83
views
0
recommends
+1 Recommend
1 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      The fate and lifespan of human monocyte subsets in steady state and systemic inflammation

      research-article

      Read this article at

      ScienceOpenPublisherPMC
      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Using stable isotope labeling, Patel et al. establish the lifespan of all three human monocyte subsets that circulate in dynamic equilibrium; in steady state, classical monocytes are short-lived precursors with the potential to become intermediate and nonclassical monocytes. They highlight that systemic inflammation induces an emergency release of classical monocytes into the circulation.

          Abstract

          In humans, the monocyte pool comprises three subsets (classical, intermediate, and nonclassical) that circulate in dynamic equilibrium. The kinetics underlying their generation, differentiation, and disappearance are critical to understanding both steady-state homeostasis and inflammatory responses. Here, using human in vivo deuterium labeling, we demonstrate that classical monocytes emerge first from marrow, after a postmitotic interval of 1.6 d, and circulate for a day. Subsequent labeling of intermediate and nonclassical monocytes is consistent with a model of sequential transition. Intermediate and nonclassical monocytes have longer circulating lifespans (∼4 and ∼7 d, respectively). In a human experimental endotoxemia model, a transient but profound monocytopenia was observed; restoration of circulating monocytes was achieved by the early release of classical monocytes from bone marrow. The sequence of repopulation recapitulated the order of maturation in healthy homeostasis. This developmental relationship between monocyte subsets was verified by fate mapping grafted human classical monocytes into humanized mice, which were able to differentiate sequentially into intermediate and nonclassical cells.

          Related collections

          Most cited references35

          • Record: found
          • Abstract: found
          • Article: not found

          Constant replenishment from circulating monocytes maintains the macrophage pool in adult intestine

          The paradigm that resident macrophages in steady-state tissues are derived from embryonic precursors has never been investigated in the intestine, which contains the largest pool of macrophages. Using fate mapping models and monocytopenic mice, together with bone marrow chimeric and parabiotic models, we show that embryonic precursors seeded the intestinal mucosa and demonstrated extensive in situ proliferation in the neonatal period. However these cells did not persist in adult intestine. Instead, they were replaced around the time of weaning by the CCR2-dependent influx of Ly6Chi monocytes that differentiated locally into mature, anti-inflammatory macrophages. This process was driven largely by the microbiota and had to be continued throughout adult life to maintain a normal intestinal macrophage pool.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: not found

            Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets.

            New official nomenclature subdivides human monocytes into 3 subsets: the classical (CD14(++)CD16(-)), intermediate (CD14(++)CD16(+)), and nonclassical (CD14(+)CD16(++)) monocytes. This introduces new challenges, as monocyte heterogeneity is mostly understood based on 2 subsets, the CD16(-) and CD16(+) monocytes. Here, we comprehensively defined the 3 circulating human monocyte subsets using microarray, flow cytometry, and cytokine production analysis. We find that intermediate monocytes expressed a large majority (87%) of genes and surface proteins at levels between classical and nonclassical monocytes. This establishes their intermediary nature at the molecular level. We unveil the close relationship between the intermediate and nonclassic monocytes, along with features that separate them. Intermediate monocytes expressed highest levels of major histocompatibility complex class II, GFRα2 and CLEC10A, whereas nonclassic monocytes were distinguished by cytoskeleton rearrangement genes, inflammatory cytokine production, and CD294 and Siglec10 surface expression. In addition, we identify new features for classic monocytes, including AP-1 transcription factor genes, CLEC4D and IL-13Rα1 surface expression. We also find circumstantial evidence supporting the developmental relationship between the 3 subsets, including gradual changes in maturation genes and surface markers. By comprehensively defining the 3 monocyte subsets during healthy conditions, we facilitate target identification and detailed analyses of aberrations that may occur to monocyte subsets during diseases.
              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Human CD14dim Monocytes Patrol and Sense Nucleic Acids and Viruses via TLR7 and TLR8 Receptors

