR7 (CD197)⁺ or CD45RA^highCD25^low central Treg; iii) Foxp3⁺CTLA4⁺IL-10⁺TGFβ-sensitive inducible Treg (iTreg)/peripheral Treg (pTreg), iv) IL-35 secreting Treg (iTreg35), v) CD62L^lowCCR7^lowCD44^highKLRG⁺CD103⁺ or CD45RA^lowCD25^high effector Treg, and vi) Foxp3⁺other transcription factor+ resident Treg (28). Of note, Helios (transcription factor), TGFb (cytokine), IL-10 (cytokine), granzyme B (multiple cellular locations), and IL-35 (cytokine) are not cell surface CD markers, which make flow cytometry challenging since the intracellular staining of monoclonal antibodies with a Golgi blocker is needed (https://www.bdbiosciences.com/ds/pm/tds/555029.pdf). Therefore, the characterization of novel CD markers for Treg will significantly advance our understanding on Treg function and homeostasis. We hypothesized that transcriptomic analysis of CD marker expression of Treg facilitate the identification of new CD markers for tissue Treg. We collected a complete list of 373 CDs from a comprehensive protein database (https://www.proteinatlas.org/search/protein_class:CD+markers). As shown in Figure 5A in comparison to the counterpart Tconv, s-Treg upregulated 8 CDs; LN-Treg upregulated 40 CDs, and downregulated 4 CDs; intestine (LP) Treg upregulated 33 CDs and downregulated 20 CDs; and VAT-Treg upregulated 40 CDs and downregulated 22 CDs.

Figure 5

We hypothesized that Treg from four tissues have shared and specific CD markers. We performed the Venn Diagram analysis to test this hypothesis. As shown in Figure 5C, the results showed that i) s-Treg shared total eight CDs with LN-Treg. S-Treg shared seven of the eight CD markers with int-Treg except CD86, which may be the naïve-like central Treg (103) markers; ii) LN-Treg had 12 specific CD markers, suggesting that these 12 CD markers were the lymphoid effector Treg markers; iii) LN-Treg shared 19 CD markers with int-Treg; iv) LN-Treg shared 18 CD markers with VAT-Treg. Of note, among 17 CD markers shared, first 10 CD markers also shared with LN-Treg. The last seven markers were non-lymphoid effector Treg-specific CD markers (103) including ALCAM (CD166), EPCAM (CD326), ITGAV (CD51), ADAM8 (CD156), TFRC (CD71), BMPR1A (CD292), and MUC1 (CD227); vii) VAT-Treg had 15 specific CD markers; and viii) all four tissue Treg shared four CD markers.

As shown in Figure 5B, the Metascape pathway analysis identified 20 pathways using upregulated CDs from four tissue Treg, which are all related to innate and adaptive immune regulations. Among 20 pathways, four pathways were shared in all four tissue Treg even including s-Treg including 1) lymphocyte activation, 2) cytokine-cytokine receptor interaction, 3) TNFs bind their physiological receptors, and 4) cytokine production. Of note, these four pathways were addition to the 27 shared Treg pathways identified from LN-Treg, int-Treg and VAT-Treg in Figure 3C.

Taken together, our results have demonstrated that first, since LN-Treg (40 upregulated CD markers) have five folds more CD markers than s-Treg (8 upregulated CD markers), suggesting that LN-Treg are more likely to be effector Treg; and s-Treg are more likely to be naïve-like central Treg (28, 103), which may just be matured from thymic Treg in peripheral lymphoid tissue. Spleen is a peripheral tissue of the circulatory system embedded with multiple LN-like structures (White pulp, WP), which functions similarly to how LNs drain and monitor antigens from tissues. In the WP, naïve and central memory T cells are activated in response to cognate antigens. CCR7 is required for T cell concentration in the T cell zone in the WP (104). However, our results in Figure 5C showed that s-Treg have no increased CCR7 levels over Tconv, resulting in a scattering of relatively naïve Treg throughout the spleen. Second, those seven CD markers shared between three Treg such as s-Treg, LN-Treg, and int-Treg are more likely to be central Treg CD markers shared by lymphoid and non-lymphoid tissue Treg; Third, CD86 shared by s-Treg and LN-Treg is the only central lymphoid Treg marker;Fourth, LN-Treg have 12 specific Treg markers, which serve as effector lymphoid Treg markers; Fifth, int-Treg have seven specific markers and VAT-Treg have 15 specific markers, which serve as tissue effector Treg markers; Sixth, seven CD markers shared by int-Treg and VAT-Treg are the non-lymphoid effector Treg markers; Seventh, four s-Treg pathways have been identified, which are shared with other three tissue Treg and four CD markers such as TNFRSF18 (CD357), CTLA4 (CD152), TNFRSF1B (CD120b), and TNFRSF4 (CD134) shared by all four tissue Treg, which are the essential core CD markers and functional pathways regardless of tissue environments; and Eighth, since these results are transcriptomic data, the future experiments will be needed to verify those novel CD markers with flow cytometry using monoclonal antibodies.

4. Seven, 49, 44, and 79 increased genes out of total 1176 cytokines have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively, suggesting that LN-Treg, int-Treg, and VAT-Treg are more active than s-Treg in generating immunosuppressive and homeostatic cytokines.

Several functional modes have been identified for Treg such as cytokine secretion such as Amphiregulin secreted from muscle Treg and acted on muscle satellite cells (68), cell surface protein interactions such as co-stimulation receptors and immune checkpoint receptors (105, 106), and granzyme B-mediated killing of target cells, etc (28, 29, 53).. We hypothesized that Treg from different tissues, lymphoid and non-lymphoid, secrete different cytokines from re-shaped Treg transcriptomes. To test this hypothesis, we collected 1,176 cytokines and their interactors (receptors) from a comprehensive protein database (https://www.proteinatlas.org/search/cytokine) as we reported recently (10). As shown in Figure 6A, 7, 49, 44, and 79 cytokines were upregulated in s-Treg, LN-Treg, int-Treg and VAT-Treg, respectively. Two, 5, 40, and 35 cytokines were downregulated in s-Treg, LN-Treg, int-Treg and VAT-Treg, respectively.

Figure 6

In addition, the Metascape analysis showed in Figure 6B that 12 signal pathways associated with upregulated cytokines were shared among four tissue Treg. These results suggest that these 12 pathways are Treg cytokines-shared pathways and functions. In addition, eight pathways including regulation of leukocyte migration, hematopoietic cell lineage, interleukin-1 family signaling, cell response to molecule of bacterial origin, interleukin-6 family signaling, interleukin-5 signaling, interleukin-20 family signaling, and leukocyte chemotaxis are only shared by LN-Treg and tissue Treg, which are innate immune function pathways.

The Venn Diagram analysis of upregulated cytokines in four tissue Treg showed in Figure 6C that seven cytokines in s-Treg all shared with LN-Treg including LIF, IL-7, IFNL1, TNF, IL12B, and SOCS1. LN-Treg had 13 specific cytokines, shared 14 cytokines and 15 cytokines with int-Treg, and VAT-Treg, respectively. Int-Treg had 16 specific cytokines, and shared 19 cytokines with VAT-Treg. VAT-Treg had 44 specific cytokines. These results have demonstrated that first, LN-Treg, int-Treg, and VAT-Treg are most secretory tissue Treg in comparison to s-Treg; second, LN-Treg, int-Treg, and VAT-Treg have their specific cytokine panels, which have clearly indicated their own immune regulatory functions in addition to the shared Treg functions carried out by only two cytokines such as IL-12B and SOCS1; and third, VAT-Treg are highly secretory, which suggests that VAT-Treg play significant homeostatic roles for whole body.

5. Eight, 31, 36, and 51 upregulated genes out of total 1,706 secretomic genes have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively, suggesting that LN-Treg, int-Treg, and VAT-Treg have 13 pathway functions more than s-Treg via additional secretomic proteins.

As we introduced in the beginning, the secretome, defined as a portion of total proteins secreted by cells to the extracellular space, secures a proper micro-environmental niche, thus maintaining tissue homeostasis (62, 63). Secreted molecules are key mediators in cell-cell interactions and influence the cross-talk with the surrounding tissues in addition to their endocrine functions in long-distance as previously demonstrated by hormones, growth factors, cytokines, adipokines, myokines, cardiokines (64), and chemokines (65). In addition to cytokines discussed in the previous section, we hypothesized that Treg in various tissues have different secretomes to fulfill their immunosuppressive and homeostatic functions. To test this hypothesis, we collected 1,706 secretomic genes from a comprehensive protein database (https://www.proteinatlas.org/search/cytokine) as we reported recently (10). As shown in Figure 7A, 8, 31, 36, and 51 secretomic genes were upregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively. Zero, 4, 25, and 24 secretomic genes were downregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively.

Figure 7

In addition, the Metascape analysis showed in Figure 7B that in upregulated secretomes, all four tissue Treg have three shared pathways. In addition, s-Treg have a specific pathway; VAT-Treg also have a specific pathway; int-Treg and VAT-Treg shared two pathways. Moreover, LN-Treg, int-Treg, and VAT-Treg shared 13 pathways.

The Venn Diagram analysis of upregulated secretomes in four tissue Treg showed in Figure 7C that s-Treg had five upregulated secretomic genes, s-Treg shared one secretomic gene with LN-Treg, and shared two secretomic genes such as TNFRSF1B and TNFRSF18 with LN-Treg and other two tissue Treg. LN-Treg have eight specific secretomic genes, seven shared with int-Treg, and six shared with int-Treg and VAT-Treg. Int-Treg had 10 specific secretomic genes, and shared six with VAT-Treg. VAT-Treg had 26 specific secretomic genes. Taken together, these results have demonstrated that first, each tissue Treg have their own secretomes in addition to cytokines to fulfill their functions; and second, LN-Treg, int-Treg, and VAT-Treg share 13 pathway functions more than s-Treg; and third, VAT-Treg have more secretomic genes upregulated than other tissue Treg to carry out unique systemic regulatory functions.

6. Two, 20, 25, and 43 increased transcription factors (TFs) out of total 1,496 TFs have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively, suggesting that non-lymphoid tissue (NLT) Treg such as int-Treg and VAT-Treg carry out half to 70% non-immunosuppressive functions, which are not shared with other tissue Treg.

We previously reported that GATA3, HDAC6, and Bcl-6 regulate Foxp3+ Treg plasticity and determine Treg conversion into either novel antigen-presenting cell-like Treg or Th1-Treg (29), suggesting that other T helper cell subsets such as Th2 TF GATA3, Tfh TF Bcl-6 and HDAC6 cooperate with Foxp3 to determine Treg transcriptomes and functions. We hypothesized that tissue Treg have a specific TF sets in addition to Foxp3. To test this hypothesis, we collected 1,496 TFs from a comprehensive protein database (https://www.proteinatlas.org/search/cytokine) as we reported recently (10). As shown in Figure 8A, 2, 20, 25, and 43 increased genes out of total 1,496 transcription factors (TFs) have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively. In addition, 0, 3, 12, and 10 downregulated genes out of total 1,496 TFs have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively.

Figure 8

In addition, the Metascape analysis showed in Figure 8B that in upregulated TFs, s-Treg had no specific pathways from two upregulated TFs; all three tissue Treg had four shared pathways including cell fate commitment, leukocyte differentiation, PID NFAT TFpathway, in utero embryonic development. In addition, VAT-Treg and LN-Treg shared two pathways such as transcriptional misregulation in cancer, and circadian rhythm; and VAT-Treg shared with int-Treg two pathways such as post-embryonic development and myeloid cell homeostasis. Moreover, int-Treg had two specific pathways such as cmyb pathway and connective tissue development.

