Many mechanisms involving TNF-α, Th1 responses, and Th17 responses are implicated in chronic inflammatory autoimmune disease. Recently, the clinical impact of anti-TNF therapy on disease progression has resulted in re-evaluation of the central role of this cytokine and engendered novel concept of TNF-dependent immunity. However, the overall relationship of TNF-α to pathogenesis is unclear. Here, we demonstrate a TNF-dependent differentiation pathway of dendritic cells (DC) evoking Th1 and Th17 responses. CD14+ monocytes cultured in the presence of TNF-α and GM-CSF converted to CD14+ CD1alow adherent cells with little capacity to stimulate T cells. On stimulation by LPS, however, they produced high levels of TNF-α, matrix metalloproteinase (MMP)-9, and IL-23 and differentiated either into mature DC or activated macrophages (Mφ). The mature DC (CD83+ CD70+ HLA-DR high CD14low) expressed high levels of mRNA for IL-6, IL-15, and IL-23, induced naive CD4 T cells to produce IFN-γ and TNF-α, and stimulated resting CD4 T cells to secret IL-17. Intriguingly, TNF-α added to the monocyte culture medium determined the magnitude of LPS-induced maturation and the functions of the derived DC. In contrast, the Mφ (CD14highCD70+CD83−HLA-DR−) produced large amounts of MMP-9 and TNF-α without exogenous TNF stimulation. These results suggest that the TNF priming of monocytes controls Th1 and Th17 responses induced by mature DC, but not inflammation induced by activated Mφ. Therefore, additional stimulation of monocytes with TNF-α may facilitate TNF-dependent adaptive immunity together with GM-CSF-stimulated Mφ-mediated innate immunity.
Not only does TNF-α have multiple physiological functions (1) but it also has been implicated in a variety of human diseases including sepsis, cerebral malaria, and certain autoimmune diseases such as rheumatoid arthritis (RA),2 psoriasis, and Crohn’s disease, as well as cancer (2). Over the past decade, anti-TNF therapy for autoimmune diseases has become widespread, efficiently preventing tissue injury. Moreover, studies in patients and experimental models have confirmed the central role of TNF-α in the etiology of several autoimmune diseases (3, 4, 5) and have advanced TNF biology. Thereby, questions as to the initial rationale supporting the use of anti-TNF (3, 4), and the mode of action of this therapy have arisen (6, 7, 8, 9, 10). In particular, although important for the etiology of disease, data on the impact of anti-TNF therapy on IL-23 production by dendritic cells (DC; Refs. 6 and 7), Th1 responses (6, 9, 10) and IL-23/IL-17 axis inflammation (11, 12, 13, 14, 15) are essentially absent. Recently, Banchereau et al. (16) proposed the existence of two different types of autoimmunity; one IFN dependent and the other TNF dependent. These cross-regulate one another (17) and involve induction of differentiation of DC, which further induce cytokine-dependent immunity (16). However, the precise differentiation pathways of these DC are unclear.
DC orchestrate a variety of immune responses by stimulating the differentiation of naive CD4 T cells into helper T effectors such as Th1, Th2, Th3, and Th17 cells (18, 19, 20, 21). The cytokine environment influencing DC differentiation may be an important factor in determining which type of Th cell is induced (22). To date, although a variety of in vitro-cultured DC has been studied (23, 24, 25, 26, 27), DC-driven TNF-dependent immunity has not been elucidated. Here, we have addressed such TNF-dependent differentiation of human CD14+ monocytes into mature DC. Uniquely, IL-4 (generally used for monocyte-derived DC differentiation) was not used here because of its anti-TNF/inflammatory effects (28, 29). Instead, a combination of TNF-α and GM-CSF was used for DC preparation, because of their effect on circulating monocytes in the sera of patients with autoimmune disease (4, 30). Although human monocytes cultured with this cytokine combination failed to generate functional DC (31), we did find that. LPS from Gram-negative bacteria induces not only maturation of DC but also production of IL-23 by these DC via TLR-4 ligation (32). The present study took advantage of human CD70+ DC, as previously described (33). These DC can persistently induce Th1 responses and the expression of IL-23 mRNA, whereas CD70− DC manifest exhaustion after transient activation (34). Such functions may be favorable for DC evocation of persistent Th1 and IL-23/IL-17 axis inflammation. The type of DC described here is another form of CD70+ DC; we analyzed the TNF-dependent differentiation pathway from CD14+ monocytes to CD70+ DC. Additionally, differentiation of these cells may also result in CD70+ macrophages (Mφ) producing large amounts of TNF-α and matrix metalloproteinase (MMP)-9, which is also important for their mechanism of action in autoimmune disease (35, 36, 37). Here, we propose a novel model for the TNF-dependent differentiation of a monocyte-DC/Mφ lineage and discuss where TNF-α orchestrates Th1 responses, and IL-23/IL-17 axis inflammation.