              Introduction Blood monocytes are bone marrow-derived phagocytes involved in the innate response to bacterial, fungal, parasitic, and viral infection (Barbalat et al., 2009; Gupta et al., 2001; Serbina et al., 2009; Serbina et al., 2008; Ströher et al., 2001). Nonetheless, the full spectrum of monocyte function, and in particular the specific functions of monocyte subsets, is not yet known. Recent work has revealed heterogeneity of monocytes, with two distinct subsets in rodents as well as in humans (Ancuta et al., 2009; Auffray et al., 2009; Geissmann et al., 2003; Ingersoll et al., 2010; Sunderkötter et al., 2004; Zhao et al., 2009; Ziegler-Heitbrock, 2007). Akin to other leucocytes, such as lymphocytes, evidence suggests that the diversity of functions ascribed to monocytes might reflect cellular heterogeneity, with distinct subsets endowed with specific functions. On the basis of in vivo studies, murine monocytes have been separated into two subclasses. A major subset of “inflammatory” Gr1+ monocytes specializes in the production of tumor necrosis factor (TNF)-α, reactive oxygen species (ROS), and nitric oxide (NO) but little interleukin (IL)-10 upon in vivo infection with bacteria such as Listeria monocytogenes or parasites such as Toxoplasma gondii (Serbina et al., 2008). Gr1+ monocytes also produce type 1 interferon in response to viral ligands (Barbalat et al., 2009). In contrast, Gr1− monocytes patrol the blood vasculature, can differentiate into macrophages after extravasation into tissues, and have been suggested to be associated with tissue repair (Auffray et al., 2007; Nahrendorf et al., 2007). Their role in response to infection is not documented. Human CD14+ CD16− monocytes resemble Gr1+ cells based on surface marker expression and gene expression arrays (Geissmann et al., 2003; Ingersoll et al., 2010; Strauss-Ayali et al., 2007) and are reportedly producers of IL-10 and weak producers of TNF-α in vitro. Human CD16+ monocytes, which resemble Gr1− cells, are described as the main producers of inflammatory cytokines such as TNF-α and IL-1β in response to bacterial lipopolysaccharides (LPSs) (Belge et al., 2002). Recent work demonstrates distinct responses of human monocyte subsets to Aspergillus fumigatus conidia (Serbina et al., 2009). Although both monocyte subsets efficiently phagocytose conidia, CD14+ CD16− monocytes inhibit conidial germination yet secrete little TNF-α, whereas CD14+ CD16+ monocytes do not inhibit conidial germination but secrete large amounts of TNF-α (Serbina et al., 2009). Of note, an additional degree of complexity was introduced by the work of Grage-Griebenow (Skrzeczyńska-Moncznik et al., 2008) and Schäkel (Schäkel et al., 2002), indicating that human CD16+ monocytes are not a homogeneous cell population. Thus, on the one hand it is increasingly clear that monocyte subsets exert critical specific functions in the response to microbes, whereas on the other hand the reported differences in the behavior of monocyte subsets in mice and human have hampered the description of potential evolutionary conserved functions (Auffray et al., 2009; Skrzeczyńska-Moncznik et al., 2008; Ziegler-Heitbrock, 2007). The murine population of Gr1− monocytes patrol the blood vasculature and is therefore ideally located to survey local tissue damage and infection (Auffray et al., 2007). The present study was therefore designed to search for putative human functional homologs to murine Gr1− patrolling monocytes and to characterize the full spectrum of their functions. Results Human Homologs to Murine Monocytes Human monocytes can be distinguished from other cell types in the peripheral blood by flow cytometry (Figure 1 and Figure S1 available online). Large CD14+ CD16− monocytes (85% of blood monocytes) can be distinguished from a population of large CD14+CD16+ monocytes (5%), on the basis of the level of expression of CD16, CD62L, CD11c, CD163, and CX3CR1 (Figures 1A–1D and Figure S1). Monocytes expressing low amounts of CD14 (CD14dim, 7% of monocytes) have a smaller size and granularity, express CD16, and have lower expression of CD11b and CD163 (Figures 1A–1D). These small CD14dimCD16+monocytes can be further subdivided into CD14dimCD16+ 6-sulfo LacNAc (Slan)+ and CD14dimCD16+Slan− (Figure 1C). Blood monocytes Gr1+ and Gr1−, from C57/Bl6 mice (n = 3, samples Gr1+1, 2, and 3 and Gr1− 1, 2, and 3), were therefore compared to the prospective human monocyte subsets from three healthy donors with whole-genome expression arrays (as described in the Experimental Procedures). For the three healthy donors, five samples were analyzed for each subset, including biological replicates from donor 1 (D1 samples 17–19, 19–21, 23–25, 26–28, and 29–31), and simplicates from donor 2 (32, 33, 34, 35, and 36) and donor 3 (37, 38, 39, 40, and 41) (see also Supplemental Experimental Procedures). Common expressed genes between human and mouse chips were selected with Ingenuity. Principal component analysis (PCA) on the human data set indicated that CD14+CD16−, CD14+CD16+, and CD14dimCD16+ subsets segregated in independent clusters (Figure 1E). Expression of Slan did not allow discrimination of subsets among CD14dimCD16+cells (Figure 1E). Human and mouse data sets were then merged and submitted to a hierarchical clustering with the Spearman correlation similarity measure with average linkage that used “all expressed genes” and “fold 2 expressed” genes (Figure 1F and Figure S1), and submitted independently to a PCA with a covariance matrix (Figure S1). These two unsupervised analyses showed comparable results, summarized in Figure 1G and Figure S1. Results from the hierarchical clustering analysis using D1, D2, and D3 with all individual D1 biological replicates (Figure 1F), with the mean of D1 replicates (Figure S1), or with individual replicates (not shown) were similar and led to an identical conclusion. In all cases the samples were split into two main groups, each of which included human and mouse samples. Gr1− murine samples always segregated with the samples of CD14dim CD16+ human monocytes samples, whereas Gr1+ murine monocyte samples segregated with all of the CD14+ human samples, independently