The Venn Diagram analysis of upregulated TFs in four tissue Treg showed in Figure 8C that s-Treg had no specific TFs and had two TFs shared with other three tissue Treg such as Foxp3 and IKZF2. In addition, LN-Treg had six specific TFs; LN-Treg shared five TFs with int-Treg; LN-Treg shared two TFs with int-Treg and VAT-Treg; and LN-Treg shared five TFs with VAT-Treg. Moreover, int-Treg had 12 specific TFs; and int-Treg shared four TFs with VAT-Treg. Finally, VAT-Treg had 30 specific TFs.

Taken together, these results have demonstrated that first, s-Treg have no specific TFs out of two upregulated TFs; LN-Treg have six specific TFs (30%) out of 20 upregulated TFs, int-Treg have 12 specific TFs (48%) out of 25 upregulated TFs; and VAT-Treg have 30 specific TFs (69.8%) out of 43 upregulated TFs for their own transcriptomes and functions; second, tissue Treg specific TFs allow those Treg to carry out more homeostatic functions than Foxp3-defined immunosuppressive functions; and third, int-Treg and VAT-Treg carry out half and 70% non-immunosuppressive functions, which are not shared with other tissue Treg. Our findings are well correlated with previous reports of others (107). For example, 1) intestine Treg can be classified into three subsets including GATA3⁺Helios⁺ (Nrp1⁺, thymic origin) (107); RORgt⁺Helios⁻ (microbial immunity), and RORgt⁻Helios⁻ (against dietary antigens); 2) contrary to obese animals, depletion of VAT-Treg in aged animals improves the metabolic parameters and rescues aging-induced insulin resistance (107); and 3) brown adipose tissue Treg help in thermogenesis (107).

7. Zero, 14, 10, and 16 increased regulators out of total 305 cell death regulatomes in 13 cell death forms have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively; LN-Treg and int-Treg have increased pyroptosis regulators but VAT-Treg have increased apoptosis regulators.

We and others previously reported that Treg cell death pathways related to disease conditions (26–28, 46–52). However, an important question remained unknown that the expression of how many cell death regulators are regulated in tissue Treg in physiological conditions. We hypothesized that the expressions of cell death pathway regulators are differentially regulated in tissue Treg. To test this hypothesis, we collected 13 newly characterized cell death pathway regulators (termed cell death regulatome), total 305 regulatory genes as we reported previously (108). As shown in Table 7A, these 13 cell death pathways include apoptosis, autophagy-dependent cell death (ADCD) regulated, anoikis (specific variant of intrinsic apoptosis initiated by the loss of integrin-dependent anchorage) related, entotic cell death [a type of regulated cell death (RCD) that originates from actomyosin-dependent cell-in-cell internalization (entosis) and is executed by lysosomes], ferroptosis [a form of RCD initiated by oxidative perturbations of the intracellular microenvironment that is under constitutive control by glutathione peroxidase 4 (GPX4) and can be inhibited by iron chelators and lipophilic antioxidants], immunogenic cell death (ICD) regulated, lysosome dependent cell death, mitotic catastrophe regulated (oncosuppressive mechanism for the control of mitosis-incompetent cells by RCD or cellular senescence), mitochondrial permeability transition (MPT)-driven necrosis, necroptosis [a modality of RCD triggered by perturbations of extracellular or intracellular homeostasis that critically depends on mixed lineage kinase domain like pseudokinase (MLKL), receptor interacting serine/threonine kinase 3 (RIPK3), and (at least in some settings) on the kinase activity of RIPK1], NETotic [a reactive oxygen species (ROS)-dependent modality of RCD restricted to cells of hematopoietic derivation and associated with neutrophil extracellular trap (NET) extrusion], ParThanatos [A modality of RCD initiated by poly(ADP-ribose) polymerase 1 (PARP1) hyperactivation and precipitated by the consequent bioenergetic catastrophe coupled to Apoptosis-inducing factor (AIF)-dependent and macrophage migration inhibitory factor (MIF)-dependent DNA degradation] (109), and inflammatory cell death [pyroptosis, a type of RCD that critically depends on the formation of plasma membrane pores by members of the gasdermin protein family, often (but not always) as a consequence of inflammatory caspase activation] (110).

As shown in Table 7A, we identified 305 regulatory genes associated 13 types of cell death pathways. The results showed that s-Treg had no the expression changes of cell death regulators in comparison to that of Tconv. LN-Treg had 14 cell death regulator upregulation including two (out of 102 in total) in apoptosis, one out of 22 regulators in ADCD regulated, two out of 23 regulators in ICD regulated, three out of 28 regulators in mitotic catastrophe regulated, one out of 10 necroptosis regulators, six out of 23 (26.1%) pyroptosis regulators (Table 7B) including Naip5, P2Rx7, caspase-1, caspase-4, IFNGR1, and TLR7. In addition, int-Treg had one regulator upregulation in each of seven cell death forms such as apoptosis, ADCD, anoikis, entotic, ferroptosis, ICD, necroptosis and three out of 23 (13%) regulator upregulation in pyroptosis. Moreover, VAT-Treg had five regulator (out of 102, 4.9%) upregulation in apoptosis, two regulator upregulation in each of five cell death forms such as entotic, ferroptosis, mitotic catastrophe, necroptosis, and pyroptosis, and one regulator upregulation in NETotic. Furthermore, LN-Treg had 14 regulator upregulation but only one regulator downregulation (up/down = 14/1); int-Treg had 10 regulator upregulation but seven regulator downregulation (up/down = 10/7); and VAT-Treg had 16 regulator upregulation but seven regulator downregulation (up/down = 16/7) (Table 7A). These results have demonstrated that first, s-Treg have no expression changes of cell death regulators in comparison to that of Tconv, suggesting that s-Treg is more protected from cell death stimulations in physiological conditions than other tissue Treg; second, tissue Treg have more upregulation than downregulation of cell death regulators, suggesting that upregulation of cell death regulators is a Treg response to stimuli; third, LN-Treg and int-Treg have more pyroptosis regulator upregulation than VAT-Treg but VAT-Treg have more apoptosis regulator upregulation than other tissue Treg, suggesting that LN-Treg and int-Treg have higher inflammatory cell death and major histocompatibility complex class II (MHC-II)/antigen epitope-T cell antigen receptor (TCR) signaling-independent innate immune potential, as we recently reported (29), in response to stimulations of DAMPs than Tconv in the same tissues and Treg in other tissues, which are well correlated with that reported (111). It has been reported that the activation of the pyroptosis effector NLRP3 inflammasome has a crucial role in the immunoprotection against pulmonary paracoccidioidomycosis by promoting the expansion of Th1/Th17 immunity and reducing the suppressive control mediated by Treg cells (112); loss of Treg cytokine IL10 signaling leads to intestinal inflammation, at least in part, through increased production of IL-1 by innate immune cells, leading to activation of CD4⁺ T cells (113); and particulate matter (PM) exposure leads to an immunosuppressive lung environment with higher recruitment of Treg in the presence of caspase-1 inhibitor (114). These results suggest that immunosuppressive Treg inhibit activation of inflammasomes/caspase-1 activation-pyroptosis. Taken together, our results that increased pyroptosis potential in LN Treg and int-Treg may suggest higher potential of these Treg develop Treg pyroptosis and plasticity than s-Treg and VAT-Treg.

8. One, 15, 19, and 31 increased kinases out of total 661 kinome have been identified for s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively.

It has been reported that kinases play important roles in regulating Treg functions including serine/threonine mammalian target of rapamycin (mTOR) pathway (115), phosphoinositide 3-kinase δ (PI3K δ) (116), PI3K/Akt/mTOR cascade (117), AMPK (118), non-receptor tyrosine kinase IL-2 inducible T cell kinase (ITK) (119), glycogen synthase kinase 3β (120), STAT5B (121), and NF-kB (101), etc. Also, reduced expression of phosphatase PTPN2 promotes pathogenic conversion of Tregs in autoimmunity (122). However, an important question remained whether tissue Treg have different kinase pathways. We hypothesized that tissue Treg have upregulated kinase pathways that fit their specific functions and Treg shared functions. To test this hypothesis, we examined the expression changes of total kinome (a complete list of 661 kinases encoded in human genome) (123) in four tissue Treg. As shown in Figure 9A, 1, 15, 19, and 31 kinases out of 661 total kinome are upregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively. In addition, 1, 3, 20, and 10 kinases were downregulated in s-Treg, LN-Treg, int-Treg, and VAT-Treg, respectively. As shown in Figure 9B for upregulated kinase pathways, six pathways were shared between int-Treg, LN-Treg, and VAT-Treg. In addition, LN-Treg shared five pathways with VAT-Treg. Moreover, int-Treg shared one pathway with LN-Treg tissue remodeling. Int-Treg shared one pathway with VAT-Treg fluid shear stress and atherosclerosis. Furthermore, int-Treg had two specific pathways such as regulation of protein secretion, and lipid phosphorylation. LN-Treg had two specific pathways. VAT-Treg had three specific pathways.

Figure 9

The Venn Diagram analysis of upregulated kinases in four tissue Treg showed in Figure 9C that s-Treg had one kinase DUSP4 shared with other three tissue Treg. In addition, LN-Treg had six specific kinases; LN-Treg shared four kinases with int-Treg; LN-Treg shared one kinase BMPR2 with int-Treg and VAT-Treg; and LN-Treg shared three kinases with VAT-Treg. Moreover, int-Treg had eight specific kinases; and int-Treg shared five kinases with VAT-Treg. Finally, VAT-Treg had 21 specific kinases. These results have demonstrated that first, LN-Treg, int-Treg, and VAT-Treg have six, eight, and 21 specific kinase pathways, respectively, for their specific functions; second, DUSP4 is the kinase pathway shared between all four tissue Treg including s-Treg; and BMPR2 is the kinase pathway shared between three Treg such as LN-Treg, int-Treg, and VAT-Treg; and third, additional three sets of two tissue Treg-shared 12 kinases: a) four kinases for LN-Treg and int-Treg, b) three kinases LN-Treg and VAT-Treg, and c) five kinases for int-Treg and VAT-Treg are all important for tissue Treg functions.

9. IL2Rβ plays essential roles in promoting all tissue Treg shared functions and specific functions.

We and others reported that IL-2 signaling pathway is essential for Treg survival (26–28, 46–52); and low dose IL-2 has been developed as a new Treg-based therapy (124). However, a few important questions remained how IL-2 regulate Treg signature genes and cytokines and chemokines. We hypothesized that IL-2 receptor signaling plays a critical role in regulating the expression of Treg signature genes and cytokines and chemokines. To examine this hypothesis, we found IL-2 receptor b (IL2Rβ, CD122) KO Treg microarray datasets (GSE14350) in the NCBI-GeoDatasets database. Of note, IL-2 binds with low affinity to IL-2Rα (CD25) or the common γ -chain (γc, CD132)-IL-2Rβ heterodimers, but receptor affinity increases ~1,000 folds when these three subunits together interact with IL-2. IL-2 shares the common γ -chain (γc)-IL-2Rβ heterodimers with IL-15; and IL2Rβ is indispensable for Treg cell function (125). As shown in Table 8A, 87 genes were upregulated and 235 genes were downregulated in IL2Rβ (IL2RB) KO Treg, suggesting that IL2RB induces 235 genes and suppresses 87 genes. We then examined whether IL2RB regulates Treg signature gene expressions by determining the expression of 68 Treg upregulated signature genes and 40 Treg downregulated genes. As shown in Table 8B, 12 out of 68 (17.6%) Treg upregulated signature genes were induced by IL-2RB; and four out of 40 (10%) Treg downregulated genes were also inhibited by IL-2RB in Treg. We further examined whether IL2Rb regulates tissue Treg genes differentially. As shown in Table 8C, 129 out of 235 IL-2Rb induced genes were upregulated in tissue Treg including 9 genes upregulated in s-Treg, 66 genes upregulated in LN-Treg, 73 genes upregulated in int-Treg, and 69 genes upregulated in VAT-Treg. The Gene Set Enrichment Analysis (GSEA, http://www.gsea-msigdb.org/gsea/index.jsp) showed that IL2RB modulate the functions of E2F3 and Foxp3 (Figure 10A). The Venn Diagram analysis of IL2RB upregulated genes in four tissue Treg showed in Figure 10B that first, s-Treg shared all nine genes with LN-Treg including two genes specifically shared with LN-Treg, four genes specifically shared with int-Treg, and three genes specifically shared with VAT-Treg; second, LN-Treg had 17 specific genes, shared 13 genes with int-Treg, shared 18 genes with int-Treg and VAT-Treg, and shared 9 genes with VAT-Treg; and third, int-Treg had 18 specific genes, shared 17 genes with VAT-Treg; finally, VAT-Treg had 22 specific genes.