Materials and Methods
Human rTNF-α, rGM-CSF, rIL-4, agonistic mAb to human CD3 (B-B11), and CD28 (B-T3) were from Diaclone Research. Human soluble type I TNF receptor (sTNFRI) was from R&D Systems. Culture grade LPS from Escherichia coli (L4516) was from Sigma-Aldrich. Keyhole limpet hemocyanin (KLH) was from Wako Pure Chemical.
T cells, monocytes, and DC were cultured in RPMI 1640 supplemented with 10% heat-inactivated FBS (Life Technologies-BRL) plus 100 U/ml penicillin G and 100 μg/ml streptomycin.
Preparation of DC and Mφ
Samples from healthy volunteers were obtained with informed consent approved by the ethics committee of Showa University (Tokyo, Japan). PMBC were isolated by Histopaque (Sigma-Aldrich) density gradient (1.077) centrifugation in a Leuco Sep (Greiner Bio-One) that can separate granulocytes from CD14+ monocytes. After depletion of platelets, monocytes were isolated using CD14 Ab-conjugated magnetic microbeads (Miltenyi Biotech) according to the manufacturer’s instructions. The isolated CD14+ monocytes (5 × 105/ml) were cultured for 7 days in the presence of GM-CSF (10 ng/ml) and/or TNF-α (10 ng/ml). Subsequently, these cells were washed three times with warmed 10% FBS containing RPMI 1640 and cultured in the presence of LPS (10 μg/ml) and KLH (10 μg/ml). Floating cells were collected by gentle pipetting at indicated times after stimulation. Adherent cells in the cultures were collected by vigorous pipetting after incubation in 2 mM EDTA-PBS solution at 37°C for 5 min.
Preparation of CD4 T cells
CD4 T cells were isolated from PBMC-depleted CD14+ cells. Cells were treated with anti-CD45RO Ab-conjugated magnetic microbeads (Miltenyi Biotech). After absorbing CD45RO+ cells to a negative selection column (MACS 25 LD; Miltenyi Biotec), CD45RO− cells in the passthrough fractions were treated with anti-CD4 Ab-conjugated magnetic microbeads (Miltenyi Biotec), and the resulting CD4+CD45RO− cells were collected by positive selection. Finally, activated T cells and APCs including DC or B cells were depleted by M-450 pan human HLA class II and M-450 CD19 Dynabeads (Dynal Biotech), respectively. The naive CD4 T cell fraction (CD4+CD45RA+ cells, >98%) was then used for these experiments. Occasionally, purified CD4 T cells (>98%) without depletion of CD45RO+ cells were used. The CD4 T cell fractions containing memory type T cells were isolated using the pan-T isolation kit (Miltenyi Biotec) and anti-CD4 Ab-conjugated magnetic microbeads.