of expression of CD16 (Figure 1F and Figure S1). These results predicted that the absence of expression of the CD14 and CD163 antigens distinguishes the putative human homologs of murine Gr1− “patrolling” monocytes, whereas both CD14+CD16− and CD14+CD16+ cells would resemble mouse Gr1+ “inflammatory” monocytes, in contrast to previous hypotheses (Geissmann et al., 2003; Ingersoll et al., 2010; Ziegler-Heitbrock, 2007). “Patrolling” CD14dim Monocytes To test this prediction, we purified human monocyte subsets from peripheral blood by flow cytometry (as described in Figure S1), labeled them with fluorescent probes, transferred them intravenously into Rag2−/− Il2rg−/− Cx3cr1gfp mice, and examined them by intravital microscopy (Figure 1H and Movies S1 and S2). Most human CD14dim monocytes attached immediately to the endothelium after intravenous transfer and crawled for an extended period of time, whereas crawling was not observed after intravenous transfer of CD14+ monocytes (Figure 1I), further suggesting that CD14dim monocytes are the functional homologs of murine “patrolling” Gr1− monocytes. The ability of human monocytes to attach and crawl on murine endothelium was consistent with interspecies conservation of integrins, in particular lymphocyte function-associated antigen-1 (LFA-1) (Vidovic et al., 2003), which is critical for murine monocyte crawling (Auffray et al., 2007). To test this hypothesis, we therefore performed intravenous injection of human LFA1 blocking antibody (clone 38[Dransfield et al., 1992]), and observed that it abolished crawling by human monocytes inside mouse blood vessels (Figure 1I). These data indicated that human CD14dim monocytes can patrol blood vessels in vivo, in a LFA1-dependent manner. “Inflammatory” CD14+CD16− and CD14+CD16+ Monocytes Respond to Bacteria-Associated Signals As expected, CD14+ monocytes were very efficient at phagocytosing latex beads (Figure 2A). CD14+CD16 − monocytes produced high level of ROS (Figure 2B) and also expressed more mRNA for myeloperoxidase and lysozyme (Figure S2). In the presence of LPS, CD14+CD16− monocytes produced very high amounts of a distinct and limited set of chemokines and cytokines, including IL-6, IL-8, CCL2, and CCL3 (Figure 2C). The double-positive CD14+CD16+ monocyte population did not produce ROS (Figure 2B), expressed mRNA for myeloperoxidase and lysozyme at low levels (Figure S2), and produced cytokines distinct from those produced by CD14+CD16− monocytes (Figure 2C). They were the main producers of IL-1β and TNF-α in response to LPS and additionally produced IL-6 and CCL3 (Figure 2C). CD14+CD16+ also accounted for the bulk of cytokine production in response to the Toll-like receptor (TLR) 2 agonist Pam3ck4 (data not shown and Figure S2). Both CD14+CD16− and CD14+CD16+ monocytes produced moderate amount of the anti-inflammatory cytokine IL-10, with slightly different kinetics (Figure S2) that may explain previous contradictory reports (Belge et al., 2002; Skrzeczyńska-Moncznik et al., 2008). Therefore, CD14+ monocytes as a population exhibited the basic characteristics of Gr1+ “inflammatory” TNF-α and ROS-producing murine monocytes (Serbina et al., 2008). Distinct cell types within this population, i.e., CD14+CD16− and CD14+CD16+, are responsible for the bulk production of ROS, IL-6, IL-8, and CCL2 and of TNF-α and IL-1, respectively, in the presence of LPS. These observations are in line with the published evidence that LPS induces expression of different cytokines by CD14+ and CD16+ monocytes (Belge et al., 2002; Skrzeczyńska-Moncznik et al., 2008). The secretion of type I IFN, IL-2, IL-4, IL-5, IL-7, IL-9, IL-12, IL-13, IL-15, IL-17, CCL11, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), platelet-derived growth factor (PDGF), and vascular endothelial growth factor (VEGF) was not detected in monocytes in these experiments (data not shown). “Patrolling” CD14dim Monocytes Respond Poorly to Bacterial Cues CD14dim cells exhibited a limited ability to uptake latex beads in comparison to CD14+ monocytes (Figure 2A), did not produce ROS (Figure 2B), expressed little mRNA for myeloperoxidase (MPO) and lysozyme (Figure S2), and gave little response to LPS or to PAM3CK4 TLR1 and two agonists (Figure 2C and Figure S2). Thus, the production of IL-1β and TNF-α in response to LPS, previously attributed to the CD16+ fraction of monocytes, is restricted to the double-positive CD14+ CD16+ cells in our experiments. As expected from PCA and clustering analysis, CD14dim monocytes expressed a distinct set of genes in comparison to both CD14+CD16+ and CD14+CD16− monocytes, including chemokine receptors, as well as different sets of apolipoproteins, scavenger receptors for lipids and dying cells, and cytokine receptors (Figure S2). CD14dim monocytes did not express inflammatory chemokine receptors such as CCR2 and expressed lower levels of CCR1, CCR5, IL-17RA, apolipoprotein (Apo) B48, CD36, MPO, and lysozyme. In contrast, they expressed higher amounts of IL-10 receptor and CXCL16 scavenger receptor ApoA and ApoE (Figure S2) and produced high amounts of IL-1 receptor antagonist during overnight culture (Figure 2D). Overall, the monocyte proinflammatory response, including phagocytosis and the production of ROS, MPO, lysozyme, and cytokines in response to well-characterized TLR1, TLR2, and TLR4 signals, appeared to be mediated by CD14+ monocytes, whereas the profile of CD14dim monocytes was, if anything, anti-inflammatory in this context. Are CD14dim Cells Monocytes? We thus considered the possibility that CD14dim cells might not be monocytes. Cytologic examination indicated that CD14dim cells had a morphology compatible with monocytes, and flow cytometric profiles indicated that they expressed the macrophage colony-stimulating factor (MCSF) receptor and CD68, but low CD11b and CD163 (see Figures 1B–1D). A subset of human natural killer (NK) lymphocytes express CD16 and HLA-DR, and recent studies have revealed that NK cells can be confused for myeloid cells (Blasius et al., 2007; Caminschi et al., 2007; Vosshenrich et al., 2007). Therefore, we tested the hypothesis that CD14dim cells might actually represent major histocompatibility complex (MHC) class II-positive NK cells by examining cells from patients deficient in the common cytokine receptor gamma chain (γc); these patients lack all lymphoid lineages. We found normal numbers of CD14dim CD16+ cells in three γc-deficient patients, whereas as expected CD16+ NK cells were absent from these patients (Figure 3A), indicating that CD14dim cells are not lymphoid. Cells expressing HLA-DR, CD16, and Slan and negative for CD14 have also been described as antigen-presenting dendritic cells (DCs) (Schäkel et al., 2006). Both CD14+ and CD14dim monocytes stimulate allogenous mixed lymphocyte reaction at high effector/target ratio (Figures S2D and S2E). However, they do not process and present antigen (Figure 3B). For investigating the ability of CD14dim CD16+ cells to present a recall protein antigen to T cells in comparison to other monocyte subsets and blood DCs, tetanus toxoid was added to coculture of monocytes or DCs and T cells from autologous vaccinated donors (Sallusto and Lanzavecchia, 1994). Freshly isolated monocytes from any subset did not promote any significant antigen-dependent T cell proliferation or IL-2 production ,whereas CD14−CD16− HLA-DR+ DCs are able to efficiently present tetanus toxoid to T cells and to trigger strong IL-2 production (Figure 3B). Therefore, CD14dim cells appeared to be unable to process antigen, and genetic and functional evidence indicated that they are a bona fide monocyte subset. Of note, in our PCA, CD14dimCD16+Slan+ cells clustered together with CD14dimCD16+Slan− cells and total CD14dimCD16+ cells (see Figure 1). CD14dim Monocytes Produce a Selected Set of Proinflammatory Cytokines in Response to Viruses The poor response via TLR1 and TLR 2 and via TLR4 does not preclude the involvement of monocytes in the inflammatory responses to other classes of pathogens. P. Ancuta and others have proposed a role for CD16+ monocytes in the immune response to HIV (Ancuta et al., 2006) and monocytic cells have been shown to produce inflammatory cytokines in the presence of viruses (Gupta et al., 2001; Ströher et al., 2001). A separate subset of white blood cells, plasmacytoid DCs, are known to be the main blood producers of type 1 IFN in response to viruses (Liu, 2005), although murine Gr1+ monocytes can contribute to production of type 1 IFN in response to viruses (Barbalat et al., 2009). Among TLRs that recognize nucleic acids and are involved in the immune response to viruses (TLR3, TLR7, TLR8, and TLR9), monocytes weakly express TLR3 and TLR9 and do not respond to TLR3 and TLR9 agonists (Jarrossay et al., 2001; Kadowaki et al., 2001) (Figure S3). However, human monocytes express TLR7 and TLR8 (Supplemental figure S3), the two other endosomal TLRs involved in the immune response to viruses (Diebold et al., 2004; Heil et al., 2004; Lund et al., 2004). We thus reasoned that human monocytes subsets may be involved in the pro-inflammatory response to viruses. CD14+ monocytes produced very high amounts of the granulocyte, B cell, and T cell “helper” chemokines and cytokines IL-6, IL-8, and CCL2, in response to measles (ssRNA [−]) and herpes simplex virus 1 (HSV-1) (dsDNA) at a multiplicity of infection (MOI) of 1, similar to what was observed for LPS. In contrast, freshly purified CD16+ monocytes, especially CD14dim CD16+ monocytes, whether Slan+ or Slan−, selectively produced very high amounts of the proinflammatory cytokines TNF-α, IL-1β, and of CCL3 in response to measles and HSV-1 (Figures 4A and 4B). Of note, secretion of type I IFN by human monocytes was very low (not shown). CD14dim Monocytes Respond to Viruses via a Unique TLR7-TLR 8-MYD88-MEK-Dependent Pathway To investigate the mechanisms of cytokine production by CD14dim and CD14+ monocytes, we examined their response to TLR7 and TLR8 agonists. The response of monocyte subsets to stimulation with TLR8 agonists (3M2 and R848) and to a lesser extent to a TLR7 agonist (3M13) was similar to their response to intact viruses. CD14dim monocytes produced very high amounts of proinflammatory cytokines, but little IL-6, IL-8, and CCL2, whereas CD14+ CD16− monocytes produced little proinflammatory cytokines, but high IL-6 and IL-8 (Figure 4C). As was true for responses to viruses, the population of CD14+ CD16+ monocytes displayed an intermediate phenotype (Figure S3). To investigate whether cytokine production by monocytes required infectious virus, we incubated monocytes with either live, ultraviolet light (UV)-inactivated, or heat-inactivated HSV1-green fluorescent protein (GFP) expressing virus. Control fibroblasts exhibited GFP fluorescence only when infected with live virus (Figure S3). In contrast, monocytes exhibited similar GFP fluorescence whether they were exposed to live, UV-inactivated, or heat-treated viruses (Figure S3), probably reflecting uptake rather than infection. Monocytes produced cytokines in response to live virus and UV-inactivated virus, but not in response to denatured virus particles obtained by heat-inactivation (Figure S3). These data indicate that monocytes are activated by HSV-1 particles, in line with a TLR-mediated response, but active HSV-1 replication may not be required for cytokine production at a MOI of 1. Similarly, replication of measles virus was not observed at a MOI of 1 (Figure S3). To confirm the role of TLRs and specifically of TLR7 and TLR8 for cytokine production, we purified CD14dim and CD14+ CD16− monocytes from the peripheral blood of patients deficient in myeloid differentiation primary response gene 88 (MYD88) and in some cases in interleukin-1 receptor-associated kinase 4 (IRAK-4) (Picard et al., 2003; von Bernuth et al., 2008) and from controls and were stimulated with measles virus and/or HSV1. Production of TNF-α and IL-1β by CD14dim monocytes was totally dependent on MyD88, whereas production of chemokines CCL5 and CXCL10 was maintained, likely due to intracytoplasmic sensors and interferon-inducible RNA helicase MDA5 and RIG-I, (Figure 4D). Production of IL-6 and IL-8 by CD14+ CD16- monocytes also required MYD88 (Figure 4E). Similar results were obtained with IRAK4-deficient monocytes (Figure 4E). Altogether, these data indicated that cytokine production by monocytes in