Figure 10

We further examined the expression of 1,176 cytokines and their receptors and 199 chemokines in IL2Rb KO Treg. As shown in Table 8D, 24 out of 1,176 cytokines and cytokine receptors and 8 out of 199 chemokines were induced by IL2Rb. Finally, our IPA analysis showed that, Figure 10C, IL2Rb induces 39 pathways. Taken together, these results have demonstrated that IL2Rb plays essential roles in promoting all tissue Treg shared functions and tissue Treg specific functions too.

10. Antibody bindings to T cell antigen receptor (TCR) CD3, co-stimulation receptors, and immune checkpoint receptors modulate the expression of 108 Treg signature genes.

Our recent report showed that co-signaling receptors including 14 T cell co-stimulation receptors, 10 co-inhibition receptors (immune checkpoint receptors), and 4 dual function receptors regulate T cell plasticity and immune tolerance (105). However, an important question remained whether a panel of antibodies to TCR-CD3, co-stimulation receptors, and immune checkpoint receptors regulate Treg signature gene expressions in four tissue Treg. We hypothesized that ligation with antibodies to TCR-CD3, co-stimulation receptors and immune checkpoint receptors regulate Treg signature gene expressions. As shown in Table 9 and Tables S2A and S2B, ligation with antibodies for 1 to 4 h (early time course) and 20 h (late time course) to B- and T-lymphocyte attenuator (BTLA)-early, BTLA-late, CD3-CD28-early, CD3-CD28-late, CD3-early, CD3-late, CD80-early, CD80-late, cytotoxic T-lymphocyte-associated protein 4 (CTLA4, CD152)-early, CTLA4-late, inducible T-cell costimulator (CD278, ICOS)-early, ICOS-late, programmed cell death protein 1 (PD-1, CD279)-early, PD-1-late resulted in increase of five to nine out of 68 Treg upregulated signature genes, respectively. Interestingly, these results have demonstrated that first, activation of Treg with anti-CD3 antibody ligation and anti-CD3 and anti-CD28 results in upregulation of Treg signature genes; second, antibody ligation of co-stimulation receptors ICOS and CD80 also lead to upregulation of Treg signature genes; and finally, blocking immune checkpoint receptors BTLA, CTLA-4, and PD-1 has interesting effects on Treg (61) in upregulating Treg signature genes, which are similar to that seen in antibody ligation of TCR and co-stimulation receptors.

11. Based on scRNA-Seq-identified six cluster markers in s-Treg, LN, intestine and VAT-Treg increase activated Treg cluster (clusters 1–3) markers; and decrease resting Treg cluster (clusters 4–6) markers.

Single cell RNA sequencing (scRNA-Seq), in an unbiased manner that does not rely on assumptions of cell-type identities, has revolutionized traditional cell type profiling approaches with fluorescence-conjugated antibodies (126, 127). A recent report with scRNA-Seq classified s-Treg into six clusters including S100a4^highS100a6^high cluster 1 (activated), Itgb1^high cluster 2 (activated), Dusp2^highNr4a1^highFoxp3^highIL2ra^high cluster 3 (activated), Ikzf2^highFoxp3^high cluster 4 (resting), Bach2^high cluster 5 (resting), and Satb1^highSell^high cluster 6 (resting) (84). This new classification has significantly improved our understanding on heterogeneity of s-Treg, However, heterogeneity of other LN-Treg, int-Treg and VAT-Treg remained poorly characterized. We hypothesized that LN-Treg, int-Treg and VAT-Treg have heterogeneity more than that of s-Treg. As shown in Table 10, we found 112 markers for the cluster 1, 112 markers for the cluster 2, 60 markers for the cluster 3, 59 markers for the cluster 4, 34 markers for the cluster 5, and 72 markers for the cluster 6. Since the six clusters were identified in s-Treg, as expected, s-Treg had four markers (3.57%) upregulated for the cluster 1, four markers (3.57%) upregulated for the cluster 2, zero markers changed for the cluster 3, zero markers changed for the cluster 4, zero markers changed for the cluster 5, one marker (1.39%) upregulated for the cluster 6. In addition, LN-Treg had 14 markers (12.5%) upregulated for the cluster 1, 14 markers (12.5%) upregulated for the cluster 2, two markers (3.33%) upregulated for the cluster 3, one marker (1.69%) upregulated for the cluster 4, one marker (2.94%) upregulated for the cluster 5, four markers (5.56%) upregulated for the cluster 6. Moreover, int-Treg had 12 markers (10.71%) upregulated for the cluster 1, 12 markers (10.71%) upregulated for the cluster 2, one marker (1.67%) upregulated for the cluster 3, zero marker upregulated and three markers (5.08%) downregulated for the cluster 4, one marker (2.94%) upregulated and five markers (14.71%) downregulated for the cluster 5, four markers (5.56%) upregulated and three markers (4.17%) downregulated for the cluster 6. Furthermore, VAT-Treg had 21 markers (18.75%) upregulated and six markers (5.36%) downregulated for the cluster 1, 21 markers (18.75%) upregulated and seven markers (6.25%) downregulated for the cluster 2, seven markers (11.67%) upregulated and two markers (3.33%) downregulated for the cluster 3, two markers (3.39%) upregulated and two markers (3.39%) downregulated for the cluster 4, one marker (2.94%) upregulated and four markers (11.76%) downregulated for the cluster 5, six markers (8.33%) upregulated and six markers (8.33%) downregulated for the cluster 6. Taken together, these results have demonstrated that tissue Treg increase more activated Treg cluster (clusters 1–3) markers than s-Treg; and tissue Treg decrease more resting Treg cluster (clusters 4–6) markers than s-Treg.

As shown in Supplementary Table 3A, s-Treg-specific upregulated cytokines and cytokine receptors discussed previously were expressed differentially in six clusters; and four out of seven cytokines (57.1%) were expressed in more than four clusters of s-Treg such as IL-7, SOCS1, TNF, and LIF. In addition, in Supplementary Table 3B, 11 out of 49 (22.4%) LN specific upregulated cytokines were expressed in more than four clusters. Moreover, in Supplementary Table 3C, 14 out of 44 (31.8%) intestine specific upregulated cytokines were expressed in more than four clusters. Furthermore, in Supplementary Table 3D, 19 out of 79 (24.1%) VAT specific upregulated cytokines were expressed in more than four clusters. Taken together, these results have demonstrated that first, in three tissue Treg, four clusters-shared cytokines and receptors were in 22.4 to 31.8% range; second, there were cytokines and receptors specific for one of two clusters in tissue Treg; and third, many cytokines and receptors in LN-Treg, int-Treg, and VAT-Treg were not expressed in six s-Treg clusters.

As shown in Supplementary Table 3E, s-Treg-specific upregulated TF discussed previously were expressed differentially in six clusters; and one out of two TF (50%) were expressed in more than four clusters of s-Treg such as Foxp3. In addition, in Supplementary Table 3F, 12 out of 20 (60%) LN specific upregulated TF were expressed in more than four clusters. Moreover, in Supplementary Table 3G, 17 out of 25 (68%) intestine specific upregulated TF were expressed in more than four clusters. Furthermore, in Supplementary Table 3H, 24 out of 43 (55.8%) VAT specific upregulated TF were expressed in more than four clusters. Taken together, these results have demonstrated that first, in three tissue Treg, four clusters-shared TF were in 55.8 to 68% range; second, there were TF specific for one of two clusters in tissue Treg; and third, a few TF in LN-Treg, int-Treg, and VAT-Treg were not expressed in six s-Treg clusters.

12. Four tissue Treg promote tissue repair by generating secretomes similar to that of stem cells; and sharing transcription factors AHR, ETV5, EGR1, and KLF4 with stem cells.

Treg have functions in various tissue repair (67) including promoting muscle repair (68), controlling neutrophil recruitment (70) and promoting repair after cardiac injury (69), facilitating skin epithelial stem cell differentiation (71) and wound healing (72), enhancing satellite cell expansion in muscle but blocking satellite cell differentiation (73), facilitating lung resolution (74), promoting lung epithelial cell proliferation (75) and lung injury repair via generating the growth factor amphire gulin (68), promoting myelin regeneration in central nerve system (76), and protecting kidney injury (77). However, molecular mechanisms underlying Treg promotion of tissue repair remained poorly defined. We hypothesized that tissue Treg promote tissue repair by generating secretome similar to that of stem cells (128). To examine this hypothesis, we collected four different stem cell (SC) secretomes including human embryonic SC secretome (hESC) with 129 proteins (129), human mesenchymal SC (hMSC) secretome with 51 proteins (130), human visceral adipose SC (hASC) secretome with 182 proteins (131), and human bone marrow SC (hBMSC) secretome with 315 proteins (132) for comparison with our results of tissue Treg secretome and cytokines (Table 11). As shown in Figure 11A, we made comparison of s-Treg upregulated cytokines and cytokine receptors with seven proteins, s-Treg upregulated secretome with eight proteins, LN-Treg upregulated cytokines and cytokine receptors with 49 proteins, LN-Treg upregulated secretome with 31 proteins, int-Treg upregulated cytokines and cytokine receptors with 44 proteins, int-Treg upregulated secretome with 37 proteins, VAT-Treg upregulated cytokines and cytokine receptors with 79 proteins, and VAT-Treg upregulated secretome with 51 proteins to stem cell secretomes, respectively. The Venn Diagram results were shown in Table 12: 1) s-Treg shared one cytokine out of seven (14.3%), LIF, with hMSC secretome and hASC secretome (Supplementary Figure 9A); 2) LN-Treg secretome shared five out of 31 (16.1%) secretory proteins with SC secretomes including PENK with hASC secretome, sharing LAMC1 with hBMSC secretome, sharing ECM1 and CD44 with secretomes of hBMSC and hASC. In addition, LN-Treg cytokines shared one cytokine, CRLF1, out of 49 upregulated cytokines (2.0%) with hASC secretome (Supplementary Figure 9B); 3) int-Treg shared four out of 37 (11.0%) secretory proteins with SC secretomes including sharing PENK with hASC, sharing ECM1 with hBMSC and hhASC secretomes, and sharing DKK3 and LAMC1 with hBMSC secretome (Supplementary Figure 9C); 4) VAT-Treg shared 12 out of 51 secretomic proteins (23.5%) with SC secretomes including sharing PENK, CXCL2, and TFPI with hASC, sharing LIF with hASC and hMSC secretomes, sharing IGFBP7, CST3, LGALS1, CD44 with hBMSC and hASC secretomes; sharing LAMC1, ANXA1, LGALS3, GRN with hBMSC secretome. In addition, VAT-Treg cytokines shared five cytokines out of 79 upregulated cytokines (6.3%) with SC secretomes including sharing MIF and TIMP1 with hBMSC and hASC secretomes and sharing CRLF1, CCL2, and CXCL6 with hASC secretome (Supplementary Figure 9D). Taken together, these results have demonstrated that first, tissue Treg share secretomes with stem cell secretomes in the ranges of 11.0 to 23.5%; second, LN-Treg and VAT-Treg share cytokines with stem cell secretomes in the ranges of 2.0 to 6.3%, but s-Treg and int-Treg do not share cytokines with stem cell secretomes, suggesting that Treg secretomes are more than Treg cytokines in sharing with stem cell secretomes; and third, for comparison, s-Treg share two out of upregulated eight secretomic genes (25% for s-Treg) with 51 upregulated VAT-Treg secretomic genes (3.9% for VAT-Treg), suggesting that the secretomic gene percentages shared by s-Treg and VAT-Treg are much smaller than that shared by four tissue Treg and stem cells, suggesting that tissue Treg play significant stem cell-like roles for tissue repair and regeneration.