Cells fixed with 4% paraformaldehyde were adjusted to a concentration of 1 × 106 cells/ml and incubated at 4°C for 30 min with appropriate Abs. After two washings with ice-cold PBS containing 0.3% BSA, cells were analyzed by FACS with CellQuest software (BD Pharmingen). Ab used for flow cytometry were as follows: FITC-conjugated mAb to HLA-ABC (w6/32; Diaclone); CD14 (M5E2; BioLegend); CD70 (Ki-24; BD Pharmingen); CD80 (B-L2; Diaclone); and PE-labeled mAb to HLA-DR (B-F1), and CD40 (B-B20) from Diaclone. mAbs to CD86 (IT2.2), and CD83 (HB15e) were from BD Pharmingen. Staining for intracellular cytokines in T cells was described in a previous report (33). Ab used for the latter experiments were rabbit anti-hIL-17 polyclonal Ab or isotype control Ab (MBL), FITC-labeled goat anti-rabbit IgG-Fab (PARIS), FITC-labeled monoclonal anti-IFN-γ-Ab (4S.B3; BioLegend), PE-labeled mAb to IL-4 (MP4-25D2; BioLegend); and TNF-α (Mab11, BioLegend).
The concentrations of cytokines in DC culture medium were determined using ELISA kits for IFN-γ, TNF-α, IL-12/IL-23p40, IL-12p70, and IL-23p19/p40 (eBioscience). Data were acquired by a VERSAmax microplate reader (Molecular Devices) and analyzed by SOFTmax PRO (Molecular Devices).
T cell proliferation assay
T cell proliferation was determined by incorporation of [3H]thymidine as described in a previous report (33). T cells (2 × 105 cells) were cultured alone or cocultured with DC for 3 days. [3H]Thymidine (0.4 μCi; Amersham Pharmacia) was added to each well, and nuclear incorporation was stopped by the addition of cold thymidine after a further 18 h of culture. Isotope incorporation was determined by liquid scintillation counting in a TracorAnalytic Mark III machine, after three washings with PBS containing 0.3% BSA.
Quantitative real-time RT-PCR
Total RNA and synthesis of first-strand cDNA was from 5 × 105 cells/assay. mRNA for several different cytokines were quantified by real-time RT-PCR using a Syber Green Realtime PCR Master Mix (Toyobo) in an ABI PRISM 7000 Sequence Detection System (Applied Biosystems Japan). Amplification was conducted in a total volume of 50 μl for 40 cycles of 15 s at 94°C, 20 s at 60°C and 1 min at 72°C. mRNA levels were determined after normalization of RNA concentration against β-actin, and values were expressed as fold increase of expression above the DC control. All samples were run in triplicate. The primers used in the experiments were as follows: β-actin: forward, 5′-GCACCACACCTTCTACATGAGC; reverse, 5′-GCACAGCTTTCTCCTTAATGTCACGC. TNF-α: forward, 5′-GAGTGACAAGCCTGTAGCCCATGTTGTAGC; reverse, 5′- GCAATGATCCCAAAGTAGCCTGCCCAGAC. IL-6: forward, 5′-GA AAGCACAAAGAGGCACT; reverse, 5′-GCGCAGAATGAGATGAGATTG; IL-15: forward, 5′-ATGAGAATTTCGAAACCACATTG; reverse, 5′-CCATTAGAAGACAAACTGTTCTTTGC. Primers for IL-12p35 (33), IL-12/23p40 (33), and IL-23p19 (33) were as described in previous reports. MMP-9 primers (HA040500) were purchased from Takara Bio.
MMP-2 and MMP-9 were measured by gelatin zymography on 10% polyacrylamide gels containing 1 mg/ml gelatin in a polymerization mixture of SDS as described in previous report (37). Samples containing equal protein concentrations were mixed with sample buffer. After SDS-PAGE, SDS was washed out of the gels with 2.5% Triton X-100 for 30 min at room temperature before they were incubated in a solution of 50 mM CaCl2 and 0.02% sodium azide (pH 7.5) at 37°C for 1 day. Gels were then stained with 0.1% Coomassie blue and destained. Clear area indicated proteolytic activity. The 72-kDa purified proenzyme MMP-2 and the 92-kDa purified proenzyme MMP-9 were used as positive control and standard for the gels (Oncogene Research Products). The zymography bands were digitized using the Gel Doc/Chemic Doc Imaging System, and the data were analyzed using the Quantity One analysis software program (Bio-Rad). Detected activities of positive controls were set at 100%.
