Dendritic cells are accessory cells for the development of anti-trinitrophenyl cytotoxic T lymphocytes.

This study establishes that dendritic cells (DC) are the critical accessory cells for the development of anti-trinitrophenol (TNP) cytotoxic T lymphocytes (CTL) in vitro. We developed a model in which nylon wool-nonadherent spleen cells were used both as the responding and stimulating cells, the latter having been TNP-modified and x-irradiated. Thy-1-bearing CTL developed in C57BL/6, B6D2F1, and CBA mice only when small numbers of DC were added. Maximal responses in 5-d cultures were achieved with 0.5-1 DC/100 responding T cells. The DC did not have to be TNP modified directly. Anti-Ia and complement inactivated accessory cells, whereas similar treatment of the responders had no effect. DC exposed to ultraviolet radiation were ineffective, but x-irradiated DC were fully active. Culture media from DC, or from DC-nylon wool-passed spleen T cell cocultures that contained abundant CTL, would not substitute for viable DC. Enriched preparations of macrophages (M phi) were obtained from blood, peritoneal cavity, and spleens of BCG-immune and unprimed mice. M phi added at doses of 0.2-4% were weak or inactive as accessory cells. The level of Ia antigens on test M phi populations was quantitated and visualized by binding of a radioiodinated monoclonal anti-I-Ab,d antibody, clone B-21. M phi that bore substantial amounts of Ia from all organs were weak accessory cells. Addition of M phi to DC-T cell cocultures produced inhibitory effects, usually at a dose of 2% M phi. In contrast, 0.5% Ia-bearing M phi from BCG-immune boosted mice inhibited > 80% of the DC-mediated CTL response. Addition of indomethacin reversed M phi inhibition, and 10(-9) M prostaglandin E2 in turn blocked the indomethacin effect. Indomethacin also restored a low level of accessory cell function in immune-boosted adherent peritoneal cells, but not in preparations of monocytes and spleen M phi. Small numbers of DC were identified in preparations of immune-boosted peritoneal cells and may have accounted for the observed accessory activity. We conclude that the development of anti-TNP CTL is an immune response in which (a) DC are the critical accessory cells; (b) Ia-bearing M phi are weak or inactive; and (c) M phi can inhibit DC-mediated response by an indomethacin-sensitive mechanism.

Dendritic cells are the principal stimulators of the primary mixed leukocyte reaction in mice.

Clone 33D1 is a mouse-rat hybridoma that secretes a specific anti-dendritic cell (DC) monoclonal antibody (14). Because the antibody kills DC in the presence of rabbit complement, it can be used to study the functional consequences of selective DC depletion. Previous data on the cell specificity of 33D1 were first extended. By cytotoxicity (rabbit complement) and indirect immunofluorescence (biotin-avidin technique), 33D1 reacted with DC but not with macrophages nor other splenocytes. In contrast, the monoclonal antibody, F4/80 (15), reacted with macrophages but not DC. The functional assay evaluated in this paper was stimulation of the primary mixed leukocyte reaction (MLR). 33D1 antibody itself did not inhibit stimulation by enriched populations of DC. In the presence of complement, 33D1 killed DC and ablated stimulatory function. The effect of 33D1 and complement on MLR stimulation by heterogenous cell mixtures was then evaluated. Removal of DC from unfractionated spleen suspensions reduced stimulatory capacity 75-90 percent, comparable to that produced with specific anti-Ia antibody and complement. Stimulation of both proliferative and cytotoxic responses was reduced. DC depletion had similar effects on MLR generated across full strain differences, or across selected subregions (H2I, H-2K/D) of the major histocompatibility complex. To further compare the functional properties of spleen DC and macrophages, MLR stimulation by adherent and nonadherent fractions of spleen were tested separately. 62 +/- 8 percent of the total stimulatory capacity of spleen was in the plastic adherent population. Activity was ablated greater than 90 percent after elimination of DC. MLR stimulation by 24-h cultures of spleen adherent cells, which contained a three- to sixfold excess of Ia(+) macrophages, was also ablated when DC were removed. Stimulation by nonadherent spleen was more resistant, but was reduced 50-75 percent by 33D1 and complement. The function of spleen cells treated with 33D1 or anti-Ia antibody and complement was restored with a small inoculum of purified DC. The latter corresponded to 0.5 percent of total stimulator cells and were enriched by previously described techniques that did not require the 33D1 antibody. We conclude that the DC, a trace component of mouse spleen, is the principal cell type required for stimulation of the primary MLR. Because other cells are not immunogenic, but do express Ia and H-2 alloantigens, DC likely represent the critical accessory cell required for the induction of lymphocyte responses.