Figure 11

To further consolidate the transcription regulatory mechanisms for tissue Treg in tissue repair, we then examined a new hypothesis that tissue Treg TFs share stem cell TFs. To test this hypothesis, we performed Venn Diagram analyses with a comparison of upregulated tissue Treg TFs including s-Treg (2 TFs), LN-Treg (20 TFs), int-Treg (25 TFs), and VAT-Treg (43 TFs) to 49 hematopoietic stem cell (HSC) TFs, 41 mesenchymal stem cell (MSC) TFs, and 61 pluripotent stem cell (PSC) TFs (133) (Table 13). As shown in Table 14, s-Treg did not share any upregulated TFs with stem cell TFs (Supplementary Figure 10A); LN TFs shared aryl hydrocarbon receptor (AHR) and E26 transformation-specific (ETS) family TF (ETV5) with MSC TFs (Supplementary Figure 10B); int-Treg shared one TF, ETV5, with MSC (Supplementary Figure 10C); VAT-Treg shared TF early growth response protein 1 (EGR1, zinc finger protein 268) with HSC, and shared one TF, Krüppel-like factor 4 (KLF4), with PSC (Supplementary Figure 10D). AHR plays a significant role in hematopoietic stem cell transcriptome regulation (134). ETV5 plays a critical role in maintaining alveolar type II cells (135), controlling cell type specification in developing mouse brain (136), and having versatile functions in male reproduction (137). EGR1 directs tendon differentiation, promotes tendon repair (138) and blocks energy expenditure via direct uncoupling protein 1 (UCP1) transcription repression and counteracts obesity (139). KLF4 is one of 2012 Nobel Laureate Yamanaka’s four key stem cell TFs (140) and a reprogramming factor, and plays an essential role for stem cell maintenance and myeloid and lymphoid cell developments (141). To further determine the causative effects that these four TFs are partially responsible for upregulating Treg genes, as shown in Table 15, deficiencies or decreased expressions of the four TFs resulted in partially downregulation of some Treg genes identified in previous eight Results sections. These results suggest that these four Treg-stem cell-shared TFs at least partially promote upregulation of all the eight groups of Treg genes including Treg signature genes, CD markers, cytokines, secretome, TFs, cell death regulators, kinome, and Treg cluster markers. Taken together, our analyses have demonstrated that first, tissue Treg both from lymphoid tissue (LN-Treg) and non-lymphoid tissues (int-Treg and VAT-Treg) partially share stem cell TFs; and second, tissue Treg upregulated TFs would make Treg play roles in regenerating lung alveolar type II cells, promoting brain cell specification, facilitating male reproduction, promoting tendon differentiation and repair, blocking energy expenditure and inhibiting obesity, and maintaining stem cell and blood cell developments.

Discussion

Treg have been under intensified investigation continuously (51) since being first reported in 1995 (142). Treg are also a new therapeutic target for numerous diseases (143) including cardiovascular disease (144), monogenic disease [immunodysregulation polyendocrinopathy enteropathy X-linked (or IPEX) syndrome], systemic lupus erythematosus, organ-specific autoimmune diseases (type I diabetes, psoriasis, myasthenia, inflammatory bowel disease, and multiple sclerosis), transplantation, and cancers (143, 145). Although many wonderful technological advances including single-cell RNA sequencings (84) have been made to profile lymphoid and non-lymphoid tissue Treg heterogeneity, some important functional issues remained poorly characterized including upregulated signature genes and signaling pathways, CD markers, cytokines and secretomes, TFs, cell death regulatomes, activation and resting status, and Treg similarity of secretomes and TFs to that of stem cells. To address those issues, we performed a comprehensive transcriptomic database mining with the strategies we pioneered (10, 43, 146, 147) to compare four tissue Treg including two lymphoid tissues s-Treg and LN-Treg and two non-lymphoid tissues such as int-Treg and VAT-Treg, and made significant findings.

Our recent paper reported the use of a new function -omics angle to determine the transcriptomic changes of all the human genome-encoded cytokines and secretomes in human peripheral blood mononuclear cells (PBMCs) in patients with chronic kidney disease and end-stage renal disease (10). General transcriptomic analysis of microarrays, RNA-Seq and single-cell RNA-Seq emphasize global transcriptomic profiling regardless the functions of transcripts. Our previous report applied a function angle to examine cytokine changes using a cytokine array in the aorta in atherogenic apolipoprotein E deficient (ApoE−/−) mice in the presence or absence of caspase-1, a key regulator of inflammatory cell death (pyroptosis) (1). In this study, we attempted to multiple function -omics angles plus differentially expressed genes to profile tissue Treg.

Based on our results, we propose a new working model. First, as shown in Figure 11A including differentially expressed genes, and seven function -omics angles such as: 1) Treg signature (master regulators and signaling), 2) pathway analysis (signaling), 3) human genome-encoded total (HGET) 1,176 cytokines and their receptors (effectors), 4) 1,706 secretomes (HGET secretory proteins) (effectors), 5) total 373 CD markers (cell surface receptors for extracellular signals, cell-cell contact effectors, and signaling initiators), 6) HGET 1,496 transcription factors (nuclear master regulators), 7) HGET 661 kinases (signaling), and 8) 305 cell death regulators functional in 13 newly formulated types of cell death to profile both lymphoid tissue Treg (s-Treg and LN-Treg) and non-lymphoid tissue Treg (int-Treg and VAT-Treg). Second, these multiple function -omics angles profile from cell surface, intracellular signaling pathways, to nuclear master regulators-transcription factors, and from cell-cell-contact effectors (CD markers) including 28 T cell co-stimulation receptors and immune checkpoint receptors (105) to secretory protein effectors (cytokines and secretomes) to maintain Treg transcriptomic signatures and functional signatures. It has been well documented that stem cell signatures are termed by their stemness with self-renewal capacity and differentiation potential (148), which are maintained by the membranes markers CD73 (5’-nucleotidase to convert AMP to adenosine), CD90 (cell-cell and cell-matrix interactions), and CD105 (endoglin, an accessory receptor for TGF-β), as well as the stemness genes homeobox transcription factor (NANOG), octamer-binding transcription factor 4 (OCT4), sex determining region Y (SRY)-box 2 transcription factor (SOX2), C2H2 zinc-finger transcription factor (REX1, Zfp-42), cell fate controlling membrane receptor NOTCH1 and, type IV intermediate filament protein NESTIN (128, 149). Similarly, we proposed a new concept of Treg-ness for Treg identity maintenance with dual functions of immunosuppression and tissue repair. As demonstrated previously, Treg compartmentalization and trafficking are tissue and organ-specific potentially mediated by distinct chemokine receptors and integrins (150). Similar to stem cells, in order for Treg to maintain their Treg-ness, tissue Treg generate stem cell-shared secretomes and transcription factors to establish a special microenvironment, Treg niche (151, 152), to maintain Treg-ness and suppress Treg plasticity (29, 105), by which tissue Treg can maintain their Treg-ness with dual functions of immunosuppression and tissue repair including promoting stem cell maintenance (153–156) (Figure 11B). As shown in Figure 11C, our results have demonstrated further that 1) stem cell-shared secretomes and transcription factors make tissue CD4⁺Foxp3⁺ Treg as first T cell type use a Treg niche to maintain their Treg-ness with dual functions of immunosuppression and tissue repair; 2) lymphoid tissue Treg share Treg niche and immunosuppressive functions with non-lymphoid tissue Treg; and 3) non-lymphoid Treg develop more stem cell promoting and tissue repair functions than lymphoid tissue Treg. Our findings have provided novel insights on tissue Treg heterogeneity and new therapeutic targets for immunosuppression, tissue repair, cardiovascular diseases, chronic kidney disease, autoimmune diseases, transplantation, and cancers.

One limitation of the current study is that due to the low throughput nature of verification techniques in the laboratories, we could not verify every result we identified with the analyses of high throughput data [see Table 1 of Dr. Lai’s paper (106), and Table 10 of Dr. Zhang’s paper (10) for explanations]. We acknowledge that carefully designed in-vitro and in-vivo experimental models will be needed to verify all the findings further and underlying mechanisms. Nevertheless, our findings provide novel insights on the roles of tissue Treg in controlling immune responses, and promoting tissue repair and regeneration as well as novel targets for the future therapeutic interventions for immunosuppression, cardiovascular diseases, autoimmune diseases, transplantation, cancers, and tissue repair.

Funding

This work was supported by the hospital fellowship to RZ.

References

• 1

  YinYLiXShaXXiHLiYFShaoYet al. Early hyperlipidemia promotes endothelial activation via a caspase-1-sirtuin 1 pathway. Arterioscler Thromb Vasc Biol (2015) 35(4):804–16. doi: 10.1161/ATVBAHA.115.305282

  • CrossRef
  • Google Scholar

• 2

  LiXFangPLiYKuoYMAndrewsAJNanayakkaraGet al. Mitochondrial Reactive Oxygen Species Mediate Lysophosphatidylcholine-Induced Endothelial Cell Activation. Arterioscler Thromb Vasc Biol (2016) 36(6):1090–100. doi: 10.1161/ATVBAHA.115.306964

  • CrossRef
  • Google Scholar

• 3

  FangPZhangDChengZYanCJiangXKrugerWDet al. Hyperhomocysteinemia potentiates hyperglycemia-induced inflammatory monocyte differentiation and atherosclerosis. Diabetes (2014) 63(12):4275–90. doi: 10.2337/db14-0809

  • CrossRef
  • Google Scholar

• 4

  FangPLiXShanHSaredyJJCuetoRXiaJet al. Ly6C(+) Inflammatory Monocyte Differentiation Partially Mediates Hyperhomocysteinemia-Induced Vascular Dysfunction in Type 2 Diabetic db/db Mice. Arterioscler Thromb Vasc Biol (2019) 39(10):2097–119. doi: 10.1161/ATVBAHA.119.313138

  • CrossRef
  • Google Scholar

• 5

  XiHZhangYXuYYangWYJiangXShaXet al. Caspase-1 Inflammasome Activation Mediates Homocysteine-Induced Pyrop-Apoptosis in Endothelial Cells. Circ Res (2016) 118(10):1525–39. doi: 10.1161/CIRCRESAHA.116.308501