response to viruses (measles and HSV-1) is mediated via a TLR- MYD88 pathway, most likely via TLR8 and TLR7. Notably, TLR3 signaling is largely MYD88 independent and thus data from Figures 4D and 4E confirm that TLR3 is not involved in monocyte activation by viruses. We then investigated whether selective monocyte subset activation via TLRs was also a feature of mouse monocytes. Murine blood monocytes were incubated with LPS or the TLR7 agonist 3M13. Gr1+ monocytes responded both to LPS and TLR7 agonist, whereas Gr1− monocytes responded better to TLR7 agonist than to LPS (Figures S1E and S1F), suggesting a similar response to TLR agonist by murine Gr1− and human CD14dim monocytes. We reasoned that monocyte subsets might display distinct effector responses (i.e., the production of inflammatory cytokines versus “helper” cytokines) in response to viruses and TLR7 and TLR8 agonists because different downstream signaling pathways might be engaged in different cell types. To test this possibility, we examined activation of various signaling pathways in different subsets after TLR7 and/or TLR8 activation. As has been previously observed (Levy et al., 2006), p38 mitogen-activated protein kinase (MAPK) was rapidly phosphorylated in CD14+ monocytes in response to TLR8 agonists. In contrast, p42 mitogen-activated protein kinase kinase 1 (MEK1) phosphorylation, but not p38 MAPK, was observed 30 min after stimulation in CD14dim monocytes in response to TLR7 and TLR8 agonists (Figure 5A), followed 90 min later by Jun N-terminal kinases (JNK) phosphorylation (Figure 5A). This suggested that different signaling pathways are activated in monocyte subsets. We therefore investigated whether pharmacological inhibitors of MEK1 or p38 affect cytokine production by monocyte subsets in response to TLR7 and TLR8 agonists. Production of proinflammatory cytokines in response to TLR7 and/or TLR8 agonists by CD14dim monocytes was strongly inhibited by the MEK1 inhibitor in a dose-dependent manner (Figure 5B), but not by a p38 inhibitor (Figure S4). In contrast, production of cytokines by CD14+ monocytes was partially inhibited by a p38 inhibitor (30% inhibition, Figure S4) but not by the MEK1 inhibitor (Figure 5B). Although pharmacological inhibitors can have off-target effects, subset-specific responses to these inhibitors indicate differential involvement of MAPK pathways. Altogether, these data indicate that distinct subset-specific pathways control cytokine production by monocytes in responses to viruses and nucleic acids. We propose a model in which production of IL-6 and IL-8 by CD14+ monocytes involves activation of p38 in addition to a classical NF-κB pathway, whereas production of TNF-α, IL-1, and CCL3 by CD14dim monocytes is controlled by an alternative p42 MEK MAP kinase pathway (Figures S4B–S4F). CD14dim Monocyte-Specific Inflammatory Response to Self-Nucleic Acids in Patients with Lupus Intracellular TLRs activated by nucleic acids, and particularly RNA, have evolved under strong selection, suggesting an essential role in host survival (Barreiro et al., 2009). However, as a possible tradeoff mechanism, TLR7- and TLR9-dependent recognition of endogenous nucleic acids is also a feature of autoimmune diseases such as systemic lupus erythematosus (SLE), a disease associated with autoantibodies to nucleosome and ribonucleoproteins, and immune complex deposition in several organs (Rahman and Isenberg, 2008). Ribonucleoprotein-containing immune complexes stimulate immune cell such as plasmacytoid DCs and B cells via TLRs (Lau et al., 2005; Lövgren et al., 2006). Although the mechanisms of Lupus-induced tissue damage, and in particular glomerulonephritis, are poorly understood, recent work in the mouse has proposed that activation of myeloid cells by glomerular immune complex, which requires both FcRγ and TLRs (Bergtold et al., 2006; Clynes et al., 1998), initiates an inflammatory response that results in glomerulonephritis. Murine Gr1− monocytes have recently been shown to be selectively expanded in nephritis, to accumulate inside the capillary vessels in the nephritic kidney, and to be activated by immune complexes in an FcRγ-dependent manner (Bergtold et al., 2006). In human, genetic studies studies have associated CD16 with lupus (Rhodes and Vyse, 2008). It was further shown that CD16+ monocytes adhere in vitro to endothelial cells better than CD14+ monocytes, a phenomenon in part dependent on the chemokine receptor CX3CR1 (Ancuta et al., 2003), and human CD14dimCD16+ monocytes patrol capillaries in vivo as shown in this report (see Figures 1H and 1I). Finally, elevated amounts of fractalkine (the CX3CR1 ligand) expression and accumulation of CD16+ monocytes inside glomerular blood vessels has been documented in human SLE (Hill et al., 2005; Yoshimoto et al., 2007). In accordance with the literature (Hill et al., 2005; Yoshimoto et al., 2007), we confirmed and observed a glomerular IgG deposition and monocytic infiltrate (CD68+ CD16+ CD14dim and CD15− CD3ɛ−) in glomerular vessels from patients with class IV-G lupus nephritis (Figure 6A). We thus hypothesized that nucleic acid-containing immune complexes present in glomerular capillaries activate CD14dim monocytes to produce TNF-α, IL-1, and CCL3 cytokines that may damage the endothelium. We compared the effects of serum from individuals with SLE and anti-ribonucleoprotein (RNP) antibodies (see Experimental Procedures) on monocyte activation. Patient sera induced strong production of TNF-α and CCL3 by CD14dim monocytes isolated from healthy donors, as well as the production of IL-6 and IL-8 by CD14+ monocytes, in comparison with sera from controls (Figure 6B). Sera were then pretreated with RNase and DNase, and/or Ig depletion, and/or addition of ribonucleoprotein, before incubation with monocytes (Figures 6C and 6D). Both nucleic acids and Ig were observed to contribute 30%–50% of the production of CCL3 and TNF-α by CD14dim monocytes; the depletion of nucleic acids and Ig together did not have an additive effect on inhibition. These results indicated that in addition to TLR7-TLR8 agonists, and viruses, ribonucleoprotein-containing immune complexes induce CCL3 and TNF-α production