The unpaired two-tailed Student t test was used to compare differences. A p value of ≤0.05 was considered statistically significant.
Combination of TNF-α and GM-CSF induces differentiation of CD14+ monocytes into CD14+ DC/Mφ precursors
CD14+ monocytes (>98% purity) isolated from PMBC were cultured in the presence of 10 ng/ml GM-CSF plus 10 ng/ml TNF-α or 10 ng/ml IL-4 for 7 days. Most of the cells cultured with GM-CSF plus TNF-α were converted into round or spindle-shaped adherent Mφ-like cells (Fig. 1,A). These cells expressed high levels of CD14, a marker of monocyte/Mφ, and substantial levels of CD86 (Fig. 1 B) and are termed CD14+ pre-DC/Mφ hereafter. They also expressed HLA-DR, CD1a, CD40, and CD80, but at lower levels than immature DC cultured with GM-CSF and IL-4 for 7 days (IL-4 DC). Furthermore, these CD14+ pre-DC/Mφ stimulated much less proliferation of allogeneic naive CD4 T cells than immature IL-4 DC (data not shown).
The production of TNF-α, IL-12/IL-23p40 (homodimer of p40 subunits), IL-12p70 (heterodimer of p35 and p40 subunits), and IL-23p19/p40 (heterodimer of p19 and p40 subunits) in response to 10 μg/ml LPS is shown in Fig. 2. CD14+ pre-DC/Mφ produced extremely large amounts of TNF (>150 ng/ml at peak), as well as IL-12/IL-23p40 and IL-23p19/p40, but very little IL-12p70. They maintained high levels of IL-12/IL-23p40 and IL-23p19/p40 production for up to at least 5 days after LPS stimulation. In contrast, IL-4 DC produced much smaller amounts of TNF, IL-12/IL-23p40, and IL-23p19/p40, whereas IL-12p70 was produced in far greater amounts. Conclusively, CD14+preDC/Mφ have Mφ-like rather than DC-like characteristics and retain the high responsiveness to LPS reflected in production of TNF-α and IL-23 rather than IL-12.
LPS induces dual differentiation of CD14+ pre-DC/Mφ into mature DC or activated Mφ-like cells
Differentiation of CD14+ pre-DC/Mφ after LPS stimulation was analyzed. Some cells began to convert to DC-like floating cells with extended dendrites (Fig. 3,A) 2 days after LPS stimulation, and the number of floating cells increased to up to one-third of the total at 6 days (Fig. 3,B). Phenotyping of the floating cells showed that these were mature DC, because substantial levels of HLA-DR, CD40, CD70, CD80, CD86, and CD83 were expressed, but only little CD14 (Fig. 3,C). In contrast, adherent cells maintained high levels of CD14 and CD86, but only low-level HLA-DR and CD83; CD70 and CD86 expression was also lower than in DC-like cells (Fig. 3,C). TNF-DC (The floating cells are termed TNF-DC hereafter.) stimulated naive CD4 T cells to the same extent as mature IL-4 DC induced by LPS (Fig. 4 A). Furthermore, TNF-DC retained IFN-γ-inducing capacity much better than mature IL-4 DC (data not shown). In contrast, the adherent cells had typical activated Mφ functions (i.e., excellent phagocytic activity and production of TNF-α) but stimulated proliferation only poorly (data not shown). These results indicate that LPS stimulation induces dual differentiation of CD14+ pre-DC/Mφ to mature DC or activated Mφ.
Comparison of T effectors induced by TNF-DC or IL-4 DC
Cytokine profiles of CD4 T cell stimulated by TNF-DC or mature IL-4 DC were compared (Fig. 4 B). Autologous naive CD4 T cells were added to DC pulsed with KLH and treated with LPS for 3 days. Intracellular cytokines were analyzed by FACS 12 days after starting the DC-T cultures. TNF-DC promoted more Th1-dominant polarization than mature IL-4 DC. T effectors stimulated by TNF-DC contained many cells producing both of IFN-γ and TNF-α, which was not the case for T cells cocultured with mature IL-4 DC. According to the current paradigm for the functions of activated DC (34), the T cell induction capacity of mature IL-4 DC shifts from Th1 to Th2 polarization shortly after LPS stimulation, but the TNF-DC investigated here did not show this increase in Th2 polarization. Even much later after LPS stimulation, Th2 polarization was very limited in TNF-DC (data not shown).