  • CrossRef
  • Google Scholar

• 6

  YangJFangPYuDZhangLZhangDJiangXet al. Chronic Kidney Disease Induces Inflammatory CD40+ Monocyte Differentiation via Homocysteine Elevation and DNA Hypomethylation. Circ Res (2016) 119(11):1226–41. doi: 10.1161/CIRCRESAHA.116.308750

  • CrossRef
  • Google Scholar

• 7

  FerrerLMMonroyAMLopez-PastranaJNanayakkaraGCuetoRLiYFet al. Caspase-1 Plays a Critical Role in Accelerating Chronic Kidney Disease-Promoted Neointimal Hyperplasia in the Carotid Artery. J Cardiovasc Transl Res (2016) 9(2):135–44. doi: 10.1007/s12265-016-9683-3

  • CrossRef
  • Google Scholar

• 8

  MonroyMAFangJLiSFerrerLBirkenbachMPLeeIJet al. Chronic kidney disease alters vascular smooth muscle cell phenotype. Front Biosci (Landmark Ed) (2015) 20:784–95. doi: 10.2741/4337

  • CrossRef
  • Google Scholar

• 9

  SunYJohnsonCZhouJWangLLiYFLuYet al. Uremic toxins are conditional danger- or homeostasis-associated molecular patterns. Front Biosci (Landmark Ed) (2018) 23:348–87. doi: 10.2741/4595

  • CrossRef
  • Google Scholar

• 10

  ZhangRSaredyJShaoYYaoTLiuLSaaoudFet al. End-stage renal disease is different from chronic kidney disease in upregulating ROS-modulated proinflammatory secretome in PBMCs - A novel multiple-hit model for disease progression. Redox Biol (2020) 34:101460. doi: 10.1016/j.redox.2020.101460

  • CrossRef
  • Google Scholar

• 11

  ChanMMYangXWangHSaaoudFSunYFongD. The Microbial Metabolite Trimethylamine N-Oxide Links Vascular Dysfunctions and the Autoimmune Disease Rheumatoid Arthritis. Nutrients (2019) 11(8):1108–821. doi: 10.3390/nu11081821

  • CrossRef
  • Google Scholar

• 12

  ShaXMengSLiXXiHMaddaloniMPascualDWet al. Interleukin-35 Inhibits Endothelial Cell Activation by Suppressing MAPK-AP-1 Pathway. J Biol Chem (2015) 290(31):19307–18. doi: 10.1074/jbc.M115.663286

  • CrossRef
  • Google Scholar

• 13

  ShaoYChengZLiXChernayaVWangHYangXF. Immunosuppressive/anti-inflammatory cytokines directly and indirectly inhibit endothelial dysfunction–a novel mechanism for maintaining vascular function. J Hematol Oncol (2014) 7:80. doi: 10.1186/s13045-014-0080-6

  • CrossRef
  • Google Scholar

• 14

  LiXWangLFangPSunYJiangXWangHet al. Lysophospholipids induce innate immune transdifferentiation of endothelial cells, resulting in prolonged endothelial activation. J Biol Chem (2018) 293:jbc.RA118.002752. doi: 10.1074/jbc.RA118.002752

  • CrossRef
  • Google Scholar

• 15

  LiASunYDrummerCTLuYYuDZhouYet al. Increasing Upstream Chromatin Long-Range Interactions May Favor Induction of Circular RNAs in LysoPC-Activated Human Aortic Endothelial Cells. Front Physiol (2019) 10:433. doi: 10.3389/fphys.2019.00433

  • CrossRef
  • Google Scholar

• 16

  Lopez-PastranaJFerrerLMLiYFXiongXXiHCuetoRet al. Inhibition of Caspase-1 Activation in Endothelial Cells Improves Angiogenesis: A NOVEL THERAPEUTIC POTENTIAL FOR ISCHEMIA. J Biol Chem (2015) 290(28):17485–94. doi: 10.1074/jbc.M115.641191

  • CrossRef
  • Google Scholar

• 17

  LiYFHuangXLiXGongRYinYNelsonJet al. Caspase-1 mediates hyperlipidemia-weakened progenitor cell vessel repair. Front Biosci (Landmark Ed) (2016) 21:178–91. doi: 10.2741/4383

  • CrossRef
  • Google Scholar

• 18

  LiXShaoYShaXFangPKuoYMAndrewsAJet al. IL-35 (Interleukin-35) Suppresses Endothelial Cell Activation by Inhibiting Mitochondrial Reactive Oxygen Species-Mediated Site-Specific Acetylation of H3K14 (Histone 3 Lysine 14). Arterioscler Thromb Vasc Biol (2018) 38(3):599–609. doi: 10.1161/ATVBAHA.117.310626

  • CrossRef
  • Google Scholar

• 19

  LiXFangPSunYShaoYYangWYJiangXet al. Anti-inflammatory cytokines IL-35 and IL-10 block atherogenic lysophosphatidylcholine-induced, mitochondrial ROS-mediated innate immune activation, but spare innate immune memory signature in endothelial cells. Redox Biol (2020) 28:101373. doi: 10.1016/j.redox.2019.101373

  • CrossRef
  • Google Scholar

• 20

  ZhangCEWeiWLiuYHPengJHTianQLiuGPet al. Hyperhomocysteinemia increases beta-amyloid by enhancing expression of gamma-secretase and phosphorylation of amyloid precursor protein in rat brain. Am J Pathol (2009) 174(4):1481–91. doi: 10.2353/ajpath.2009.081036

  • CrossRef
  • Google Scholar

• 21

  ZhangDFangPJiangXNelsonJMooreJKKrugerWDet al. Severe hyperhomocysteinemia promotes bone marrow-derived and resident inflammatory monocyte differentiation and atherosclerosis in LDLr/CBS-deficient mice. Circ Res (2012) 111(1):37–49. doi: 10.1161/CIRCRESAHA.112.269472

  • CrossRef
  • Google Scholar

• 22

  NelsonJWuYJiangXBerrettaRHouserSChoiEet al. Hyperhomocysteinemia suppresses bone marrow CD34+/VEGF receptor 2+ cells and inhibits progenitor cell mobilization and homing to injured vasculature-a role of beta1-integrin in progenitor cell migration and adhesion. FASEB J Off Publ Fed Am Soc Exp Biol (2015) 29(7):3085–99. doi: 10.1096/fj.14-267989

  • CrossRef
  • Google Scholar

• 23

  ShaoYChernayaVJohnsonCYangWYCuetoRShaXet al. Metabolic Diseases Downregulate the Majority of Histone Modification Enzymes, Making a Few Upregulated Enzymes Novel Therapeutic Targets-“Sand Out and Gold Stays”. J Cardiovasc Trans Res (2016) 9(1):49–66. doi: 10.1007/s12265-015-9664-y

  • CrossRef
  • Google Scholar

• 24

  LuYSunYDrummerCTNanayakkaraGKShaoYSaaoudFet al. Increased acetylation of H3K14 in the genomic regions that encode trained immunity enzymes in lysophosphatidylcholine-activated human aortic endothelial cells - Novel qualification markers for chronic disease risk factors and conditional DAMPs. Redox Biol (2019) 24:101221. doi: 10.1016/j.redox.2019.101221

  • CrossRef
  • Google Scholar

• 25

  ZhongCYangXFengYYuJ. Trained Immunity: An Underlying Driver of Inflammatory Atherosclerosis. Front Immunol (2020) 11:284. doi: 10.3389/fimmu.2020.00284

  • CrossRef
  • Google Scholar

• 26

  XiongZSongJYanYHuangYCowanAWangHet al. Higher expression of Bax in regulatory T cells increases vascular inflammation. Front Biosci (2008) 13:7143–55. doi: 10.2741/3217

  • CrossRef
  • Google Scholar

• 27

  XiongZYanYSongJFangPYinYYangYet al. Expression of TCTP antisense in CD25(high) regulatory T cells aggravates cuff-injured vascular inflammation. Atherosclerosis (2009) 203(2):401–8. doi: 10.1016/j.atherosclerosis.2008.07.041

  • CrossRef
  • Google Scholar

• 28

  YangWYShaoYLopez-PastranaJMaiJWangHYangXF. Pathological conditions re-shape physiological Tregs into pathological Tregs. Burns Trauma (2015) 3(1):1–11. doi: 10.1186/s41038-015-0001-0

  • CrossRef
  • Google Scholar

• 29

  XuKYangWYNanayakkaraGKShaoYYangFHuWet al. gaTa3, hDac6, and Bcl6 regulate FOXP3+ Treg Plasticity and Determine Treg conversion into either novel antigen-Presenting cell-like Treg or Th1-Treg. Front Immunol (2018) 9:45. doi: 10.3389/fimmu.2018.00045

  • CrossRef
  • Google Scholar

• 30

  WebbLMCLintermanMA. Signals that drive T follicular helper cell formation. Immunology (2017) 152(2):185–94. doi: 10.1111/imm.12778

  • CrossRef
  • Google Scholar

• 31

  GowthamanUChenJSZhangBFlynnWFLuYSongWet al. Identification of a T follicular helper cell subset that drives anaphylactic IgE. Science (2019) 365(6456):eaaw6433. doi: 10.1126/science.aaw6433

  • CrossRef
  • Google Scholar

• 32

  DuPageMBluestoneJA. Harnessing the plasticity of CD4(+) T cells to treat immune-mediated disease. Nat Rev Immunol (2016) 16(3):149–63. doi: 10.1038/nri.2015.18

  • CrossRef
  • Google Scholar

• 33

  PawankarRHayashiMYamanishiSIgarashiT. The paradigm of cytokine networks in allergic airway inflammation. Curr Opin Allergy Clin Immunol (2015) 15(1):41–8. doi: 10.1097/ACI.0000000000000129

  • CrossRef
  • Google Scholar

• 34

  TakeuchiASaitoT. CD4 CTL, a Cytotoxic Subset of CD4+ T Cells, Their Differentiation and Function. Front Immunol (2017) 8:194. doi: 10.3389/fimmu.2017.00194

  • CrossRef
  • Google Scholar

• 35

  FarberDL. Form and function for T cells in health and disease. Nat Rev Immunol (2020) 20(2):83–4. doi: 10.1038/s41577-019-0267-8

  • CrossRef
  • Google Scholar

• 36

  SaigusaRWinkelsHLeyK. T cell subsets and functions in atherosclerosis. Nat Rev Cardiol (2020) 17:387–401. doi: 10.1038/s41569-020-0352-5

  • CrossRef
  • Google Scholar

• 37

  MirlekarB. Co-expression of master transcription factors determines CD4(+) T cell plasticity and functions in auto-inflammatory diseases. Immunol Lett (2020) 222:58–66. doi: 10.1016/j.imlet.2020.03.007

  • CrossRef
  • Google Scholar

• 38

  KeXWangJLiLChenIHWangHYangXF. Roles of CD4+CD25(high) FOXP3+ Tregs in lymphomas and tumors are complex. Front Biosci (2008) 13:3986–4001. doi: 10.2741/2986

  • CrossRef
  • Google Scholar

• 39

  PastranaJLShaXVirtueAMaiJCuetoRLeeIAet al. Regulatory T cells and Atherosclerosis. J Clin Exp Cardiolog (2012) 2012(Suppl 12):2. doi: 10.4172/2155-9880.S12-002

  • CrossRef
  • Google Scholar

• 40

  SharirRSemoJShaishALanda-RoubenNEntin-MeerMKerenGet al. Regulatory T cells influence blood flow recovery in experimental hindlimb ischaemia in an IL-10-dependent manner. Cardiovasc Res (2014) 103(4):585–96. doi: 10.1093/cvr/cvu159

