Converging Discoveries: The First Reports of IL-4

Interleukin-4 was one of the first cytokines to be intensively studied and remains one of the most interesting and biologically important (1). It has a distinctive spectrum of biological activities, many of which are unique to IL-4 or to both IL-4 and IL-13 but are not shared by other cytokine families. The ability to produce IL-4 is the defining characteristic of an entire subset of T cell immunity, the Th2 response, and, at the same time, it is the key signal to induce Th2 differentiation from naive T cells (2). As a consequence, IL-4 remains particularly relevant to studies in allergy and asthma (3), as well as helminth infection (4). Thirty years ago, two papers published in a single issue of The Journal of Experimental Medicine described quite distinct biological activities mediated by the cytokine we now call IL-4 (5, 6). However, they were not coordinated submissions by two groups aware they were working in the same area. On the contrary, several years would elapse before it became clear that these papers represented the first descriptions of the same cytokine.

In one of these papers, Maureen Howard, Bill Paul, and their collaborators (5) described an activity in supernatants of a T cell lymphoma that enhanced the proliferative response of mouse B cells activated with anti-IgM Abs. This activity was not absolutely required for proliferation, but provided substantial enhancement of tritiated thymidine incorporation under conditions of suboptimal cell density and anti-IgM concentration. This activity was mediated by a secreted protein of ∼18,000 kDa and was distinct by multiple criteria from IL-2, the only well-characterized cytokine known then to have B cell stimulating activity. This newly identified cytokine was initially named B cell growth factor, reflecting the activity on which its identification was based, but was renamed B cell stimulatory factor-1 (BSF-1) upon the recommendation of an international nomenclature committee (7).

The second paper, from Ellen Vitetta, Peter Krammer, and their colleagues (6), asked whether T cells could produce soluble factors that could stimulate the switching of IgM, IgD-positive B cells to IgG-producing cells. They established a simple assay based on adding supernatants from activated T cell clones or hybridomas to mouse B cells cultured with LPS. LPS itself induces some IgG production, but supernatants from two alloreactive T cell clones and one T cell hybridoma led to a substantial increase in IgG production and in the number of IgG plaque-forming cells. The more striking finding, however, was that these supernatants profoundly altered the subclasses of IgG produced. Levels of IgG2b and IgG3, the principal IgG subclasses induced by LPS, were not altered, but the active supernatants induced high levels of IgG1 production, accounting for most of the IgG produced. To show that this reflected a selective induction of isotype switching, the experiment was repeated in cultures of B cells rigorously depleted of IgG-bearing cells. The results were comparable to those with unfractionated B cells, demonstrating that most of the IgG1 induced by LPS plus the active supernatants came from B cells that had not undergone isotype switching prior to the culture. Based on this activity, the active factor was christened B cell differentiation factor-γ (BCDF-γ). Isakson et al. (6) did not report biochemical characterization of BCDF-γ in this paper, although its presence in supernatants of various T cell lines did not correlate with expression of several other cytokines or activities, including IL-2.

Collaboration between these two teams in 1985 showed definitively that the BSF-1 and BCDF-γ activities were mediated by the same protein, using a mAb produced against highly purified BSF-1 (8). This finding was a significant milestone in cytokine research for several reasons. A widely held assumption in the early days of cytokine discovery was that proliferation and differentiation of immune cell types were likely mediated by different factors, reflecting the long-standing idea that growth and differentiation were distinct, sequential processes in the development of most hematopoietic lineages. Earlier studies of BSF-1 had already begun to raise doubts about this distinction by showing that BSF-1 acted on resting B cells to induce MHC class II expression (9, 10) and to prepare cells for entry into cell cycle upon receiving a subsequent mitogenic signal (11, 12). The demonstration that BSF-1/BCDF-γ was a potent mediator both of proliferation and isotype switching, a key differentiation step, in the same cell lineage clearly showed this notion to be incorrect. This re-evaluation removed a significant impediment to understanding the biological functions of known cytokines and the discovery of new ones. The subsequent demonstration that BSF-1 was a growth factor for both T cells and mast cells (13) expanded the range of biological activities of this single cytokine and further eroded the traditional idea of distinct growth and differentiation factors specific for each major hematopoietic lineage.

Second, it was one of these experimental systems that provided the bioassay needed to clone the gene for BSF-1/BCDF-γ, a necessary step for making the cytokine routinely available for research. In the 1980s, a widely used strategy for cloning genes encoding biologically active proteins was “expression cloning,” that is, screening for biological activity in supernatants from cell lines transfected with increasingly smaller pools of cDNA clones constructed with mammalian expression vectors. Interestingly, it was the IgG1 switching assay, not the seemingly simpler B cell proliferation assay, that was employed by the groups of Tasaku Honjo and Eva Severinson to clone mouse BSF-1/BCDF-γ, which was renamed once again as IL-4 (14). Soon thereafter, a group from the DNAX Research Institute independently cloned IL-4 using multiple biological assays in parallel, including the analogous assay for switching to IgE (15).

The demonstration of a soluble factor that could promote switching to one specific H chain isotype was, in itself, a landmark in understanding T cell regulation of B cell function. By 1982, it was accepted that isotype switching was primarily, although not exclusively, regulated by T cells. There was considerable speculation, but little hard data, that specific T cell subsets controlled switching to each isotype, possibly through secretion of isotype-specific switch factors. The competing idea was that switch recombination was a stochastic process, not specified by T cells, and that appropriate isotypes were subsequently selected, possibly with influence from T cells. The demonstration by Isakson et al. (6) of the existence of a T cell–produced factor capable of inducing a switch to one specific isotype profoundly changed the debate to favor a switching mechanism directed by T cells. However, it was not until the definition of the Th1 and Th2 subsets (16) with their distinctive patterns of cytokine-enhanced isotype switching (17, 18) that it became clear that there was not a specialized T cell subset specific for each of the IgH isotypes, but rather a distinct pattern of switch recombination induced by each Th subset and regulated by cytokines that induced or inhibited specific isotypes.