The T-lymphocyte population is divisible into several subclasses; each subclass possesses a distinctive genetic program which combines information for cell-surface phenotype and function (1). In the mouse, there is evidence that T cells which express the Thyl+Lyl+Ly23 -surface phenotype CLyl cells") are programmed for helper (TH) 1 function. In contrast, T cells that express the Thyl+Lyl-Ly23 ÷ surface phenotype CLy23 cells") are programmed for suppressor (Ts) function (1). Isolation of these two T-cell subclasses in mice depleted of T cells CB mice") has indicated that each belongs to an independent line or branch of thymus-dependent differentiation (2). A third major T-cell subclass, expressing the surface phenotype Lyl+2+3 +, can react to antigen and differentiate to Ly23 + cytotoxic effector cells (3), suggesting that this subclass probably contains precursor cells that have acquired receptors for antigen but have not yet become committed to either TH or Tc~s function (3).These findings, and others, are consistent with the view that functionally distinct T-cell sets carry cell surface components that are invariably associated with particular immunologic function. According to this idea, cells carrying the Lyl+Ly23 -surface phenotype are programmed for helper and not suppressive activity regardless of external conditions, such as the mode or type of antigen stimulation. To test this hypothesis we have stimulated purified populations of Lyl+2 -T cells with antigen in vitro, by using conditions devised to induce unselected T cells to express optimal levels of antigen-specific suppressive activity (4). We find that (a) stimulation of purified Lyl cells under these conditions results in the generation of TH but not Ts activity and (b) such hyperimmune Lyl cells also induce a subset of nonimmune T cells to exert potent suppressive effects upon the antibody response. The surface phenotype of the T-cell set responsible for "feedback" inhibition is described in this study.

We have shown that (a) purified T-helper cells induce cells of another T-cell set-, expressing the Ly123+Qa1+ surface phenotype, to exert potent suppressive activity, (b) this T-T interaction plays an important role in regulating in vivo immune responses, and (c) this interaction represents an important barrier to protocols intended to augment the immune status of individuals by adoptive (or active) immunotherapy. Our results also indicate that the Ly123+ T-cell set mediating feedback suppression in vivo is sensitive to both low doses of cyclophosphamide and removal of the thymus in adult life. The importance of this T-T interaction to normal, physiologic regulation of the immune system is emphasized by the finding that the major T-cell deficit of NZB mice (an inbred strain of mice that spontaneously develops an autoimmune disorder) is the absence or malfunction of an Ly123+ T-cell set responsible for feedback inhibition.

Antigen-stimulated Lyl cells induce B cells to secrete antibody and induce a nonimmune set of T cells (surface phenotype Ly123+Qal +) to participate in specific suppressor activity (1, 2). We have referred to this suppressive T-T interaction as feedback inhibition because (a) the level of suppression exerted by a fixed number of nonimmune T cells increase in direct proportion to the numbers of antigen-stimulated Lyl cells (0.5-5 × 105) in cell culture and (b) one consequence of Ly123-assoeiated suppression is a reduction in the delivery of T-helper activity to B cells.These observations indicate that cells of the Ly 1 set play a central role in regulating antibody formation. It is therefore important to determine whether all Ly 1 cells are programmed to induce nonimmune T cells to exert feedback inhibition or whether this immunologic function is invested in a specialized set of Lyl cells. If the latter explanation were the case, then the intensity and duration of the immune response after stimulation by antigen would depend in part upon the degree of activation of Lyl cells that induce feedback inhibition.A direct approach to this question comes from the finding that a portion of Lyl + cells also express Qal surface components. 1 We find that (a) Lyl:Qal + cells are responsible for induction of feedback inhibition and (b) signals from both Ly 1 :Qal + and Lyl :Qal-ceils are required for optimal formation of antibody by B cells. Materials and MethodsMice. C57BL/6 (B6) mice 10-to 14-wk of age were obtained from the Jackson Laboratory, Bar Harbor, Maine. The congenic lines B6-Lyl a and B6-Lyla3 a, phenotypes Lyl. 1,2.2,3.2 and Lyl.2,2.1,3.1, respectively, and B6-Tla a were produced and supplied by Dr. E. A. Boyse, Memorial Sloan-Kettering Cancer Center, New York.Production and Use of Antisera. Congenic anti-Thyl.2, anti-Lyl.2, anti-Ly2.2, and anti-Ly3.2 were prepared as described previously (3). The antiserum (B6 × A-Tla ~) anti-A strain leukemia ASL1, heretofore termed anti-TL in reference to its reaction with thymocytes, and herein termed anti-Qal in reference to its reaction with peripheral T cells, is described elsewhere (4).

No abstract