TNF-α stimulation of monocytes determines maturation and Th1-inducing activity of TNF-DC
The effects on subsequent LPS-induced DC maturation and function of TNF-α added into monocyte cultures were analyzed. Because low levels of TNF-α were detected in the cultures of monocytes cultured for 7 days with GM-CSF alone (∼50 pg/ml), we tested monocytes cultured with GM-CSF plus anti-TNF reagents (50 ng/ml sTNFRI) or titrated concentrations of TNF-α (0–5 ng/ml). Cells in these cultures were also converted to floating DC or adherent Mφ after LPS stimulation. Differences between phenotypes of cells cultured in the presence or absence of TNF-α were not apparent. However, after LPS stimulation, expression of DC maturation markers (HLA-DR, CD70, and CD86) were markedly different (Fig. 5,A). Activity was increased in a dose-dependent manner by adding TNF-α into the monocyte culture medium. Likewise, the Th1-inducing capacity of these cells was analyzed. Quantitative analysis of IFN-γ production in combination cultures with naive CD4 T cells also revealed that the Th1-inducing capacity rose in parallel with the dose of TNF-α in the monocyte cultures (Fig. 5,B). FACS also revealed that the addition of TNF-α enhanced Th1-dominant responses induced by DC (data not shown). Supporting these data, expression of mRNA for IL-12 family members (IL-12p35, IL-12/IL-23p40, and IL-23p19) and proinflammatory cytokines (TNF-α, IL-6, and IL-15) 2 days after stimulation was also amplified in the same manner (Fig. 5 C). These results indicate that CD14+ monocytes are primed with TNF-α before LPS-induced maturation into TNF-DC, and this controls the magnitude of Th1 inducing capacity of CD70+ DC. Additionally, TNF-α stimulation of monocytes had to be from the beginning of the culture, having no effect 3 days later.
TNF priming facilitates capacity of DC to induce IL-17 production from resting CD4 T cells
The capacity of TNF-DC to induce IL-17 production by purified CD4 T cells was tested. T cells were extensively depleted of APC or activated T cells because crude T cell fractions contained cells spontaneously producing IL-17. CD4 T cells were cultured alone or with LPS-induced DC pretreated with GM-CSF alone (GM-CSF DC), or GM-CSF plus sTNFRI (anti-TNF DC), or 10 ng/ml TNF (TNF-DC) for 6 days. T cells alone contained only low levels of IL-17-producing cells. T cells stimulated with 1 μg/ml agonistic anti-CD3 Ab and anti-CD28 Ab in the absence of DC did not yield increased ratios of Th17 cells. IFN-γ- and/or IL-17-positive cells cocultured with LPS-stimulated DC are presented in Fig. 6,A. TNF-DC induced three types of T effectors (IL-17+IFN-γ−, IL-17+IFN-γ+, and IL-17−IFN-γ+). In experiments using isotypic control IgG instead of rabbit anti-IL17 Ab, IL-17+ cells were significantly reduced (Fig. 6,A). The increase in all these cell types was facilitated by TNF priming. Similar results were obtained when data on total IL-17-producing cells (IL-17+IFN-γ− and IL-17+IFN-γ+ cells) from four independent experiments were considered (Fig. 6 B). In experiments on Th17 induction from naive T cells (CD4+CD45RA+ >98%), TNF-DC could induce small but significantly increased numbers of these cells (negative control; 0.04 ± 0.03%, n = 8, vs TNF-DC; 0.24 ± 0.05%, n = 8, p < 0.005). These were almost all IL-17+IFN-γ+ (data not shown). Therefore, these data indicate that the TNF-DC had the capacity to induce Th17 responses and that the priming of monocyte with TNF-α facilitated this activity.
