GM-CSF, IL-1 alpha, IL-1 beta, IL-6, IL-8, IL-10, ICAM-1 and VCAM-1 gene expression and cytokine production in human duodenal fibroblasts stimulated with lipopolysaccharide, IL-1 alpha and TNF-alpha.

The role of mucosal fibroblasts in intestinal inflammatory reactions is not known. In this study, we demonstrate that fibroblasts grown from histologically normal human duodenal biopsy tissues expressed mRNA genes for granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-1 alpha, IL-1 beta, IL-6, IL-8, IL-10, intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) when stimulated with lipopolysaccharide (LPS) or IL-1 alpha. The increased mRNA expression of GM-CSF, IL-1 alpha, IL-1 beta, IL-6 and IL-8 in response to IL-1 alpha and LPS stimulation was time- and dose-dependent. In contrast, IL-10 was weakly expressed when fibroblasts were stimulated with LPS, IL-1 alpha or tumour necrosis factor-alpha (TNF-alpha), but the expression was enhanced in the presence of cycloheximide combined with optimal concentrations of LPS, IL-1 alpha or TNF-alpha, IL-1 alpha was a more potent stimulator than LPS for GM-CSF, IL-6, IL-8 and IL-10 expression, but not for IL-1 alpha and IL-1 beta. Increased GM-CSF, IL-6 and IL-8 gene expression was associated with the production of cytokine proteins in culture supernatant, but IL-1 alpha and IL-1 beta remained undetectable. Dexamethasone suppressed both gene expression and protein production of GM-CSF, IL-6 and IL-8 when fibroblasts were exposed to IL-1 alpha. TNF-alpha stimulated the release of GM-CSF, IL-6 and IL-8 and, combined with IL-1 alpha, cytokine production was enhanced synergistically. Finally, both LPS and IL-1 alpha up-regulated ICAM-1 and VCAM-1 gene expression. These findings implicate duodenal fibroblasts in the initiation and/or regulation of intestinal inflammation.     

Direct and mononuclear cell mediated effects on interleukin 6 production by glioma cells in infection with herpes simplex virus type 1

To clarify the mechanism of interleukin (IL)-6 elevation in the cerebrospinal fluid of viral meningitis and/or encephalitis patients, we investigated how herpes simplex virus type 1 (HSV1)-infection enhances IL-6 production in human glioma cells (the U373MG and T98G cells). Although human glioma cells did not show enhanced IL-6 production by direct HSV1-infection, the cell-free supernatant from HSV1-stimulated mononuclear cells (MNC) culture and lipopolysaccharide, as a positive control, markedly elevated IL-6 production at both mRNA and polypeptide levels. Ultra violet-irradiated HSV1 induced the secretion of the IL-6 inducing factor(s) from MNC, whereas heat-inactivated HSV1 did not show this activity. This finding indicated that the adsorption of virus on the surface of MNC may be sufficient for induction of secretion. The supernatant from the culture of HSV1-stimulated MNC contained detectable amounts of IL-1beta, tumor necrosis factor (TNF) alpha, interferon (IFN) gamma and IL-6, and its IL-6-inducing activity was inhibited only by anti-IL-1beta antibodies. Moreover, recombinant IL-1beta markedly enhanced IL-6 production in glioma cells with a concomitant elevation of its mRNA level. Taken together, the results suggest that in HSV1-infection of the CNS, enhancement of IL-6 production in glial cells is mediated not by direct infection to glial cells but rather by IL-1beta released from HSV1-stimulated MNC. These findings may help elucidate the mechanisms underlying cerebro-parenchymal inflammatory progression and repair in herpes simplex encephalitis.     

Lipopolysaccharides of Bacteroides intermedius (Prevotella intermedia) and Bacteroides (Porphyromonas) gingivalis induce interleukin-8 gene expression in human gingival fibroblast cultures.