by CD14dim monocytes. Discussion This study establishes CD14dim human monocytes as a bona fide myeloid immune cell population that patrol blood vessels and exert specific effector functions in the inflammatory response to viruses and nucleic acids. Genome-wide expression analysis (Figure 1) predicted that these monocytes, which lack CD14 expression (CD14dim) and represent a minority (∼7%) of human blood monocytes, resemble mouse “patrolling” Gr1− monocytes and can be distinguished from lymphoid cells, dendritic cells, and CD14+ CD16− and CD14+ CD16+ monocytes. CD14dim monocytes patrol capillaries after adoptive transfer in vivo, do not produce ROS, exhibit low or absent myeloperoxidase and lysozyme production in a steady state, constitutively produce IL-1RA, and are weak producers of inflammatory cytokines after exposure to LPS or a TLR1 and TLR2 agonist. Nevertheless CD14dim cells belong to the monocytic lineage, given that they develop in individuals lacking lymphoid precursors due to loss of function mutations in X-linked common γ chain, implying that they are developmentally distinct from NK CD16+ lymphocytes. CD14dim monocytes exhibited an anti-inflammatory phenotype in a steady state and respond poorly to surface-associated TLR stimulation, but specialize in the production of the proinflammatory cytokines TNF-α, IL-1β, and of CCL3 in response to viruses and nucleic acids, via a unique TLR7-8, MyD88 and MEK-dependent pathway. Of note, murine “patrolling” Gr1− monocytes also responded better to TLR7 agonist than to LPS (Figure S1), suggesting a degree of functional similarity between human and murine patrolling monocytes. Although they can stimulate the proliferation of allogenic T cells at high effector-to-target ratio (Sallusto and Lanzavecchia, 1994), blood monocytes, including the CD14dim subset, do not process and present a recall antigen to T cells, in contrast to blood DCs. This is in contrast to some previous reports that identified DCs as a population expressing CD16 and Slan (Schäkel et al., 2002). However, we observed that monocytes, in particular the CD14dim subset, can promote autologous T cell proliferation, in an antigen-independent manner, and without detectable increase in IL-2 production. Therefore, although monocytes are clearly distinct from DCs, further studies will be needed to clarify the functional outcomes of their interaction with T cells (Evans et al., 2009). Reported differences in the behavior of monocyte subsets in mice and human (Auffray et al., 2009; Skrzeczyńska-Moncznik et al., 2008; Ziegler-Heitbrock, 2007) have stimulated efforts to identify human homologs to monocyte subsets (Ingersoll et al., 2010). The PCA and hierachical clustering analysis reported here suggested that monocytes expressing CD14 (CD14+), which represent more than 90% of blood monocytes, resemble mouse “inflammatory” Gr1+ monocytes, whether or not they express CD16. Indeed, they share their key functional characteristics; they are responsible for phagocytosis, ROS production, and cytokine production in response to LPS via a classical MyD88-, p38-, and NF-κB dependent pathway. Within this subset, distinct cells, i.e., CD14+CD16− and CD14+CD16+ cells, are responsible for the bulk of ROS, IL-8, and IL-6 production and for TNF-α and IL-1 production respectively in response to LPS. This observation may appear different from a recent work that concluded that CD14+ CD16− monocytes were homologous to mouse Gr1+ monocytes, whereas CD14+ CD16+ monocytes were homologous to mouse Gr1− monocytes (Ingersoll et al., 2010). However, this discrepancy can be accounted for by the fact that the Ingersoll study compared only two human subsets (CD14+ CD16− monocytes and human CD16+ monocytes) with two murine subsets (Gr1+ and Gr1− monocytes). Their CD16+ monocyte population may have contained both the CD14dim subset and the CD14+ CD16− subset. We therefore propose that CD14dim CD16+ monocytes represent a monocyte subset that patrol blood vessels and selectively detect virally infected and damaged cells to produce proinflammatory cytokines. This role of CD14dim CD16+ monocytes is reminiscent of the main function of plasmacytoid DCs (PDCs), which are also specialized in response to viral stimulation via TLR7 and TLR9 (Liu, 2005). However, PDCs secrete type I interferons, whereas CD14+ monocytes produce IL-6 and IL-8, and patrolling monocytes produce TNF-α, IL-1β, and CCL3, thus exhibiting complementary functions in the TLR-dependent antiviral response. Of note, we did not detect significant secretion of type I interferon by either CD14+ or CD14dim monocytes in this study. In the murine system Gr1+ monocytes (which resemble CD14+ human monocytes) secrete IFN-β after activation by viruses (Barbalat et al., 2009). During antiviral responses, a systemic type 1 interferon response is triggered by PDCs and Gr1+ monocytes and a local type 1 interferon response by infected fibroblasts and other tissue cells. Our data suggest that in addition, monocytes and in particular patrolling CD14dim monocytes, participate in a regional antiviral response characterized by the production of proinflammatory cytokines such as TNF-α and IL-1β and of CCL3 and are potentially involved in virus-induced immunopathology. PCA and functional experiments indicated that CD14+CD16−, CD14dim and CD14+CD16+ monocytes represent distinct cell populations. The expression of CD16 by CD14+ monocytes may correspond to an activation and/or differentiation state of CD14+ monocytes, as suggested by several groups (Ziegler-Heitbrock, 2000). They both respond to extracellular TLR such as TLR4 and TLR2 and share a high phagocytic index. In some donors the CD14+CD16+ population displayed an intermediate cytokine response to viruses and TLR7- TLR8 agonists between that of CD14dim and CD14+CD16− monocytes, and future studies will need to investigate potential heterogeneity of this “double positive” population. Although it is conceivable that CD14dim may also derive from CD14+ monocytes, there is no experimental evidence to support this claim, and results from this study argue that, whether CD14dim arise from CD14+ monocytes or directly from progenitors, their distinct