Studies of the cell surface of mouse dendritic cells and other leukocytes.

The surface of dendritic cells (DC) has been analyzed by means of monoclonal antibodies (Ab) and lactoperoxidase (LPO)-mediated radioiodination. Antigens and other exteriorily disposed polypeptides of purified spleen DC were compared with those of tissue macrophages (Mphi), monocytes, and other bone marrow-derived elements. Quantitative binding studies and autoradiography with (125)I-Ab established that DC expressed high levels of I-A and H-2D, 2 x 10(5) and 1 x 10(5) Ab binding sites per cell, respectively. DC from conventional, germ-free, and specific pathogen-free mice were all rich in Ia. Expression of Ia on B cells was 5-10 percent of that on DC and increased fivefold during lipopolysaccharide mitogenesis. More than 70-90 percent of purified Mphi and monocytes from specific pathogen-free mice were Ia negative, but increased levels of Ia were noted on cells from mice reared under conventional conditions. Thus large amounts of Ia on DC is a constitutive trait, whereas the expression of Ia by other cell types may be governed by the environmental and immunological status of the host. The 2.4G2 Fc receptor Ag was not detected on DC. Peritoneal and spleen Mphi had 10(5) 2.4G2 binding sites/cell, whereas monocytes and lymphocytes were less reactive (1 x 10(4)-3 x 10(4) binding sites/cell). Four other Mphi-related antigens were evaluated. Each had a distinctive tissue distribution and none bound exclusively to Mphi and monocytes. Neither 1.21J (Mac-1) nor F4/80 reacted with DC. Immunoprecipitation studies of externally ((125)I) and biosynthetically ([(35)S]methionine)dabeled cells confirmed the binding data. Sensitive binding assays with (125)I-Ab confirmed previous observations that DC lack Ig and Thy-1. Lyt-1 was also not found on DC, but 5-12 percent of the cells in purified DC preparations expressed both Lyt-2 and Ia. All DC expressed the leukocyte common antigens at levels similar to other leukocytes. The spectrum of surface polypeptides labeled by LPO-mediated iodination was different on Mphi, DC, and lymphocytes. Polypeptides migrating at molecular weights of 155,000, 85,000, and 62,000 appeared to be restricted to DC. These observations establish that the cell surface of DC differs considerably from other leukocytes, including the blood monocyte, and suggest that the DC is part of a unique Ia-rich leukocyte differentiation pathway.

A monoclonal antibody specific for mouse dendritic cells.

Dendritic cells (DC) are a small subpopulation of lymphoid cells with distinctive cytologic features, surface properties, and functions. This report describes the DC-specific antibody (Ab) secreted by clone 33DI. Rat spleen cells immune to mouse DC were fused to the P3U myeloma. Hybrid culture supernatants were screened simultaneously against DC, a macrophage (M phi) cell line, and mitogen-stimulated lymphoblasts. 33DI Ab specifically killed 80-90% of DC from spleen and lymph node, but no other leukocytes, including Ia+ and Ia- M phi (Ia, I-region-associated antigen,). Quantitative binding studies with 5H-labeled 33D1 Ab showed that DC had an average of 14,000 binding sites per cell. Binding to DC was inhibited with Fab fragment of 33D1 Ab but not with a panel of other monoclonal antibodies, including anti-Ia Ab. Adherence and flotation procedures that enriched for DC enriched for 3H-labeled 33D1 Ab binding in parallel. 33D1 antigen was not detectable on: M phi from spleen, peritoneal cavity, and blood; three M phi cell lines; lymphocytes; granulocytes; platelets; and erythroid cells. DC continued to express the 33D1 antigen after 4 days in culture, whereas M phi and lymphocytes did not acquire it. Quantitative and autoradiographic studies confirmed that spleen and lymph node suspensions contain less than 1% DC. We conclude that 33D1 Ab detects a stable and specific DC antigen and can be used to monitor DC content in complex lymphoid mixtures.