Lipopolysaccharides (LPS) prepared from Bacteroides intermedius (Prevotella intermedia) and Bacteroides (Porphyromonas) gingivalis by hot phenol-water extraction induced interleukin-8 (IL-8) mRNA in normal human gingival fibroblast cultures, as demonstrated by Northern (RNA) blot analysis. IL-8 mRNA levels began to increase after a 2-h exposure, reached a maximum after 12 h, and then dropped to the unstimulated level at 48 h. IL-8 mRNA levels were also enhanced in a dose-dependent manner. By contrast, LPS specimens from various Salmonella species with S and R chemotypes and bacterial [corrected] and synthetic lipid A preparations did not increase IL-8 mRNA levels in fibroblasts. Although recombinant human IL-1 alpha induced IL-8 mRNA expression in fibroblast cultures, an antiserum to recombinant human IL-1 alpha did not decrease the IL-8 mRNA accumulation induced by B. intermedius LPS. Fibroblasts primed with natural human gamma interferon (IFN-gamma) expressed higher IL-8 mRNA levels upon stimulation with B. intermedius LPS, but not with Salmonella LPS, compared with nontreated cells. Natural human IFN-beta exhibited a similar priming effect on the fibroblasts, and antiserum to IFN-beta added to the cultures together with B. intermedius LPS decreased the IL-8 mRNA levels. Therefore, endogenous IFN-beta enhanced IL-8 mRNA production in response to B. intermedius LPS in fibroblasts.     

Cytokine inhibition by a novel steroid, mometasone furoate

Mometasone furoate (9 alpha, 21 dichloro-11 beta, 17 alpha dihydroxy-16 alpha methyl-1,4 pregnadiene-3, 20 dione-17-[2'] furoate) was an unexpectedly potent inhibitor of the in vitro production of three inflammatory cytokines, IL-1(1), IL-6, and TNF-alpha. The potency of mometasone furoate in inhibiting cytokine production was compared to that of hydrocortisone, betamethasone, dexamethasone, and beclomethasone. IL-6 and TNF-alpha were both produced by WEHI-265.1 (murine myelomonocytic leukemia) cells following stimulation by lipopolysaccharide (LPS). Twenty-four hours after stimulation by LPS, the cell-free supernatant fluids were removed. Their cytokine content was analyzed using ELISAs specific for each cytokine. IL-1 synthesis was induced in the harvested peritoneal macrophages of BALB/c mice by incubation with LPS for twenty-four hours. The IL-1 content in the cell-free supernatant fluids was determined by the thymocyte-costimulator bioassay. Using these systems, mometasone furoate was found to be the most potent steroid tested for inhibiting the production of the three cytokines. The IC50's were 0.05 nM (IL-1), 0.15 nM (IL-6), and 0.25 nM (TNF-alpha). The inhibition of the production of proinflammatory mediators by extremely low concentrations of mometasone furoate suggests that this steroid should be highly effective in various disorders.     

Normal human alveolar macrophages have more ability than blood monocytes to produce cell-associated interleukin-1-alpha

Human alveolar macrophages (AM) were obtained by bronchoalveolar lavage from healthy donors, and their abilities to produce extracellular and cell-associated interleukin 1 (IL-1) in response to various activation stimuli were compared with those of autologous blood monocytes. The production of IL-1 alpha and IL-1 beta by monocytes and AM was examined by thymocyte co-stimulation assay and enzyme immunoassays (EIA). Results showed that when activated with lipopolysaccharide (LPS) or desmethyl muramyl dipeptide (norMDP), AM released much less extracellular IL-1 beta than did blood monocytes. In contrast, these activated AM produced more cell-associated IL-1 than did blood monocytes. When the IL-1 activity was examined by the thymocyte assay, the extracellular and cell-associated IL-1 produced by the two cell types were largely IL-1 beta and IL-1 alpha, respectively, as shown by antibody neutralization. The cell-associated IL-1 activity of AM induced by the synergistic actions of suboptimal concentrations of recombinant interferon-gamma (rIFN-gamma) and norMDP was also higher than that of autologous blood monocytes. Consistent with these findings on AM, macrophages generated in vitro by maturation of blood monocytes produced higher levels of cell-associated IL-1 activity than did freshly isolated monocytes. These observations suggest that AM may play a critical role in situ regulation of pulmonary inflammatory and immune reactions through production of cell-associated IL-1 alpha.     

Effect of interferon on the induction of human monocyte secretion of interleukin-1 activity.