Production of MMPs and TNF-α by DC and Mφ
Production of MMP-9 by LPS-stimulated DC or Mφ was analyzed. Conditioned medium from cultures of DC 3 days before and after LPS stimulation were tested in gelatin zymography. In Fig. 7,A, MMP-9 activity is shown to be facilitated in cultures of anti-TNF-DC, GM-CSF DC, and TNF-DC, but not IL-4-DC, after LPS stimulation. In contrast, MMP-2 activity was consistently detected in all samples. Semi quantitative analysis using the Gel Doc/Chemic Doc Imaging System showed that the MMP-9 activity of TNF-DC cultures was ∼5 times of that of IL-4 DC. However, MMP-9 activity of anti-TNF-DC, GM-CSF, and TNF-DC was significantly high. Analysis by real-time RT-PCR revealed that expression of MMP-9 mRNA was high in Mφ, but very low in DC, especially IL-4 DC (Fig. 7,B). These data suggest that large amounts of MMP-9 in cultures of anti-TNF-DC, GM-CSF-DC, and TNF-DC were mainly produced by Mφ. Likewise, TNF-α mRNA expression by Mφ in TNF-DC culture (1640 ± 531, n = 4) was much higher than DC (Fig. 5 B). Mean values were not significantly different from those of other Mφ (data not shown). These data suggest that LPS-stimulated Mφ produced large amounts of MMP-9 and TNF-α in a manner independent of TNF priming.
In this report, we document TNF-dependent differentiation of CD14+ monocytes into mature DC that evoke Th1 and Th17 responses. This takes place in a two-step stimulation model. First, monocytes are stimulated with TNF-α; second, mature DC are induced by LPS stimulation. The first step determined the maturation and functions of the derived DC. The mode of differentiation is different from that of conventional IL-4-DC. This represents a unique role of TNF-α in stimulating monocytes for DC differentiation, in addition to its accepted maturational effect on immature DC (38). The former effect of TNF-α is more important than the latter, as supported by several studies using clinical samples. First, sera of RA patients stimulate monocytes from healthy donors to differentiate into active DC, which is reduced by anti-TNF therapy (30); second, monocyte-DC from patients are more active than those from healthy donors (39); third, a combination of TNF-α and GM-CSF not only enhances DC differentiation (40, 41) but also facilitates autologous MLR (42); Fourth, serum TNF/ sTNFR (endogenous agonists/inhibitors) ratios correlate with the severity of autoimmune disease (43); fifth, injection of anti-TNF reagents into the blood rather than joint space is effective (4). Taken together, TNF-α elevated in patients’ sera may render monocytes susceptible to the TNF priming described here for TNF-DC differentiation. Intriguingly, CD14 positivity of DC precursors in RA synovial fluid (42) is similar to pre-TNF-DC/Mφ, but not IL-4-DC. Further in vivo studies remain necessary to clarify the roles of TNF priming for the differentiation of DC that evoke abnormal inflammatory responses.
Knowledge of how to maintain TNF priming could be important for persistence in Th1 and Th17 responses. This could be provided by large amounts of TNF produced by pre-TNF-DC/Mφ stimulated with LPS. Such phenomena were noted in previous experimental animal models where hemorrhagic necrosis was induced (44). Alternatively, T effectors producing TNF-α or IL-17 induced by TNF-DC could stimulate circulating monocytes and Mφ (45). Such routes for positive feedback stimulation could maintain persistent inflammation (46). Thus, anti-TNF therapy could reduce Th1 and Th17 responses by blocking TNF-dependent differentiation of DC and thence the TNF-dependent positive feedback system. Hence, we propose a model for the differentiation pathway of DC and Mφ as shown in Fig. 8. In this model, inflammation would be terminated once pathogens are cleared in the case of a bacterial infection. Unlike infections, endogenous agonists for TLR-4 could evoke unremitting inflammation (47). Endogenous TLR-4 agonists such as hyaluronan, fibrinogen, fibronectin, high mobility group box 1 protein, and heat shock proteins may be involved in this process (48). However, this model cannot apply to every autoimmune disease such as multiple sclerosis and systemic lupus erythematosus that involve in IFN-dependent immunity (17).
