The realization during the past few years that CD4+ T cells (TH cells) can be subdivided into distinct subsets on the basis of both function and cytokine expression (Mosmann & Coffman 1987, 1989) has helped clarify a number of outstanding questions in immunology. One such question, for example, is whether B-ceil help and delayed-type hypersensitivity (DTH) are mediated by the same type of T cell. Not surprisingly, this revised view of TH cells has raised a new series of questions aboul T cell-mediated regulatory and effector functions. Among the most interesting, and perhaps complex, of the questions are: From what precursors do THI and TH2 cells arise? and: What factors regulate this differentiation process?

Most evidence now favors the view that THI and TH2 cells develop from a common postthymic precursor, probably after T cell receptor-mediated activation. Experiments in which primary murine TH clones were derived by limiting dilution suggest that many TH cells differentiate first into cells that express both THland TH2-specific cytokines when restimulated (THO cells) and that THO clones often differentiate into either TH 1 or TH2 clones after additional cycles of antigen stimulation (Firestein et al. 1989, Street et al. 1990). Swain and colleagues have identified, at the population level, a precursor (presumably a virgin TH cell) that makes initially only IL-2, but then differentiates into cells making the THO cytokine pattern (Swain et al. 1991). Support for this differentiation pathway in humans comes from observations that human TH clones specific for antigens to which the donor is nol immunized are similar to murine THO clones (based, especially, on their coexprcssion of IL-4 and lI'N-j'), whereas most clones for which the donor was hyperimmune resemble murine TH! or TH2 clones (Bacchet-