This study investigates the effect of interferons on the induction of human monocyte secretion of interleukin-1 (IL-1) activity by lipopolysaccharides (LPS). Monocytes do not spontaneously produce IL-1 and the addition of interferons to the culture does not lead to detectable secretion. Addition of LPS alone induces the release of measurable amounts of IL-1 activity. The addition of low doses (1-10 units/ml) of alpha, beta, or gamma interferon to the LPS-stimulated cultures further increases this secretion by 50%. The addition of 1000 units/ml of alpha or beta interferon leads to inhibition of IL-1 release. By contrast, gamma interferon is a dose dependent enhancer of IL-1 release. The effect of gamma interferon is on the production of IL-1 and is not an enhancement of IL-1 activity in the biological assay. Results demonstrate that addition of gamma interferon to monocytes increases the rate of secretion of IL-1 by these cells. Gamma interferon also appears to abrogate the loss in the ability of monocytes to produce IL-1 activity after overnight culture. This last result parallels the maintainence of the expression of the HLA-DR surface marker on monocytes by gamma interferon. These results may help define a mechanism involving IL-1 generation which could have bearing on the in vivo pyrogenic effects of purified gamma interferon.     

Comparison of Interleukin 1 Release and Interleukin 1 mRNA Expression of Human Monocytes Activated by Bacterial Lipopolysaccharide or Synthetic Lipid A

Several of the biological effects of bacterial lipopolysaccharide (LPS) can be induced by the lipid A part of the molecule. Here we show that in human peripheral blood monocytes, synthetic E. coli lipid A is as effective as the whole LPS molecule in inducing the production of interleukin 1 (IL-1) bioactivity which remains associated to the cells (i.e. IL-1 alpha). In contrast, LPS- but not lipid A-stimulated cells released the bioactive IL-1 produced into the culture supernatant (mainly IL-1 beta). Northern blotting analysis demonstrated, however, that LPS and lipid A are equally effective in inducing the accumulation of IL-1 alpha and IL-1 beta mRNA. These data support the hypothesis that induction of IL-1 biosynthesis and activation of the secretory mechanism for IL-1 are independent phenomena.     

Interleukin-1 and tumor necrosis factor synergistically stimulate lung fibroblast interleukin-1 alpha production

We determined whether normal human lung fibroblasts expressed cell-associated thymocyte-stimulating activity in response to recombinant interleukin-1 (rIL-1) (alpha and beta) and recombinant tumor necrosis factor (rTNF). Individually, rIL-1 and rTNF induced fibroblast expression of thymocyte-stimulating activity, with rIL-1 being significantly more potent. Importantly, combining rIL-1 and rTNF resulted in a synergistic increase in fibroblast thymocyte-stimulating activity. This synergistic interaction was dose dependent for both cytokines and was not noted when gamma-interferon was combined with rIL-1 or rTNF. In all cases, the thymocyte-stimulating activity was the result of an IL-1 alpha-like moiety whose maximal production required protein synthesis. IL-1 alpha activity could be detected after as little as 4 h, peaked after 24 h, and returned toward normal with longer periods of cytokine-fibroblast incubation. However, cytokine-stimulated fibroblasts that no longer expressed IL-1 alpha activity could be induced to re-express this activity with repeat cytokine challenge. Induction of fibroblast IL-1 alpha by IL-1 and/or TNF may be an important mechanism amplifying IL-1-mediated biologic events at sites of local inflammation.     