Although induction of human Th-17 by DC has not been reported thus far, TNF-DC induced IL-17 production by highly purified resting CD4 T cells, which did not respond to stimulation with anti-CD3/CD28 Ab. The persistent IL-23 production in response to LPS (Fig. 2) is a feature that would evoke IL-23/IL-17 axis inflammation in TNF-DC differentiation, but not IL-4-DC. Furthermore, the TNF priming enhanced both of IL-23 mRNA expression (Fig. 5,C) and induction of IL-17-producing cells by TNF-DC (Fig. 6). Thus, TNF-α may drive IL-23/IL-17 axis inflammation due to promoting monocytes to differentiate into TNF-DC.
To date, some differences between human and murine system have been found concerning the induction of Th17 cells. Th1-type IL-17 producing cells (IFN-γ+IL-17+ cells) as shown in Fig. 6 can be found in patients with autoimmune disease (49), but not in mice, because IFN-γ inhibits IL-17 production (18, 20). In addition, IL-17 production from human naive CD4 T cells cannot be induced by stimuli as can be done in mice (50). Accordingly, it is suggested that IL-17 induction may be more complex than in mice, and conditions for inducing substantial Ag-specific Th17 cells in humans are largely unknown. Although TNF-DC had potential ability to promote naive CD4 T cells to Th17 cells, the activity was rather low. Because precise definition of these conditions is ultimately crucial for understanding the pathogenesis of autoimmune disease, further studies will be required to delineate specific condition in vitro. In this study, we also investigated monocyte-derived Mφ differentiation by stimulation with LPS after culture with GM-CSF alone or GM-CSF plus anti-TNF or TNF-α. These findings suggested that such differentiated Mφ could evoke abnormal inflammation because they produced large amounts of TNF-α and MMP-9, even, without TNF-priming. However, GM-CSF is known to induce Mφ, termed proinflammatory type 1 Mφ (51), which have important roles in defense against infection (52), but a limited role in adaptive immunity (53). These data are consistent with our results. Namely, GM-CSF plus TNF stimulation of monocytes may form a bridge to adaptive immunity from the innate immunity mediated through TLR-4 signaling. In contrast, TLR-7 or -9 signaling, which leads to IFN-α production by plasmacytoid DC initiates IFN-dependent immunity by generating DC derived form monocytes cultured with GM-CSF plus IFN-α (16). Concerning infectious immunity based on GM-CSF-dependent differentiation to type 1 Mφ, the additional cytokine stimulation of monocytes could modify the mode of differentiation of DC/Mφ appropriate to adaptive immunity against bacterial or viral infection.
CD70 positivity of the TNF-DC and Mφ presented in this report is unusual. CD70+ DC and Mφ are found in lymph nodes in experimental infectious models in mice (54) and in atherosclerotic plaques in humans (55), respectively. Interestingly, CD70+ DC augment CD4 T cell-independent cellular immunity (32, 56) through CD27/CD70 alternative pathway interactions. These CD70+ cells may share functions in infectious immunity or pathogenesis of chronic inflammatory diseases. Further studies are required to clarify the precise details.
In conclusion, this differentiation of TNF-DC has provided a new rationale to explain the mechanism of action of TNF-dependent Th1 and Th17 responses. Hence, the model presented here is useful to better understand TNF immunity and the pathology of disease. In future, support for this model needs to be acquired in animals, but some studies, in particular, regarding IL-17-dependent inflammation must use human materials.
We thank Professors Yoichiro Iwakura, Kouji Matsushima, Seigi Shioda, and Ken Takeda for helpful suggestions and discussion.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Abbreviations used in this paper: RA, rheumatoid arthritis; DC, dendritic cell; Mφ, macrophage; MMP, matrix metalloproteinase; sTNFRI, soluble type I TNF receptor; KLH, keyhole limpet hemocyanin.