CD23, the low-affinity IgE receptor, is up-regulated on interleukin (IL)-4-stimulated B cells and monocytes, with a concomitant increase in the release of soluble fragments of CD23 (sCD23) into the medium by proteolytic processing of the surface-bound intact CD23. The effect of inhibition of the processing of CD23 on IgE production in human and mouse cells and in a mouse model in vivo was evaluated. CD23 processing to sCD23 from RPMI 8866 (a human Epstein-Barr virus-transformed B cell line) cell membranes was inhibited by a broad-spectrum matrix-metalloprotease inhibitor, batimastat, with an IC50 of 0.15 microM. Batimastat also inhibited CD23 processing in whole RPMI 8866 cells as well as in IL-4-stimulated purified human monocytes with similar IC50. Batimastat inhibited IgE production from IL-4/anti-CD40-stimulated human tonsil B cells as well as mouse splenic B cells in a manner consistent with inhibition of CD23 processing. Release of soluble fragments of CD23 in the cell supernatants of tonsil B cells was inhibited over the concentration range of 1-10 microM batimastat and intact cell surface CD23 was increased on mouse splenic B cells in the presence of these concentrations of batimastat. IgE production of IL-4-stimulated human peripheral blood mononuclear cells was also blocked by 1-10 microM batimastat, again with comparable inhibition of sCD23 release over the same concentration range. Finally, in a mouse model of IgE production, batimastat inhibited IgE production in response to ovalbumin challenge as determined by serum IgE levels. Taken together, the data support a role of CD23 in IgE production and point to CD23 processing to sCD23 as a therapeutically relevant control point in the regulation of IgE synthesis.

No abstract

No abstract

Regression of the L-asparaginase-sensitive Gardner lymphosarcoma was investigated in normal CBA mice and in CBA mice deficient in thymus-processed lymphocytes ( T cells). I n normal animals, solid subcutaneous grafts of the tumour disappeared completely following one intraperitoneal injection of L-asparaginase (50 IU) given 10 days after tumour inoculation. The local tumour was impalpable after 3 or 4 days and there was early and massive destruction of the graft though surviving tumour cells persisted in the vicinity of blood vessels for up to 24 h after treatment. These cells subsequently disappeared and there was an intense local inflammatory response dominated by macrophages. In mice deficient in T cells, the Gardner lymphosarcoma initially

CD23 is a multifunctional molecule expressed by cells of lymphoid, myeloid and hematopoietic lineages. As a cell surface molecule CD23 acts both as a low-affinity receptor for IgE (Fc epsilon RII) and as a cell adhesion molecule. CD23 can undergo autoproteolysis to release soluble 37-25-kDa CD23 (s-CD23) molecules with a range of cytokine activities. Here we show a causal link between the two apparently disparate functions of autoproteolysis and cell adhesion. The Epstein-Barr virus-transformed B cell line RPMI-8866 formed macroscopic cell clusters solely via CD23. Cell adhesion was inhibited by mAb to CD23 and by IgE. Cell adhesion was also dependent on serum as cells grown in serum-free media failed to form clusters. In serum-free conditions cell adhesion could be induced by the addition of not only 10% FCS but also s-CD23. As s-CD23 is reported to possess proteolytic activity we screened a range of proteases to determine whether they also could induce cell adhesion in serum-free medium. It was found that chymotrypsin and elastase induced cell:cell adhesion in RPMI-8866 cells. The same panel of proteases were screened against a range of CD23-positive (Jijoye, AF-10, T2, U937, ICH-1) and CD23-negative (RPMI-8226, U266, MOLT-4, Ramos) cell lines. It was found that chymotrypsin and elastase induce cell adhesion only in cells expressing CD23. Peptide mapping studies showed that chymotrypsin and elastase cleaved immunoprecipitated CD23 near the same site by which 37-kDa s-CD23 is released (Ala 80). Serum demonstrated no proteolytic activity towards CD23. However, it was found that cells grown in serum-free medium released 25-kDa s-CD23 without the need for prior cleavage at the 37-kDa cleavage site. To confirm the role of proteolysis in CD23-mediated cell adhesion we screened a range of protease inhibitors for their ability to antagonize this process. It was found that tosyl-lysine chloromethyl ketone inhibited CD23-mediated cell adhesion. Lactoperoxidase treatment, which inhibits CD23 cleavage, also inhibited cell adhesion. Addition of chymotrypsin and elastase to lactoperoxidase-treated cells induced cell adhesion. From these data we propose that intact CD23 has no demonstrable role in cell adhesion; instead, the portion of CD23 remaining on the cell surface following cleavage appears to mediate cell adhesion.