Biological Characterization of Lipopolysaccharide from Treponema pectinovorum

This study investigated the endotoxic and biological properties of purified lipopolysaccharide (LPS) isolated from an oral spirochete, Treponema pectinovorum. Endotoxicity, measured by Limulus amoebocyte lysate kinetic assay, showed that the LPS contained 1.28 endotoxin units per microg of purified LPS, which was approximately 4,000 times less than Escherichia coli O55:B5 LPS. To determine in vivo endotoxicity, LPS responder mice were administered LPS following galactosamine (GalN) sensitization. The LPS induced neither endotoxic symptoms nor lethality for 96 h, suggesting negligible or very low endotoxicity. In contrast, infection with live T. pectinovorum induced 100% lethality within 12 h in GalN-sensitized LPS responder mice, indicating an endotoxin-like property of this treponeme. Heat-killed microorganisms exhibited no lethality in GalN-sensitized mice, suggesting that the endotoxicity was associated with heat-labile components. To determine cytokine and chemokine induction by LPS, human gingival fibroblasts were stimulated and secretion of interleukin 1beta (IL-1beta), granulocyte-macrophage colony-stimulating factor, gamma interferon, IL-6, IL-8, and monocyte chemoattractant protein 1 (MCP-1) was assessed. The purified LPS induced significant amounts of only IL-6, IL-8, and MCP-1, although they were substantially lower than levels after challenge with live T. pectinovorum. After injection of LPS or live or heat-killed T. pectinovorum, serum was collected from mice and analyzed for proinflammatory cytokines IL-1beta, tumor necrosis factor alpha (TNF-alpha), and IL-6. LPS induced only IL-6 consistently. Both live and heat-killed T. pectinovorum induced serum IL-6, which was higher than the level detected following LPS administration. Importantly, live bacteria elicited systemic TNF-alpha and IL-1beta levels similar to those induced by a lethal dose of live E. coli O111. The results indicated that T. pectinovorum LPS has very low or no endotoxicity, although it can elicit low levels of cytokines from host cells. In contrast to the LPS, live T. pectinovorum demonstrated in vivo toxicity, which was associated with serum IL-1beta, TNF-alpha, and IL-6, suggesting an endotoxin-like property of a heat-labile molecule(s) of the spirochete.     

Enhanced interleukin-1beta, interleukin-6 and tumor necrosis factor-alpha production by LPS stimulated human monocytes isolated from HIV+ patients

Periodontal disease and tooth loss is a common finding among advanced HIV+ patients. In addition to local oral lipopolysaccharide (LPS) stimulation, systemic up-regulation of monocyte pro-inflammatory cytokine secretion may also be involved in the pathogenesis of HIV disease. A study was undertaken to investigate IL-1beta, IL-6 and TNF-alpha production by resting and LPS stimulated monocytes isolated from HIV+ patients and also to investigate the relationship of the patient's HIV viral load status to the cytokine production. Whole blood samples in EDTA were collected from 39 HIV-1 infected patients and 20 age and sex matched uninfected controls. Plasma was separated by centrifugation. Viral load was determined using a quantitative RT-PCR. Monocytes were isolated by Ficoll-hypaque gradient separation followed by overnight plastic adherence. Cultured monocytes (1x10(6)/ml) were stimulated with LPS (1 microg/ml) of either P. gingivalis or F. nucleatum for 2, 8, 24 and 48 h and supernatant fluids were collected. IL-1beta, IL-6, and TNF-alpha levels in supernatant fluids were estimated by ELISA. Increased overall production of IL-1beta, IL-6 and TNF-alpha by LPS stimulated monocytes isolated from HIV-1 infected patients was observed when compared to HIV-1 uninfected controls. LPS stimulated monocytes from HIV-1 infected patients with high viral load (HVL) produced significant (p<0.05) elevations in these pro-inflammatory cytokines when compared to HIV-1 uninfected controls. Both LPS of P. gingivalis and F. nucleatum produced a comparable cytokine production by monocytes after 8 h of stimulation. These data suggest that enhanced IL-1beta, IL-6 and TNF-alpha is produced by monocytes/macrophages isolated from HVL HIV+ patients and may be involved in the overall pathogenesis of HIV-1 infection.     

Abnormal production of interleukin-1 by microglia from trisomy 16 mice.

The production of interleukin-1 (IL-1) was examined in cultured CNS microglia obtained from trisomy 16 (Ts16) fetal mouse brain, a model system for studies relevant to Down syndrome (DS). When compared to microglia from their normal littermates, Ts16 microglia produced significantly higher levels of IL-1 activity both before and following stimulation with lipopolysaccharide (LPS). IL-1 release was stimulated by alpha/beta interferon (IFN) in the normal but not Ts16 microglial cultures. The overall level of IL-1 production in normal littermates, however, was still less than that seen in Ts16. Thus, microglia from Ts16 mice may function in an inappropriate manner and, if this abnormality occurs in vivo, may have wide ranging effects on a developing nervous system.     