  • CrossRef
  • Google Scholar

• 41

  BurzynDBenoistCMathisD. Regulatory T cells in nonlymphoid tissues. Nat Immunol (2013) 14(10):1007–13. doi: 10.1038/ni.2683

  • CrossRef
  • Google Scholar

• 42

  DiSpiritoJRZemmourDRamananDChoJZilionisRKleinAMet al. Molecular diversification of regulatory T cells in nonlymphoid tissues. Sci Immunol (2018) 3(27):eaat5861. doi: 10.1126/sciimmunol.aat5861

  • CrossRef
  • Google Scholar

• 43

  LiXMaiJVirtueAYinYGongRShaXet al. IL-35 is a novel responsive anti-inflammatory cytokine–a new system of categorizing anti-inflammatory cytokines. PloS One (2012) 7(3):e33628. doi: 10.1371/journal.pone.0033628

  • CrossRef
  • Google Scholar

• 44

  LiXFangPYangWYWangHYangX. IL-35, as a newly proposed homeostasis-associated molecular pattern, plays three major functions including anti-inflammatory initiator, effector, and blocker in cardiovascular diseases. Cytokine (2017) 122:154076. doi: 10.1016/j.cyto.2017.06.003

  • CrossRef
  • Google Scholar

• 45

  RaffinCVoLTBluestoneJA. Treg cell-based therapies: challenges and perspectives. Nat Rev Immunol (2020) 20(3):158–72. doi: 10.1038/s41577-019-0232-6

  • CrossRef
  • Google Scholar

• 46

  BrakchNYangXFCrinePCohenPBoileauG. Predominant basolateral proteolytic processing of prosomatostatin into somatostatin-28 in polarized LLC-PK1 cells. Neuropeptides (1997) 31(5):393–8. doi: 10.1016/S0143-4179(97)90030-5

  • CrossRef
  • Google Scholar

• 47

  YangYYangFXiongZYanYWangXNishinoMet al. An N-terminal region of translationally controlled tumor protein is required for its antiapoptotic activity. Oncogene (2005) 24(30):4778–88. doi: 10.1038/sj.onc.1208666

  • CrossRef
  • Google Scholar

• 48

  YangXF. Factors regulating apoptosis and homeostasis of CD4+ CD25(high) FOXP3+ regulatory T cells are new therapeutic targets. Front Biosci (2008) 13:1472–99. doi: 10.2741/2775

  • CrossRef
  • Google Scholar

• 49

  YangXFFangPMengSJanMXiongXYinYet al. The FOX transcription factors regulate vascular pathology, diabetes and Tregs. Front Biosci (Schol Ed) (2009) 1:420–36. doi: 10.2741/s35

  • CrossRef
  • Google Scholar

• 50

  YangXFFangPMengSJanMXiongXYinYet al. The forkhead transcription factors play important roles in vascular pathology and immunology. Adv Exp Med Biol (2009) 665:90–105. doi: 10.1007/978-1-4419-1599-3_7

  • CrossRef
  • Google Scholar

• 51

  YangXFYinYWangH. VASCULAR INFLAMMATION AND ATHEROGENESIS ARE ACTIVATED VIA RECEPTORS FOR PAMPs AND SUPPRESSED BY REGULATORY T CELLS. Drug Discovery Today Ther Strateg (2008) 5(2):125–42. doi: 10.1016/j.ddstr.2008.11.003

  • CrossRef
  • Google Scholar

• 52

  YanYXiongZZhangSSongJHuangYThorntonAMet al. CD25high T cells with a prolonged survival inhibit development of diabetes. Int J Immunopathol Pharmacol (2008) 21(4):767–80. doi: 10.1177/039463200802100401

  • CrossRef
  • Google Scholar

• 53

  Lopez-PastranaJShaoYChernayaVWangHYangXF. Epigenetic enzymes are the therapeutic targets for CD4(+)CD25(+/high)Foxp3(+) regulatory T cells. Transl Res (2015) 165(1):221–40. doi: 10.1016/j.trsl.2014.08.001

  • CrossRef
  • Google Scholar

• 54

  KitzADominguez-VillarM. Molecular mechanisms underlying Th1-like Treg generation and function. Cell Mol Life Sci (2017) 74(22):4059–75. doi: 10.1007/s00018-017-2569-y

  • CrossRef
  • Google Scholar

• 55

  ButcherMJFilipowiczARWaseemTCMcGaryCMCrowKJMagilnickNet al. Atherosclerosis-Driven Treg Plasticity Results in Formation of a Dysfunctional Subset of Plastic IFNgamma+ Th1/Tregs. Circ Res (2016) 119(11):1190–203. doi: 10.1161/CIRCRESAHA.116.309764

  • CrossRef
  • Google Scholar

• 56

  PandiyanPZhuJ. Origin and functions of pro-inflammatory cytokine producing Foxp3+ regulatory T cells. Cytokine (2015) 76(1):13–24. doi: 10.1016/j.cyto.2015.07.005

  • CrossRef
  • Google Scholar

• 57

  SharirRSemoJShimoniSBen-MordechaiTLanda-RoubenNMaysel-AuslenderSet al. Experimental myocardial infarction induces altered regulatory T cell hemostasis, and adoptive transfer attenuates subsequent remodeling. PloS One (2014) 9(12):e113653. doi: 10.1371/journal.pone.0113653

  • CrossRef
  • Google Scholar

• 58

  AndersenMH. Anti-regulatory T cells. Semin Immunopathol (2017) 39(3):317–26. doi: 10.1007/s00281-016-0593-x

  • CrossRef
  • Google Scholar

• 59

  LarsenSK. Cellular immune responses towards regulatory cells. Dan Med J (2016) 63(1):B5188.

  • Google Scholar

• 60

  LevineAGMedozaAHemmersSMoltedoBNiecRESchizasMet al. Stability and function of regulatory T cells expressing the transcription factor T-bet. Nature (2017) 546(7658):421–5. doi: 10.1038/nature22360

  • CrossRef
  • Google Scholar

• 61

  LuccaLEDominguez-VillarM. Modulation of regulatory T cell function and stability by co-inhibitory receptors. Nat Rev Immunol (2020) 20(11):680–93. doi: 10.1038/s41577-020-0296-3

  • CrossRef
  • Google Scholar

• 62

  FarhanHRabouilleC. Signalling to and from the secretory pathway. J Cell Sci (2011) 124(Pt 2):171–80. doi: 10.1242/jcs.076455

  • CrossRef
  • Google Scholar

• 63

  RabouilleC. Pathways of Unconventional Protein Secretion. Trends Cell Biol (2017) 27(3):230–40. doi: 10.1016/j.tcb.2016.11.007

  • CrossRef
  • Google Scholar

• 64

  PlanavilaAFernandez-SolaJVillarroyaF. Cardiokines as Modulators of Stress-Induced Cardiac Disorders. Adv Protein Chem Struct Biol (2017) 108:227–56. doi: 10.1016/bs.apcsb.2017.01.002

  • CrossRef
  • Google Scholar

• 65

  LipphardtMSongJWMatsumotoKDadafarinSDihaziHMullerGet al. The third path of tubulointerstitial fibrosis: aberrant endothelial secretome. Kidney Int (2017) 92(3):558–68. doi: 10.1016/j.kint.2017.02.033

  • CrossRef
  • Google Scholar

• 66

  MakridakisMRoubelakisMGVlahouA. Stem cells: insights into the secretome. Biochim Biophys Acta (2013) 1834(11):2380–4. doi: 10.1016/j.bbapap.2013.01.032

  • CrossRef
  • Google Scholar

• 67

  LiJTanJMartinoMMLuiKO. Regulatory T-Cells: Potential Regulator of Tissue Repair and Regeneration. Front Immunol (2018) 9:585. doi: 10.3389/fimmu.2018.00585

  • CrossRef
  • Google Scholar

• 68

  BurzynDKuswantoWKolodinDShadrachJLCerlettiMJangYet al. A special population of regulatory T cells potentiates muscle repair. Cell (2013) 155(6):1282–95. doi: 10.1016/j.cell.2013.10.054

  • CrossRef
  • Google Scholar

• 69

  WeiratherJHofmannUDBeyersdorfNRamosGCVogelBFreyAet al. Foxp3+ CD4+ T cells improve healing after myocardial infarction by modulating monocyte/macrophage differentiation. Circ Res (2014) 115(1):55–67. doi: 10.1161/CIRCRESAHA.115.303895

  • CrossRef
  • Google Scholar

• 70

  CarboneFNencioniAMachFVuilleumierNMontecuccoF. Pathophysiological role of neutrophils in acute myocardial infarction. Thromb Haemost (2013) 110(3):501–14. doi: 10.1160/TH13-03-0211

  • CrossRef
  • Google Scholar

• 71

  AliNZirakBRodriguezRSPauliMLTruongHALaiKet al. Regulatory T Cells in Skin Facilitate Epithelial Stem Cell Differentiation. Cell (2017) 169(6):1119–29 e11. doi: 10.1016/j.cell.2017.05.002

  • CrossRef
  • Google Scholar

• 72

  NosbaumAPrevelNTruongHAMehtaPEttingerMScharschmidtTCet al. Cutting Edge: Regulatory T Cells Facilitate Cutaneous Wound Healing. J Immunol (2016) 196(5):2010–4. doi: 10.4049/jimmunol.1502139

  • CrossRef
  • Google Scholar

• 73

  CastiglioniACornaGRigamontiEBassoVVezzoliMMonnoAet al. FOXP3+ T Cells Recruited to Sites of Sterile Skeletal Muscle Injury Regulate the Fate of Satellite Cells and Guide Effective Tissue Regeneration. PloS One (2015) 10(6):e0128094. doi: 10.1371/journal.pone.0128094

  • CrossRef
  • Google Scholar

• 74

  D’AlessioFRTsushimaKAggarwalNRWestEEWillettMHBritosMFet al. CD4+CD25+Foxp3+ Tregs resolve experimental lung injury in mice and are present in humans with acute lung injury. J Clin Invest (2009) 119(10):2898–913. doi: 10.1172/JCI36498

  • CrossRef
  • Google Scholar

• 75

  MockJRGaribaldiBTAggarwalNRJenkinsJLimjunyawongNSingerBDet al. Foxp3+ regulatory T cells promote lung epithelial proliferation. Mucosal Immunol (2014) 7(6):1440–51. doi: 10.1038/mi.2014.33

  • CrossRef
  • Google Scholar

• 76

  DombrowskiYO’HaganTDittmerMPenalvaRMayoralSRBankheadPet al. Regulatory T cells promote myelin regeneration in the central nervous system. Nat Neurosci (2017) 20(5):674–80. doi: 10.1038/nn.4528

  • CrossRef
  • Google Scholar

• 77

  VenetFPachotADebardALBoheJBienvenuJLepapeAet al. Human CD4+CD25+ regulatory T lymphocytes inhibit lipopolysaccharide-induced monocyte survival through a Fas/Fas ligand-dependent mechanism. J Immunol (2006) 177(9):6540–7. doi: 10.4049/jimmunol.177.9.6540

  • CrossRef
  • Google Scholar

• 78

  LiYFNanayakkaraGSunYLiXWangLCuetoRet al. Analyses of caspase-1-regulated transcriptomes in various tissues lead to identification of novel IL-1beta-, IL-18- and sirtuin-1-independent pathways. J Hematol Oncol (2017) 10(1):40. doi: 10.1186/s13045-017-0406-2

