SummaryCostimulation mediated by the CD28 molecule plays an important role in optimal activation of T cells. However, CD28-deficient mice can mount effective T cell-dependent immune responses, suggesting the existence of other costimulatory systems. In a search for other costimulatory molecules on T cells, we have developed a monoclonal antibody (mAb) that can costimulate T cells in the absence of antigen-presenting cells (APC). The molecule recognized by this mAb, 9D3, was found to be expressed on almost all mature T cells and to be a protein of ,'-~24 kD molecular mass. By expression cloning, this molecule was identified as CD9. 9D3 (anti-CD9) synergized with suboptimal doses of anti-CD3 mAb in inducing proliferation by virgin T cells. Costimulation was induced by independent ligation of CD3 and CD9, suggesting that colocalization of these two molecules is not required for T cell activation. The costimulation by anti-CD9 was as potent as that by anti-CD28. Moreover, anti-CD9 costimulated in a CD28-independent way because anti-CD9 equally costimulated T cells from the CD28-deficient as well as wild-type mice. Thus, these results indicate that CD9 serves as a molecule on T cells that can deliver a potent CD28-independent costimulatory signal. Full activation of the T cell has been shown to require two independent signals (1). The first signal is provided by antigen-specific T cell receptor (TCR) interacting with processed antigen peptides plus major histocompatibitity complex (MHC) molecules on APC. This signal leads to an effective T cell response only when accompanied by a second costimulatory signal(s) presented by the APC. The lack of costimulation not only prevents activation but also induces tolerance called anergy (1). Identifying molecules capable of delivering costimulatory signals has been the subject of a large number of recent investigations (2-5). CD28 expressed on T cells was found to be a receptor for the costimulatory molecules CD80 and CD86 on APC (6). CD28 engagement, by either anti-CD28 mAb or ligands (CD80/ CD86), has been shown to costimulate T cells in the absence of APC, resulting in T cell activation (7-9). Conversely, the blocking of CD28-1igand interactions induced substantial inhibition of T cell activation (2). These observations indicated that the CD28-CD80/CD86 interaction functions as a critical pathway of T cell costimulation. Nevertheless, recent studies have revealed that CD28-deficient mice can develop normal in vivo immune responses (10) and that T cells from these mice mount APC-dependent responses for T cell activa~:ion in vitro although the response is reduced compared to T cells from wild-type mice (10, 11). Thus, these results strongly suggest that there may exist other molecules capable of providing costimulatory activity.In this report, we have developed a rat IgG mAb (9D3) by immunization with cells of a murine thymic stromal clone (12). This mAb recognized a protein of"-,24 kD that is expressed on immunizing thymic stromal cells as well as murine T cells. By cDNA expr...

T cell activation requires two signals: a signal from the TCR and a co-stimulatory signal provided by antigen-presenting cells (APC). In addition to CD28, multiple molecules on the T cell have been described to deliver co-stimulatory signals. Here, we investigated whether there exist quantitative or qualitative differences in the co-stimulatory capacity between CD28 and other molecules. Anti-CD28 monoclonal antibody (mAb) and mAb against CD5, CD9, CD2, CD44 or CD11a all induced activation of naive T cells in the absence of APC when co-immobilized with a submitogenic dose of anti-CD3 mAb. [ 3 H]Thymidine incorporation determined 2 days after co-stimulation was all comparable. In contrast to progressive T cell proliferation induced by CD28 co-stimulation, co-stimulation by other T cell molecules led to a decrease in viable cell recovery along with the induction of apoptosis of once activated T cells. This was associated with a striking difference in IL-2 production; CD28 co-stimulation induced progressively increasing IL-2 production, whereas co-stimulation by other molecules produced limited amounts of IL-2. Addition of recombinant IL-2 to the latter cultures corrected the induction of apoptosis, resulting in levels of cellular proliferation comparable to those observed for CD28 co-stimulation. These results indicate that a fundamental difference exists in the nature of co-stimulation between CD28 and other molecules, which can be evaluated by the levels of IL-2 production, but not simply by [ 3 H]thymidine incorporation.

T cell activation requires two signals: a signal from the TCR and a co-stimulatory signal provided by antigen-presenting cells (APC). In addition to CD28, multiple molecules on the T cell have been described to deliver co-stimulatory signals. Here, we investigated whether there exist quantitative or qualitative differences in the co-stimulatory capacity between CD28 and other molecules. Anti-CD28 monoclonal antibody (mAb) and mAb against CD5, CD9, CD2, CD44 or CD11a all induced activation of naive T cells in the absence of APC when co-immobilized with a submitogenic dose of anti-CD3 mAb. [3H]Thymidine incorporation determined 2 days after co-stimulation was all comparable. In contrast to progressive T cell proliferation induced by CD28 co-stimulation, co-stimulation by other T cell molecules led to a decrease in viable cell recovery along with the induction of apoptosis of once activated T cells. This was associated with a striking difference in IL-2 production; CD28 co-stimulation induced progressively increasing IL-2 production, whereas co-stimulation by other molecules produced limited amounts of IL-2. Addition of recombinant IL-2 to the latter cultures corrected the induction of apoptosis, resulting in levels of cellular proliferation comparable to those observed for CD28 co-stimulation. These results indicate that a fundamental difference exists in the nature of co-stimulation between CD28 and other molecules, which can be evaluated by the levels of IL-2 production, but not simply by [3H]thymidine incorporation.

By the use of ferritin-conjugated antibody (conjugate) indirect immunoelectron microscopy, NADPH-cytochrome c reductase was localized on rat liver microsomes. Most microsomes in the sections had from 1 to 12 conjugates on their outer surfaces. Among the conjugates, 83% was estimated to bind to NADPHcytochrome c reductase at a molecular ratio of 1:1, 12% at the ratio of 2:1, and 5% at the ratio of 3 or 4:1. The correlation between immunochemical and morphological data confirmed that most of the NADPH-cytochrome c reductase reacted with the conjugates. Subsequent morphological analyses have revealed that the enzyme is distributed homogeneously on the outer surfaces of microsomes but heterogeneously within microsomes in groups of three to five enzyme molecules.

No abstract