Induction of Interleukin 1α (IL-1α) and IL-1β mRNA Expression and Cellular IL-1 Production by Anti-HLA-DR Antibodies in Human Monocytes

We studied the role of HLA class II antigens in the regulation of interleukin 1 (IL-1) production in human monocytes. Monocytes were cultured with monoclonal anti-HLA-DR antibodies for 24 h after which cellular (i.e. intracellular and membrane-associated) IL-1 production, IL-1 secretion, and the expression of IL-1 alpha and IL-1 beta mRNA were determined. One of the anti-HLA-DR antibodies tested (anti-HLA-DR, Becton Dickinson) clearly induced IL-1 alpha and IL-1 beta mRNA expression and cellular IL-1 production. The other anti-HLA-DR antibody tested (OKIa1, Ortho) had no effect on IL-1 production. The stimulatory effect of anti-HLA-DR was enhanced by IFN-gamma in both fresh and aged monocytes. A synergistic effect by anti-HLA-DR and suboptimal doses of LPS (1 ng/ml) on both cellular IL-1 production and secretion was also demonstrated. The possibility of contaminating LPS causing the IL-1-inducing effect of anti-HLA-DR was excluded by the inability of polymyxin B to abolish the anti-HLA-DR-induced IL-1 production.     

Selective enhancement of interleukin 1 beta production in myelomonocytic cell lines by insulin and its related cytokines.

The production of interleukin 1 (IL-1) by lipopolysaccharide (LPS)-stimulated myelomonocytic cell lines ML-1, THP-1 and PL-21 was significantly enhanced by the addition of insulin, insulin-like growth factor (IGF)-I or IGF-II into the cell cultures. The IL-1 activity in the supernatants from cell cultures stimulated with LPS and insulin was completely neutralized by anti-IL-1 beta antibody. Anti-IL-1 alpha antibody had no inhibitory effect. Insulin itself did not stimulate IL-1 beta production directly, but increased it in the mitogen activated cells. However, insulin had no enhancing effect on the production of IL-1 alpha by human T cell lymphotropic virus-I (HTLV-I)-infected T cell lines or on IL-2 production by mitogen-stimulated leukemia T cell lines. Thus, insulin and its related cytokines are shown here as other molecules selectively modulating the production of IL-1 beta in myelomonocytic cell lines.     

Human gingival fibroblasts rescue butyric acid-induced T-cell apoptosis

We previously demonstrated that butyric acid, an extracellular metabolite from periodontopathic bacteria, induces cytotoxicity and apoptosis in murine thymocytes, splenic T cells, and human Jurkat T cells. In this study, we used a cell-to-cell interaction system to examine the contribution of gingival fibroblasts to the regulation of T-cell death induced by butyric acid. Butyric acid slightly suppressed fibroblast viability in a concentration-dependent fashion. However, DNA fragmentation assays indicated that butyric acid did not induce apoptosis for up to 21 h in human gingival fibroblasts (Gin 1, F41-G, and H. pulp cells). The culture supernatants were assayed for interleukin 1alpha (IL-1alpha), IL-1beta, IL-6, IL-8, IL-11, tumor necrosis factor alpha, and transforming growth factor beta, but only the IL-6, IL-8, and IL-11 levels were significantly increased by addition of butyric acid. Butyric acid- or Fas-induced Jurkat-cell apoptosis was attenuated when Jurkat cells were cocultured with either F41-G or Gin 1 cells that had been preincubated for 6 h with butyric acid. IL-8 slightly stimulated butyric acid- or Fas-induced Jurkat-cell apoptosis in a dose-dependent manner, although a low dose of IL-8 had a mildly inhibitory effect on apoptosis. In contrast, IL-6 and IL-11 significantly suppressed butyric acid- or Fas-induced apoptosis in a dose-dependent fashion. Furthermore, the addition of monoclonal antibodies against human IL-6 and IL-11 to cocultures of gingival fibroblasts and Jurkat cells partially eliminated T-cell recovery. These results suggest that the proinflammatory cytokines such as IL-6 and IL-11, produced in fibroblasts stimulated with butyric acid, are involved in the attenuation of T-cell apoptosis by gingival fibroblasts.     

Interleukin-12-stimulated natural killer cells can activate human macrophages to inhibit growth of Mycobacterium avium.