  • CrossRef
  • Google Scholar

• 79

  EisenbergELevanonEY. Human housekeeping genes, revisited. Trends Genet (2013) 29(10):569–74. doi: 10.1016/j.tig.2013.05.010

  • CrossRef
  • Google Scholar

• 80

  WangLFuHNanayakkaraGLiYShaoYJohnsonCet al. Novel extracellular and nuclear caspase-1 and inflammasomes propagate inflammation and regulate gene expression: a comprehensive database mining study. J Hematol Oncol (2016) 9(1):122. doi: 10.1186/s13045-016-0351-5

  • CrossRef
  • Google Scholar

• 81

  RoggeLBianchiEBiffiMBonoEChangSYAlexanderHet al. Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat Genet (2000) 25(1):96–101. doi: 10.1038/75671

  • CrossRef
  • Google Scholar

• 82

  DuguetFLocard-PauletMMarcellinMChaouiKBernardIAndreolettiOet al. Proteomic Analysis of Regulatory T Cells Reveals the Importance of Themis1 in the Control of Their Suppressive Function. Mol Cell Proteomics (2017) 16(8):1416–32. doi: 10.1074/mcp.M116.062745

  • CrossRef
  • Google Scholar

• 83

  ValzasinaBGuiducciCDislichHKilleenNWeinbergADColomboMP. Triggering of OX40 (CD134) on CD4(+)CD25+ T cells blocks their inhibitory activity: a novel regulatory role for OX40 and its comparison with GITR. Blood (2005) 105(7):2845–51. doi: 10.1182/blood-2004-07-2959

  • CrossRef
  • Google Scholar

• 84

  ZemmourDZilionisRKinerEKleinAMMathisDBenoistC. Single-cell gene expression reveals a landscape of regulatory T cell phenotypes shaped by the TCR. Nat Immunol (2018) 19(3):291–301. doi: 10.1038/s41590-018-0051-0

  • CrossRef
  • Google Scholar

• 85

  SengAYankeeTM. The Role of the Ikaros Family of Transcription Factors in Regulatory T cell Development and Function. J Clin Cell Immunol (2017) 8(2):495. doi: 10.4172/2155-9899.1000495

  • CrossRef
  • Google Scholar

• 86

  YangSWangJBrandDDZhengSG. Role of TNF-TNF Receptor 2 Signal in Regulatory T Cells and Its Therapeutic Implications. Front Immunol (2018) 9:784. doi: 10.3389/fimmu.2018.00784

  • CrossRef
  • Google Scholar

• 87

  YanDFaracheJMingueneauMMathisDBenoistC. Imbalanced signal transduction in regulatory T cells expressing the transcription factor FoxP3. Proc Natl Acad Sci U S A (2015) 112(48):14942–7. doi: 10.1073/pnas.1520393112

  • CrossRef
  • Google Scholar

• 88

  TischnerDGagglIPeschelIKaufmannMTuzlakSDrachMet al. Defective cell death signalling along the Bcl-2 regulated apoptosis pathway compromises Treg cell development and limits their functionality in mice. J Autoimmun (2012) 38(1):59–69. doi: 10.1016/j.jaut.2011.12.008

  • CrossRef
  • Google Scholar

• 89

  XieMMFangSChenQLiuHWanJDentAL. Follicular regulatory T cells inhibit the development of granzyme B-expressing follicular helper T cells. JCI Insight (2019) 4(16):e128076. doi: 10.1172/jci.insight.128076

  • CrossRef
  • Google Scholar

• 90

  PanduroMBenoistCMathisD. Tissue Tregs. Annu Rev Immunol (2016) 34:609–33. doi: 10.1146/annurev-immunol-032712-095948

  • CrossRef
  • Google Scholar

• 91

  PovoleriGAMNova-LampertiEScottaCFanelliGChenYCBeckerPDet al. Human retinoic acid-regulated CD161(+) regulatory T cells support wound repair in intestinal mucosa. Nat Immunol (2018) 19(12):1403–14. doi: 10.1038/s41590-018-0230-z

  • CrossRef
  • Google Scholar

• 92

  KitohAOnoMNaoeYOhkuraNYamaguchiTYaguchiHet al. Indispensable role of the Runx1-Cbfbeta transcription complex for in vivo-suppressive function of FoxP3+ regulatory T cells. Immunity (2009) 31(4):609–20. doi: 10.1016/j.immuni.2009.09.003

  • CrossRef
  • Google Scholar

• 93

  OnoM. Control of regulatory T-cell differentiation and function by T-cell receptor signalling and Foxp3 transcription factor complexes. Immunology (2020) 160(1):24–37. doi: 10.1111/imm.13178

  • CrossRef
  • Google Scholar

• 94

  KwonHKChenHMMathisDBenoistC. Different molecular complexes that mediate transcriptional induction and repression by FoxP3. Nat Immunol (2017) 18(11):1238–48. doi: 10.1038/ni.3835

  • CrossRef
  • Google Scholar

• 95

  ParkJEBottingRADominguez CondeCPopescuDMLavaertMKunzDJet al. A cell atlas of human thymic development defines T cell repertoire formation. Science (2020) 367(6480):eaay3224. doi: 10.1101/2020.01.28.911115

  • CrossRef
  • Google Scholar

• 96

  HuangXGongRLiXVirtueAYangFYangIHet al. Identification of novel pretranslational regulatory mechanisms for NF-kappaB activation. J Biol Chem (2013) 288(22):15628–40. doi: 10.1074/jbc.M113.460626

  • CrossRef
  • Google Scholar

• 97

  OhHGrinberg-BleyerYLiaoWMaloneyDWangPWuZet al. An NF-kappaB Transcription-Factor-Dependent Lineage-Specific Transcriptional Program Promotes Regulatory T Cell Identity and Function. Immunity (2017) 47(3):450–65 e5. doi: 10.1016/j.immuni.2017.08.010

  • CrossRef
  • Google Scholar

• 98

  RuanQChenYH. Nuclear factor-kappaB in immunity and inflammation: the Treg and Th17 connection. Adv Exp Med Biol (2012) 946:207–21. doi: 10.1007/978-1-4614-0106-3_12

  • CrossRef
  • Google Scholar

• 99

  Grinberg-BleyerYCaronRSeeleyJJDe SilvaNSSchindlerCWHaydenMSet al. The Alternative NF-kappaB Pathway in Regulatory T Cell Homeostasis and Suppressive Function. J Immunol (2018) 200(7):2362–71. doi: 10.4049/jimmunol.1800042

  • CrossRef
  • Google Scholar

• 100

  Grinberg-BleyerYOhHDesrichardABhattDMCaronRChanTAet al. NF-kappaB c-Rel Is Crucial for the Regulatory T Cell Immune Checkpoint in Cancer. Cell (2017) 170(6):1096–108 e13. doi: 10.1016/j.cell.2017.08.004

  • CrossRef
  • Google Scholar

• 101

  FerrandinoFGrazioliPBellaviaDCampeseAFScrepantiIFelliMP. Notch and NF-kappaB: Coach and Players of Regulatory T-Cell Response in Cancer. Front Immunol (2018) 9:2165. doi: 10.3389/fimmu.2018.02165

  • CrossRef
  • Google Scholar

• 102

  YanYXiongZSongJYinYCowanAWangHet al. Expression Of TCTP Antisense In CD4+ CD25 High Regulatory T Cells Weakens The Cell Survival And Aggravates Vascular Inflammation. Am Heart Assoc (2008) 203(2):401–8. doi: 10.1016/j.atherosclerosis.2008.07.041

  • CrossRef
  • Google Scholar

• 103

  RothsteinDMCamirandG. New insights into the mechanisms of Treg function. Curr Opin Organ Transplant (2015) 20(4):376–84. doi: 10.1097/MOT.0000000000000212

  • CrossRef
  • Google Scholar

• 104

  LewisSMWilliamsAEisenbarthSC. Structure and function of the immune system in the spleen. Sci Immunol (2019) 4(33):eaau6085. doi: 10.1126/sciimmunol.aau6085

  • CrossRef
  • Google Scholar

• 105

  ShenHWuNNanayakkaraGFuHYangQYangWYet al. Co-signaling receptors regulate T-cell plasticity and immune tolerance. Front Biosci (Landmark Ed) (2019) 24:96–132. doi: 10.2741/4710

  • CrossRef
  • Google Scholar

• 106

  LaiBWangJFagensonASunYSaredyJLuYet al. Twenty Novel Disease Group-Specific and 12 New Shared Macrophage Pathways in Eight Groups of 34 Diseases Including 24 Inflammatory Organ Diseases and 10 Types of Tumors. Front Immunol (2019) 10:2612. doi: 10.3389/fimmu.2019.02612

  • CrossRef
  • Google Scholar

• 107

  SharmaARudraD. Emerging Functions of Regulatory T Cells in Tissue Homeostasis. Front Immunol (2018) 9:883. doi: 10.3389/fimmu.2018.00883

  • CrossRef
  • Google Scholar

• 108

  Wang JLBNanayakkaraGYangQSunYLuYShaoYet al. Experimental data-mining analyses reveal new roles of low-intensity ultrasound in differentiating cell death regulatome in cancer and non-cancer cells via potential modulation of chromatin long-range interactions. Front Oncol (2019) 9:2019.00600. doi: 10.3389/fonc.2019.00600

  • CrossRef
  • Google Scholar

• 109

  WangYAnRUmanahGKParkHNambiarKEackerSMet al. A nuclease that mediates cell death induced by DNA damage and poly(ADP-ribose) polymerase-1. Science (2016) 354(6308):aad6872. doi: 10.1126/science.aad6872

  • CrossRef
  • Google Scholar

• 110

  GalluzziLVitaleIAaronsonSAAbramsJMAdamDAgostinisPet al. Molecular mechanisms of cell death: recommendations of the Nomenclature Committee on Cell Death 2018. Cell Death Differ (2018) 25(3):486–541. doi: 10.1038/s41418-018-0102-y

  • CrossRef
  • Google Scholar

• 111

  AlvarezFAl-AubodahTAYangYHPiccirilloCA. Mechanisms of TREG cell adaptation to inflammation. J Leukoc Biol (2020) 108(2):559–71. doi: 10.1002/JLB.1MR0120-196R

  • CrossRef
  • Google Scholar

• 112

  FeriottiCde AraujoEFLouresFVda CostaTAGaldinoNALZamboniDSet al. NOD-Like Receptor P3 Inflammasome Controls Protective Th1/Th17 Immunity against Pulmonary Paracoccidioidomycosis. Front Immunol (2017) 8:786. doi: 10.3389/fimmu.2017.00786

  • CrossRef
  • Google Scholar

• 113

  ShouvalDSBiswasAKangYHGriffithAEKonnikovaLMascanfroniIDet al. Interleukin 1beta Mediates Intestinal Inflammation in Mice and Patients With Interleukin 10 Receptor Deficiency. Gastroenterology (2016) 151(6):1100–4. doi: 10.1053/j.gastro.2016.08.055

  • CrossRef
  • Google Scholar

• 114

  ColarussoCDe FalcoGTerlizziMRoviezzoFCerquaISirignanoMet al. The Inhibition of Caspase-1- Does Not Revert Particulate Matter (PM)-Induced Lung Immunesuppression in Mice. Front Immunol (2019) 10:1329. doi: 10.3389/fimmu.2019.01329

  • CrossRef
  • Google Scholar

• 115

  HuangHLongLZhouPChapmanNMChiH. mTOR signaling at the crossroads of environmental signals and T-cell fate decisions. Immunol Rev (2020) 295(1):15–38. doi: 10.1111/imr.12845