morphology, behavior in vivo, and signal transduction apparatus make them two independent functional cell types. Results presented here indicate that a TLR7-8, MyD88, and MEK pathway is critical for the secretion of a distinct set of cytokines by CD14dim monocytes, whereas p38 is important in the MyD88-dependent response of CD14+ monocytes. Activation of MAPK ERK1 and ERK2 has been previously shown to play an important role in regulating cytokine production by murine myeloid cells (Agrawal et al., 2006), and the regulation of distinct cytokine genes by different members of the MAPK family (ERK versus p38 MAPK) has been previously reported in DCs (Dillon et al., 2004). We used computational modeling to examine the potential relevance of this model. Computational modeling of this forward biochemical cascades for TLR7 and TLR8 signaling in monocyte subsets simulates in silico TNF-α, IL-1β, IL-8, and IL-6 expression profiles that are consistent with our experimental data. However, more investigations will be needed to define the consequences of distinct signaling pathways being involved in the effector functions of monocyte subsets. TLR7 and TLR8 have evolved under strong selection, indicating an essential role in host survival (Barreiro et al., 2009). Our present identification of a specific human monocyte subset relying on TLR7 and TLR8 to detect viruses and nucleic acids for the production of the proinflammatory cytokines TNF-α and IL-1β further suggests that TLR7 and TLR8 are endowed with an essential physiological role. However, there have been no reports of susceptibility to viral infections in human patients with impaired MYD88 and IRAK-4 signaling, who are unresponsive to TLR7, TLR8, and TLR9 (Ku et al., 2007; von Bernuth et al., 2008; Yang et al., 2005). Patients unresponsive to TLR3, TLR7, TLR8, and TLR9 (UNC-93B deficiency) share the narrow viral phenotype of TLR3-deficient patients, with herpes simplex encephalitis (Casrouge et al., 2006; Zhang et al., 2007). Thus, TLR7 and TLR8 are apparently redundant for host defense against most common viruses, although it is possible that TLR7 and TLR8 are required for the defense against yet unidentified virus(es). Alternatively, they may have been selected by virtue of their role in recognition of endogenous nucleic acids. In support of a role in the sensing of endogenous nucleic acids for TLR7-TLR8 expressing monocytes, our results indicate that activation of CD14dim monocytes by endogenous nucleic acids can contribute to the pathogenesis of inflammatory autoimmune diseases such as SLE, during which immune complexes containing nucleoproteins accumulate in tissues and in particular the glomeruli. Because glomerular damage associated with SLE leads to kidney failure, requiring dialysis or transplantation, CD14dim monocytes may therefore represent a potentially useful cellular therapeutic target in selected inflammatory diseases. Experimental Procedures Monocyte Phenotyping and Purification Anticoagulated whole blood was collected from healthy volunteers, MYD88-deficient, IRAK-4-deficient, and γc-deficient patients after informed consent. Red blood cell lysis was used to avoid activation of monocytes by ficoll density centrifugation. Monocytes (CD14+CD16−, CD14dimCD16+ and CD14+CD16+ cells) were analyzed and purified as described in Figure S1 and the Supplemental Experimental Procedures. Intravital Microscopy and Adoptive Transfer Intravital microscopy and adoptive transfer were performed as described in the Supplemental Experimental Procedures. Array Analysis Samples The human samples come from three different healthy donors (D1, D2, and D3) but the samples from donor D1 are in triplicate. Twenty-five “Whole Human Genome” chips were therefore obtained from five prospective subsets and three different donors, including triplicate samples from one donor (see Supplemental Experimental Procedures). Six “Whole Mouse Genome” chips from Gr1+ and Gr1− mouse monocyte subsets were generated as described (Auffray et al., 2007). MIAME data are available at www.ebi.ac.uk/arrayexpress: Geissman_Miltenyi_Human, ArrayExpress accession: E-MEXP-2544 and Geissman_Miltenyi_Mouse: ArrayExpress accession: E-MEXP-2545. We submitted this series of expression arrays to an unsupervised analysis in order to unveil possible correspondences between samples from the two species. The two sets of data were obtained and processed independently as described in the Supplemental Experimental Procedures. Stimulation of Monocytes CD14hiCD16lo, CD14loCD16hi, and CD14hiCD16hi cells were sorted from the blood of healthy controls, and MyD88 and IRAK deficient patients when indicated, as described in the Supplemental Experimental Procedures. TLR Agonist Stimulations. A total of 104 cells/well were laid in a 96-well plate in 100 μL and stimulated with PBS control or TLR agonists. Optimal concentration of TLR agonists were determined in preliminary experiments. In the experiments reported in this paper, the following concentration were used: LPS (TLR4 agonist) 1 ng/ml, Pam3ck4 (TLR2) 100 μg/mL, 3M2 (TLR8) 3 μg/ml, R838 (TLR8) 3 μg/ml, and 3M13 (TLR7) 3 μg/ml, in 10% FCS, 1% P/S, and incubated for 6, 24, or 48 hr at 37°C and 5% CO2. When indicated, monocytes were preincubated (30 min) with the MEK inhibitor PD98059 (10–100 μM), the P38 inhibitor SB-203580 (50 μM), or DMSO vehicle control before exposure to TLR agonists. Viral Stimulations. We stimulated 2 × 104 monocytes or peripheral blood mononuclear cells (PBMCs) per well from controls and, when indicated, from MYD88-deficient and IRAK-4-deficient patients, with the indicated intact viruses (Yang et al., 2005) at a multiplicity of infection of 1 and placed on ice for 30 min to obtain a synchronous infection. Cell supernatant was collected and analyzed for cytokine production with the Biorad multiplex assay or ELISA tests. Human Serum. Human serum were obtained and treated as described in the Supplemental Experimental Procedures. ROS production, phagocytosis assay, antigen presentation assay, cytokine measurements, protein phosphorylation, and histological studies were performed as described in the Supplemental Experimental Procedures.
                Bookmark