Interleukin-12 (IL-12) is a critical cytokine that affects many of the biological functions of NK cells and T cells. We have previously shown that both human and murine NK cells are important in host defense against Mycobacterium avium complex and act by secreting cytokines that induce macrophages to inhibit the growth of intracellular M. avium. To define the role of IL-12 in M. avium complex infection, we stimulated human NK cells with recombinant human IL-12 at 0.01 to 1 ng/ml for 24 h and used the tissue culture supernatant to treat human monocyte-derived macrophage monolayers infected with M. avium. IL-12 had no direct effect on M. avium-infected macrophages, but culture supernatant from IL-12-treated NK cells activated macrophages to inhibit the growth of intracellular M. avium in a dose-dependent manner. Stimulation of NK cells with IL-12 in combination with tumor necrosis factor alpha (TNF-alpha) or IL-1 increased the ability of supernatant from NK-cell culture to limit M. avium growth within macrophages, compared with that of culture supernatant from IL-12-treated NK cells. Results with supernatant from nonstimulated NK cells were similar to those with supernatant from untreated controls. Treatment of supernatant from IL-12-stimulated NK cells with anti-TNF-alpha, anti-granulocyte-macrophage colony-stimulating factor, but not anti-gamma interferon antibodies decreased the ability of NK-cell supernatant to induce anti-M. avium activity in infected macrophages. Treatment of macrophage monolayers with anti-transforming growth factor beta antibody before adding supernatant from IL-12-stimulated NK cells was associated with an increase of anti-M. avium activity compared with that of supernatant from IL-12-treated NK cells. These results suggest that IL-12 has a role in host defense against M. avium and that the effect of IL-12 is dependent chiefly on TNF-alpha and granulocyte-macrophage colony-stimulating factor.     

Capsular polysaccharide of Cryptococcus neoformans induces proinflammatory cytokine release by human neutrophils.

Human polymorphonuclear leukocytes from normal subjects produced proinflammatory cytokines in response to stimulation with Cryptococcus neoformans yeast cells. The cytokines released after stimulation of neutrophils included interleukin-1beta (IL-1beta), IL-6, IL-8 and tumor necrosis factor alpha. The magnitude of the cytokine response was related to the yeast capsule size. Cells of a large-capsule isolate stimulated release of greater amounts of cytokine than did a thinly encapsulated isolate, which, in turn, stimulated release of greater amounts of cytokine than an acapsular isolate. Cytokine release was also stimulated by supernatant fluids from cryptococcal cells that were preincubated with 10% human serum, suggesting the generation of a soluble mediator. The major capsular polysaccharide, glucuronoxylomannan, stimulated release of tumor necrosis factor alpha, IL-1beta, and IL-8 in a dose-dependent fashion. These results differ from previous studies of cytokine secretion by human monocytes in several important respects, including the importance of encapsulation in stimulation of cytokine secretion and the ability of purified glucuronoxylomannan to induce cytokine secretion.     

Extracts of scabies mites (Sarcoptidae: Sarcoptes scabiei) modulate cytokine expression by human peripheral blood mononuclear cells and dendritic cells

We performed a series of experiments to determine if human peripheral blood mononuclear cells (PBMCs) from a healthy donor and dendritic cells (NHDCs) derived from these PBMCs reacted to molecules in a scabies extract. PBMCs extravasate from the circulatory system and enter tissues such as scabietic lesions, where monocytes become macrophages. Cells were cultured in medium alone or medium containing 50 microg/ml of Sarcoptes scabiei (SS) extract, 50 ng/ml E. coli lipopolysaccharide (LPS), or SS + LPS together. Supernatants were collected and assayed by enzyme-linked immunosorbent assay (ELISA) for specific cytokines. PBMCs stimulated with SS or LPS exhibited moderately upregulated production of interleukin (IL)-1beta and huge increases in secretions of IL-6, IL-8 and TNF-alpha. Cells co-stimulated with both SS and LPS generally secreted more of these cytokines than cells stimulated with either SS or LPS alone. LPS induced a small amount of IL-1alpha secretion, whereas SS did not, and neither additive resulted in the production of IL-10. NHDCs did not produce IL-1alpha, IL-1beta, IL-6, IL-8, or IL-10 in response to stimulation with SS. These cells did produce IL-6, IL-8, and tumor necrosis factor (TNF)-alpha in response to LPS. When cells were co-stimulated with both LPS and SS, the production of IL-6 and IL-8 was significantly reduced compared with the levels secreted after LPS stimulation alone. These studies show that molecules in a whole body extract of S. scabiei modulate the function of PBMCs (probably monocytes) and dendritic cells.