Single cell suspensions of rat lymphoid and nonlymphoid tissues were fractionated on discontinuous gradients of bovine serum albumin into high density and low density subfractions. In general, accessory activity required for responses of periodate-treated T lymphocytes was recovered only in a low density population containing a small percent of the total fractionated cells from lymph nodes, spleen, liver, skin, and peritoneal exudates. Further purification always led to an increase of both accessory activity and number of dendritic cells present in nonrosetting and nonadherent populations. After purification, a high recovery of the total accessory activity was found in fractions that contained a high percentage of dendritic cells resulting in a more than 1,000-fold enrichment in accessory activity per cell. No other fraction obtained during the purification contained significant accessory activity. In all cases, macrophage-enriched populations lacked accessory cell activity. With the exception of peritoneal exudate cell preparations, which contained an inhibitory cell, the level of accessory activity in a given population was always found to be a function of the number of dendritic cells present. Dendritic cells from all sources were nonadherent, nonphagocytic, radio- resistant, and nonspecific esterase negative. They expressed Ia antigens and lacked Fc receptors. Both epidermal and lymph node dendritic cells contain Birbeck granules, subcellular structures previously described only for Langerhans cells. Accessory activity requires viable dendritic cells but is unaffected by 1,000 rad of γ-irradiation. However, ultraviolet irradiation abolished the activity of accessory cells. The cells that responded to periodate were IgG-negative T cells, whereas IgG-positive B cells could not be stimulated under the same conditions. Only periodate-treated T cells and dendritic cells were needed for responses to occur; removal of virtually all macrophages from these purified preparations had no effect. Dendritic cells were also required as stimulators in mixed leukocyte cultures, whereas macrophages, even though Ia positive, were inert.

(125)I-labelled asialo-fetuin, administered intravenously, rapidly accumulates in rat liver and the radioactivity is subsequently cleared from the liver within 60min. Plasma radioactivity reaches a minimum between 10 and 15 min after injection and rises slightly during the period of liver clearance. Free iodide is the only radioactive compound found in plasma during this latter period. Fractionation of rat liver at 5 and 13min after injection of (125)I-labelled asialo-fetuin supports the hypothesis that asialo-glycoprotein is taken into liver by pinocytosis after binding to the plasma membrane and is then hydrolysed by lysosomal enzymes. At 5min, radioactivity was concentrated 23-fold in a membrane fraction similarly enriched in phosphodiesterase I, a plasma-membrane marker enzyme, whereas at 13min the radioactivity appeared to be localized within lysosomes. Separation of three liver fractions (heavy mitochondrial, light mitochondrial and microsomal) on sucrose gradients revealed the presence of two populations of radioactive particles. One population banded in a region coincident with a lysosomal marker enzyme. The other, more abundant, population of radioactive particles had a density of 1.13 and contained some phosphodiesterase, but very little lysosomal enzyme. These latter particles appear to be pinocytotic vesicles produced after uptake of the asialo-fetuin bound by the plasma membrane. Lysosomal extracts extensively hydrolyse asialo-fetuin during incubation in vitro at pH4.7 and iodotyrosine is completely released from the iodinated glycoprotein. Protein digestion within lysosomes was demonstrated by incubating intact lysosomes containing (125)I-labelled asialo-fetuin in iso-osmotic sucrose, pH7.2. The radioactive hydrolysis product, iodotyrosine, readily passed through the lysosomal membrane and was found in the external medium. These results are not sufficient to account for the presence of free iodide in plasma, but this was explained by the observation that iodotyrosines are deiodinated by microsomal enzymes in the presence of NADPH.

Transformation of T lymphocytes, induced by treatment with periodate or with neuraminidase plus galactose oxidase, requires the participation of accessory cells. Procedures were developed for the fractionation of rat lymph node cells, by which most of the lymphocytes can be recovered as a major population of cells that do not respond to mitogenic stimulation unless accessory cells from a separated minor population are added. Further purification led to a 1000-fold overall increase in accessory activity per cell, with a 50-70% yield. The purest preparations were virtually free of macrophages and contained more than 90% typical dendritic cells. Maximum responses occurred at a ratio of only one dendritic cell per 200 periodate-treated lymphocytes. This evidence thus indicates strongly that in rats, dendritic cells--not macrophages--function as accessory cells. Further, the number of dendritic cells in a preparation governed the magnitude of the mitogenic response and was limiting in the case of unfractionated lymph node cells. In addition, when oxidized with periodate or with neuraminidase plus galactose oxidase, the dendritic cell served as a very potent indirect stimulator of untreated responder lymphocytes. Both functions of the dendritic cell appeared to lack species specificity, since mouse dendritic cells were very active when tested with rat responder lymphocytes. A soluble factor (accessory cell-replacing factor), produced by cultures of lymph node or spleen cells subjected to oxidative mitogenesis, enabled otherwise unresponsive mitogen-treated lymphocytes to respond. Dendritic cells were required for the production of this factor but may not be solely responsible for its production.

No abstract

The biosynthesis of carnitine in the rat was studied by following the metabolism of two radioactive derivatives of asialo-fetuin. The first contained 14C-labelled methyl groups covalently bound to the 6-N-amino fraction of its lysine residues as 6-N-monomethyl- and dimethyl-lysine. By treating this protein with iodomethane, a second derivative was produced in which the radioactivity was preferentially incorporated as 6-N-[Me-14C]-trimethyl-lysine. These desialylated glycoproteins, like other asialo-proteins, were immediately cleared from the blood by rat liver. Within hepatocyte lysosomes, the 14C-labelled proteins were rapidly hydrolysed, producing free amino acids containing the various 6-N-[Me-14C]methylated lysine residues. The radioactive amino acids crossed the lysosomal membrane and were further metabolized in the cytosol. Carnitine was the major radioactive metabolite detected in extracts of the rat carcass and liver after intravenous injection of 6-N-[Me-14C]trimethyl-lysine-labelled asialo-fetuin. Within 3h, at least 34.6% of the trimethyl-lysine in the administered protein was converted into carnitine. Similarly, an isolated perfused rat liver converted 30% of the added peptide-bound trimethyl-lysine into carnitine within 90 min. On the other hand, in numerous attempts we failed to detect radioactive carnitine in both rat liver and carcass between 20 min and 22 h after injection of 6-N-[Me-14C]-monomethyl- and -dimethyl-lysine-labelled asialo-fetuin. These data provide evidence for a pathway of carnitine biosynthesis that involves trimethyl-lysine as a peptide-bound precursor as proposed by R.A. Cox & C.L. Hoppel [(1973) Biochem. J. 136, 1083-1090] and V. Tanphaichitr & H.P. Broquist [(1973) J. Biol. Chem. 248, 2176-2181]. The findings also show that rat liver can synthesize carnitine without the aid of other tissues, but cannot convert free partially methylated lysines into trimethyl-lysine.