                Author and article information

                Journal
                J Exp Med
                J. Exp. Med
                jem
                jem
                The Journal of Experimental Medicine
                The Rockefeller University Press
                0022-1007
                1540-9538
                03 July 2017
                03 July 2017
                : 214
                : 7
                : 1913-1923
                Affiliations
                [1 ]Division of Medicine, University College London, University of London, London, England, UK
                [2 ]Institute for Infection and Immunity, St. George’s, University of London, London, England, UK
                [3 ]Theoretical Immunology Group, Faculty of Medicine, Imperial College London, London, England, UK
                [4 ]Department of Immunobiology, Yale University, New Haven, CT
                [5 ]Howard Hughes Medical Institute, Yale University, New Haven, CT
                [6 ]Newcastle University Medical School, Newcastle University, Newcastle Upon Tyne, England, UK
                [7 ]St. George’s University Hospitals NHS Foundation Trust, London, England, UK
                Author notes
                Correspondence to Simon Yona: s.yona@ 123456ucl.ac.uk
                [*]

                R.A. Flavell, D.W. Gilroy, B. Asquith, and D. Macallan contributed equally to this paper.

                A. Rongvaux’s present address is Program in Immunology, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA.

                Author information
                http://orcid.org/0000-0001-6054-1309
                http://orcid.org/0000-0002-4855-9255
                http://orcid.org/0000-0003-3786-5545
                http://orcid.org/0000-0002-3017-2474
                http://orcid.org/0000-0003-4461-0778
                http://orcid.org/0000-0002-3014-7148
                http://orcid.org/0000-0002-3984-2008
                Article
                20170355
                10.1084/jem.20170355
                5502436
                28606987
                4518fd7a-7da9-4114-9d88-d392e0c1510d
                © 2017 Patel et al.

                This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/licenses/by/4.0/).

                History
                : 23 February 2017
                : 28 April 2017
                : 28 April 2017
                Funding
                Funded by: Engineering and Physical Sciences Research Council, DOI http://dx.doi.org/10.13039/501100000266;
                Funded by: Wellcome Trust, DOI http://dx.doi.org/10.13039/100004440;
                Funded by: Medical Research Council, DOI http://dx.doi.org/10.13039/501100000265;
                Award ID: G1001052
                Funded by: Wellcome Trust, DOI http://dx.doi.org/10.13039/100004440;
                Award ID: 093053/Z/10/Z
                Funded by: Bloodwise, DOI http://dx.doi.org/10.13039/501100007903;
                Award ID: 15012
                Funded by: Wellcome Trust, DOI http://dx.doi.org/10.13039/100004440;
                Award ID: 103865
                Funded by: Medical Research Council, DOI http://dx.doi.org/10.13039/501100000265;
                Award ID: J007439
                Award ID: G1001052
                Funded by: European Union Seventh Framework Program, DOI http://dx.doi.org/10.13039/100011102;
                Award ID: FP7/2007–2013
                Award ID: 317040
                Funded by: Leukemia and Lymphoma Research, DOI http://dx.doi.org/10.13039/501100000651;
                Award ID: 15012
                Categories
                Research Articles
                Brief Definitive Report
                311
                315

                Medicine
                Medicine

                Comments

                Comment on this article