  • CrossRef
  • Google Scholar

• 116

  LimELOkkenhaugK. Phosphoinositide 3-kinase delta is a regulatory T-cell target in cancer immunotherapy. Immunology (2019) 157(3):210–8. doi: 10.1111/imm.13082

  • CrossRef
  • Google Scholar

• 117

  FanMYTurkaLA. Immunometabolism and PI(3)K Signaling As a Link between IL-2, Foxp3 Expression, and Suppressor Function in Regulatory T Cells. Front Immunol (2018) 9:69. doi: 10.3389/fimmu.2018.00069

  • CrossRef
  • Google Scholar

• 118

  RajaeiEHaybarHMowlaKZayeriZD. Metformin one in a Million Efficient Medicines for Rheumatoid Arthritis Complications: Inflammation, Osteoblastogenesis, Cardiovascular Disease, Malignancies. Curr Rheumatol Rev (2019) 15(2):116–22. doi: 10.2174/1573397114666180717145745

  • CrossRef
  • Google Scholar

• 119

  ElmoreJPMcGeeMCNidetzNFAnannyaOHuangWAugustA. Tuning T helper cell differentiation by ITK. Biochem Soc Trans (2020) 48(1):179–85. doi: 10.1042/BST20190486

  • CrossRef
  • Google Scholar

• 120

  ShabanEBaylissGMalhotraDKSheminDWangLJGohhRet al. Targeting Regulatory T Cells for Transplant Tolerance: New Insights and Future Perspectives. Kidney Dis (Basel) (2018) 4(4):205–13. doi: 10.1159/000490703

  • CrossRef
  • Google Scholar

• 121

  GoropevsekAHolcarMAvcinT. The Role of STAT Signaling Pathways in the Pathogenesis of Systemic Lupus Erythematosus. Clin Rev Allergy Immunol (2017) 52(2):164–81. doi: 10.1007/s12016-016-8550-y

  • CrossRef
  • Google Scholar

• 122

  BernardNJ. How Treg cells lose FOXP3. Nat Rev Rheumatol (2019) 15(3):127. doi: 10.1038/s41584-019-0171-4

  • CrossRef
  • Google Scholar

• 123

  de BoussacHBruyerAJourdanMMaesARobertNGourzonesCet al. Kinome expression profiling to target new therapeutic avenues in multiple myeloma. Haematologica (2020) 105(3):784–95. doi: 10.3324/haematol.2018.208306

  • CrossRef
  • Google Scholar

• 124

  ScheineckerCGoschlLBonelliM. Treg cells in health and autoimmune diseases: New insights from single cell analysis. J Autoimmun (2020) 110:102376. doi: 10.1016/j.jaut.2019.102376

  • CrossRef
  • Google Scholar

• 125

  ChinenTKannanAKLevineAGFanXKleinUZhengYet al. An essential role for the IL-2 receptor in Treg cell function. Nat Immunol (2016) 17(11):1322–33. doi: 10.1038/ni.3540

  • CrossRef
  • Google Scholar

• 126

  KleinAMMazutisLAkartunaITallapragadaNVeresALiVet al. Droplet barcoding for single-cell transcriptomics applied to embryonic stem cells. Cell (2015) 161(5):1187–201. doi: 10.1016/j.cell.2015.04.044

  • CrossRef
  • Google Scholar

• 127

  WagnerARegevAYosefN. Revealing the vectors of cellular identity with single-cell genomics. Nat Biotechnol (2016) 34(11):1145–60. doi: 10.1038/nbt.3711

  • CrossRef
  • Google Scholar

• 128

  YangXF. Immunology of stem cells and cancer stem cells. Cell Mol Immunol (2007) 4(3):161–71.

  • Google Scholar

• 129

  SarkarPRandallSMMuddimanDCRaoBM. Targeted proteomics of the secretory pathway reveals the secretome of mouse embryonic fibroblasts and human embryonic stem cells. Mol Cell Proteomics (2012) 11(12):1829–39. doi: 10.1074/mcp.M112.020503

  • CrossRef
  • Google Scholar

• 130

  MaumusMJorgensenCNoelD. Mesenchymal stem cells in regenerative medicine applied to rheumatic diseases: role of secretome and exosomes. Biochimie (2013) 95(12):2229–34. doi: 10.1016/j.biochi.2013.04.017

  • CrossRef
  • Google Scholar

• 131

  LeeMJKimJKimMYBaeYSRyuSHLeeTGet al. Proteomic analysis of tumor necrosis factor-alpha-induced secretome of human adipose tissue-derived mesenchymal stem cells. J Proteome Res (2010) 9(4):1754–62. doi: 10.1021/pr900898n

  • CrossRef
  • Google Scholar

• 132

  KimJMKimJKimYHKimKTRyuSHLeeTGet al. Comparative secretome analysis of human bone marrow-derived mesenchymal stem cells during osteogenesis. J Cell Physiol (2013) 228(1):216–24. doi: 10.1002/jcp.24123

  • CrossRef
  • Google Scholar

• 133

  NiwaH. The principles that govern transcription factor network functions in stem cells. Development (2018) 145(6):ev157420. doi: 10.1242/dev.157420

  • CrossRef
  • Google Scholar

• 134

  BennettJASinghKPUnnisaZWelleSLGasiewiczTA. Deficiency in Aryl Hydrocarbon Receptor (AHR) Expression throughout Aging Alters Gene Expression Profiles in Murine Long-Term Hematopoietic Stem Cells. PloS One (2015) 10(7):e0133791. doi: 10.1371/journal.pone.0133791

  • CrossRef
  • Google Scholar

• 135

  ZhangZNewtonKKummerfeldSKWebsterJKirkpatrickDSPhuLet al. Transcription factor Etv5 is essential for the maintenance of alveolar type II cells. Proc Natl Acad Sci U.S.A. (2017) 114(15):3903–8. doi: 10.1073/pnas.1621177114

  • CrossRef
  • Google Scholar

• 136

  NewtonKDuggerDLSengupta-GhoshAFerrandoREChuFTaoJet al. Ubiquitin ligase COP1 coordinates transcriptional programs that control cell type specification in the developing mouse brain. Proc Natl Acad Sci U.S.A. (2018) 115(44):11244–9. doi: 10.1073/pnas.1805033115

  • CrossRef
  • Google Scholar

• 137

  EoJSongHLimHJ. Etv5, a transcription factor with versatile functions in male reproduction. Clin Exp Reprod Med (2012) 39(2):41–5. doi: 10.5653/cerm.2012.39.2.41

  • CrossRef
  • Google Scholar

• 138

  GuerquinMJCharvetBNourissatGHavisERonsinOBonninMAet al. Transcription factor EGR1 directs tendon differentiation and promotes tendon repair. J Clin Invest (2013) 123(8):3564–76. doi: 10.1172/JCI67521

  • CrossRef
  • Google Scholar

• 139

  MiletCBleherMAllbrightKOrgeurMCoulpierFDuprezDet al. Egr1 deficiency induces browning of inguinal subcutaneous white adipose tissue in mice. Sci Rep (2017) 7(1):16153. doi: 10.1038/s41598-017-16543-7

  • CrossRef
  • Google Scholar

• 140

  TakahashiKYamanakaS. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell (2006) 126(4):663–76. doi: 10.1016/j.cell.2006.07.024

  • CrossRef
  • Google Scholar

• 141

  ParkCSShenYLewisALacorazzaHD. Role of the reprogramming factor KLF4 in blood formation. J Leukoc Biol (2016) 99(5):673–85. doi: 10.1189/jlb.1RU1215-539R

  • CrossRef
  • Google Scholar

• 142

  SakaguchiSSakaguchiNAsanoMItohMTodaM. Immunologic self-tolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J Immunol (1995) 155(3):1151–64.

  • Google Scholar

• 143

  SharabiATsokosMGDingYMalekTRKlatzmannDTsokosGC. Regulatory T cells in the treatment of disease. Nat Rev Drug Discovery (2018) 17(11):823–44. doi: 10.1038/nrd.2018.148

  • CrossRef
  • Google Scholar

• 144

  MengXYangJDongMZhangKTuEGaoQet al. Regulatory T cells in cardiovascular diseases. Nat Rev Cardiol (2016) 13(3):167–79. doi: 10.1038/nrcardio.2015.169

  • CrossRef
  • Google Scholar

• 145

  TogashiYShitaraKNishikawaH. Regulatory T cells in cancer immunosuppression - implications for anticancer therapy. Nat Rev Clin Oncol (2019) 16(6):356–71. doi: 10.1038/s41571-019-0175-7

  • CrossRef
  • Google Scholar

• 146

  NgBYangFHustonDPYanYYangYXiongZet al. Increased noncanonical splicing of autoantigen transcripts provides the structural basis for expression of untolerized epitopes. J Allergy Clin Immunol (2004) 114(6):1463–70. doi: 10.1016/j.jaci.2004.09.006

  • CrossRef
  • Google Scholar

• 147

  YinYYanYJiangXMaiJChenNCWangHet al. Inflammasomes are differentially expressed in cardiovascular and other tissues. Int J Immunopathol Pharmacol (2009) 22(2):311–22. doi: 10.1177/039463200902200208

  • CrossRef
  • Google Scholar

• 148

  GehartHCleversH. Tales from the crypt: new insights into intestinal stem cells. Nat Rev Gastroenterol Hepatol (2019) 16(1):19–34. doi: 10.1038/s41575-018-0081-y

  • CrossRef
  • Google Scholar

• 149

  Gonzalez-GarzaMTCruz-VegaDECardenas-LopezAde la RosaRMMoreno-CuevasJE. Comparing stemness gene expression between stem cell subpopulations from peripheral blood and adipose tissue. Am J Stem Cells (2018) 7(2):38–47.

  • Google Scholar

• 150

  WeiSKryczekIZouW. Regulatory T-cell compartmentalization and trafficking. Blood (2006) 108(2):426–31. doi: 10.1182/blood-2006-01-0177

  • CrossRef
  • Google Scholar

• 151

  ShevyrevDTereshchenkoV. Treg Heterogeneity, Function, and Homeostasis. Front Immunol (2019) 10:3100. doi: 10.3389/fimmu.2019.03100

  • CrossRef
  • Google Scholar

• 152

  VasanthakumarAChisangaDBlumeJGlouryRBrittKHenstridgeDCet al. Sex-specific adipose tissue imprinting of regulatory T cells. Nature (2020) 579(7800):581–5. doi: 10.1038/s41586-020-2040-3

  • CrossRef
  • Google Scholar

• 153

  BitonMHaberALRogelNBurginGBeyazSSchnellAet al. T Helper Cell Cytokines Modulate Intestinal Stem Cell Renewal and Differentiation. Cell (2018) 175(5):1307–20 e22. doi: 10.1016/j.cell.2018.10.008

  • CrossRef
  • Google Scholar

• 154

  MaryanovichMFrenettePS. T-Regulating Hair Follicle Stem Cells. Immunity (2017) 46(6):979–81. doi: 10.1016/j.immuni.2017.06.011

  • CrossRef
  • Google Scholar

• 155

  HirataYFuruhashiKIshiiHLiHWPinhoSDingLet al. CD150(high) Bone Marrow Tregs Maintain Hematopoietic Stem Cell Quiescence and Immune Privilege via Adenosine. Cell Stem Cell (2018) 22(3):445–53 e5. doi: 10.1016/j.stem.2018.01.017

  • CrossRef
  • Google Scholar

• 156

  ChoILuiPPAliN. Treg regulation of the epithelial stem cell lineage. J Immunol Regener Med (2020) 8:100028. doi: 10.1016/j.regen.2020.100028

  • CrossRef
  • Google Scholar