Eur. J. Immunol. 1987.17: 849-854 Life spans of “natural”Ig-secreting cells 849 zy Martine Levy’, Paul0 Vieira+, Antiinio Coutinho’ and Anthio Freitas’ Laboratory of Immunobiology’, Department of Immunology, Pasteur Institute, Paris and Department of Immunology, Institute for Genetics’, University of Cologne, Cologne The majority of “natural” immunoglobulin-secreting cells are short-lived and the progeny of cycling lymphocytes Treatment of mice with hydroxyurea to selectively kill all cycling cells has been used to study population dynamics and life expectancy of “natural” immunoglobulin-se- creting cells in the bone marrow and spleen of nonimmunized animals. The results show that 50 to 90% of those cells are eliminated 2 to 3 days after one cycle of hydroxyurea administration, demonstrating their recent origin from cycling precur- sors. Using a protocol of long-term hydroxyurea treatment which abrogates cell pro- duction from the cycling precursors compartments, it was shown that “natural” immunoglobulin-secreting cells have a very short half renewal time, in the range of 15-60 h. 1 Introduction Studies on the internal activity of the immune system should consider the physiology of “naturally” activated lymphocytes. Normal immune systems maintain constant levels of circulat- ing serum immunoglobulins (Ig) displaying a broad spectrum of reactivities. Given the half-lives of serum Ig molecules, esti- mated to 3-10 days for IgG [l-31 and 12 h for IgM [2, 31, these constant levels must be ensured through the continuous production of Ig by Ig-secreting cells in the various lymphoid organs. The population dynamics of “natural” plasma cells is therefore susceptible to give indications on the degree of natural antibody repertoire organization and variation. Studies on the renewal rate and life expectancy of this set of cells could provide useful information on the physiology of immune systems. In terms of cell population kinetics, Ig-secreting cells belong to the terminal effector compartment of the B cell lineage zyxwvu [4]. Arrest of cell input into a terminal compartment can be achieved by selectively killing their precursors. If these precur- sors are cycling cells, they can be eliminated by treatment with antimitotic agents. The antimitotic drug hydroxyurea (HU) has been chosen for these studies as it eliminates all cycling cells zyxwvutsrq in zyxwvutsrqponml vivo zyxwvutsr IS, 61 and is deprived of any toxic effects towards noncycling cells [7, 81. The rates of decay of Ig-secreting cells in lymphoid organs after HU treatment were then used as a parameter to define life expectancy of this cell population. The results show that most natural Ig-secreting cells have a very short life expectancy and represent the progeny of cycling cells. 2 Materials and methods 2.1 Drug treatment Eight to 12-week-old mice of the C57BW6, BALB/c and C3H/ Tif strains bred at the Pasteur Institute, Paris, France, were [I 60711 Correspondence: Martine Ltvy, Unitt d’hmunobiologie, Institut Pasteur, 28, rue du Docteur Row, F-75724 Pans Ctdex 15, France Abbreviations: BM Bone marrow HRT: Half renewal time(s) zyxwvutsrqpon W: Hydroxyurea Ig(s): Immunoglobulin(s) LN: Lymph node(s) LPS: Lipopolysaccharide PFC: Plaque-forming cells used. Hydroxyurea (HU; Sigma, St. Louis, MO) was dissolved in 0.9% NaCl and two doses of 1 mg/g of body weight were injected i.p. 7 h apart, according to Hodgson et al. [9]. Two injections with a 7-h interval are called one cycle of HU administration. In the experiments reported here, mice received 1, 2, 3 or 4 cycles of HU, according to a protocol previously described [lo]. 2.2 Cell suspensions Spleen, bone marrow (BM) and lymph node (LN) suspensions were prepared as described [ll]. In all experiments reported here, mesenteric and inguinal LN have been taken. Where stated, the number of B cells was estimated using a rhodamine-labeled rabbit anti-mouse antibody as described PI zyxw * 2.3 Plaque assay Determinations of Ig-secreting cells were performed by a mod- ified hemolytic plaque assay using protein A-coated sheep red blood cells, rabbit antisera specific for the different mouse isotypes and guinea pig complement (Gibco, Grand Island, NY) [12]. 2.4 Serum Ig estimation Ig concentrations in the serum were determined in a solid- phase radioimmunoassay as described 1131. Briefly, plates were coated with goat antibodies directed against mouse zy x, p and y1 chains and the bound Ig were determined with radioiodinated goat anti-mouse corresponding antibodies. 2.5 In vivo polyclonal activation C57BL/6 mice were treated with a single i.p. injection of 5 wg of lipopolysaccharide (LPS from Salmonella abortus zy equi, Difco, Detroit, MI). 3 Results 3.1 Effects of antimitotictreatment on the natural Ig-secreting cell pool Adult mice were killed at various times after they were given two consecutive injections of HU, and the number of natural 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1987 0014-2980/87/0606-0849$02.50/0