Purification to homogeneity of a human hematopoietic growth factor that stimulates the growth of a murine interleukin 3-dependent cell line

A human lymphokine derived from the 5637 bladder carcinoma has been purified to homogeneity by using sequential reverse-phase high pressure liquid chromatography. A high recovery of biological activity is obtained by using this purification. The NH2-terminal amino acid sequence shows no homology to human interleukin 1 (IL 1), human IL 2, murine IL 3, or human granulocyte-macrophage colony-stimulating factor. The growth-promoting properties of the 5637-derived factor can be rapidly assayed by using the murine IL 3-dependent 32D c1-23 cell line. The amino acid sequence described is identical to that recently described for a human granulocyte colony-stimulating factor.     

B cell stimulatory factor-2 is involved in the differentiation of cytotoxic T lymphocytes

The induction of cytotoxic T lymphocytes (CTL) from precursor T cells requires both antigen and lymphokine signals. Previous work from our laboratory has indicated that three lymphokines are required for the induction of CTL from murine thymocytes; interleukin 2, interferon-gamma (IFN-gamma), and a partially characterized factor referred to as cytotoxic differentiation factor (CDF). While attempting to clone CDF from the human T cell line C10-MJ2, we found that a gene encoding CDF-like activity is identical to the gene encoding the factor known variously as B cell stimulatory factor-2 (BSF-2), IFN-beta 2, and 26-kDa protein. We report here that BSF-2 can induce the differentiation of Ly-2+ CTL from murine thymocytes in the presence of interleukin 2 and that the level of cytotoxicity is augmented by the addition of murine IFN-gamma. Serine esterase, a marker for cytotoxic granules in CTL, was induced only in the presence of BSF-2, and the level of serine esterase activity correlated with the level of serine esterase activity correlated with the level of cytotoxicity. These data suggest that BSF-2 is a differentiation factor for CTL and that it functions in part by inducing proteins required for mediating target cell lysis.     

Murine T-cell clones specific for chicken erythrocyte alloantigens

We have established murine T-cell clones which respond to allotypic and species-specific determinants found on chicken erythrocytes (cRBC). Their relative antigen specificities were determined by assessing lymphokine production and proliferation in response to syngeneic spleen cells and cRBC obtained from chickens homozygous for major histocompatibility complex (MHC) antigens. The specificity pattern suggested that the T-cell clones recognized a more restricted set of cRBC MHC-associated allodeterminants than do antibody-producing cells. The antigen-specific responses required antigen processing, and were MHC restricted and antigen dose dependent. Approximately 20% of T-cell clones from appropriate strains of mice were also Mls alloreactive. This second reactivity showed no correlation with nominal cRBC specificity. The induction-specific lymphokine activities of T-cell growth factor, mast cell growth factor, and Ia induction factor were identified as interleukin 2 (IL-2), interleukin 3 (IL-3), and interferon-gamma respectively.     

B cell growth modulating and differentiating activity of recombinant human 26-kd protein (BSF-2, HuIFN-beta 2, HPGF).

The human "26-kd protein' is a secreted glycoprotein expressed, for example, in (blood) leukocytes, in epithelial cells treated with various inducers, but most strongly in interleukin-1 (IL-1)-treated fibroblasts. After finding it has antiviral and 2-5A synthetase-inducing activity, one group of authors called this protein IFN-beta 2. However, recently the full-length 26-kd cDNA sequence was shown to be identical with that of a B-cell-differentiating lymphokine called BSF-2, and another report suggested that the 26-kd protein could support the growth of some transformed murine B cell lines. To define its biological activities, we expressed the recombinant 26-kd protein by translating in Xenopus laevis oocytes a pure, synthetic chimeric mRNA containing the 26-kd protein coding region surrounded by Xenopus laevis beta-globin untranslated regions. A similar construction, but containing the HuIFN-beta cDNA coding region, was used to produce HuIFN-beta by the same procedure. Both recombinant glycoproteins were secreted, glycosylated, and their amounts were measured by [35S]methionine incorporation by the oocyte. Here we show that the recombinant 26-kd protein exhibits a high growth factor activity when assayed on an IL-HP1-dependent murine B cell hybridoma (sp. act. approximately 2 X 10(8) U/mg) as well as a potent differentiating activity on human CESS cells (sp. act. approximately 5 X 10(7) U/mg). While rHuIFN-beta was inactive in the latter two assays, it had the expected antiviral activity of 1-5 X 10(8) U/mg. The parallel recombinant 26-kd protein preparations had no detectable antiviral activity (i.e. a maximal specific activity of 1-3 X 10(2) U/mg, if any). The 26-kd protein is thus clearly an interleukin, and considering the confusing nomenclature now in use, this factor may better be renamed "interleukin 6'.     

A short synthetic peptide fragment of human interleukin 1 with immunostimulatory but not inflammatory activity

Short peptide fragments of human and murine interleukin 1 (IL 1) were synthesized on the basis of their predicted exposure on the surface of the molecule in an attempt to identify the minimal structure responsible for the immunostimulatory activity of IL 1. One of these peptides, a fragment of nine residues of human IL 1 beta (VQGEESNDK, fragment 163-171), showed high T cell activation capacity, as judged by its ability to stimulate murine thymocyte proliferation and to potently induce interleukin 2 production in spleen cells. On the other hand, the 163-171 peptide was devoid of prostaglandin-inducing capacity in vitro and pyrogenic activity in vivo, two inflammatory features peculiar to the entire hu IL 1 beta molecule. Thus we propose that this peptide may represent one of the portions of hu IL 1 beta responsible for its immunostimulatory capacity.     

Epidermal cell (keratinocyte)-derived thymocyte-activating factor (ETAF)

In order to determine whether keratinocytes play a role in the modulation of the immune response, we investigated the murine keratinocyte cell line Pam 212. In culture these cells generate a substance with a biologic activity that greatly enhances phytohemagglutinin-induced thymocyte proliferation. We have, therefore, called this substance epidermal cell thymocyte-activating factor (ETAF). This keratinocyte-derived supernatant activity is mainly produced at the onset of the logarithmic growth phase and is directly mitogenic for murine thymocytes. Although ETAF by itself exhibits no T cell growth factor activity, ETAF enhances Interleukin 2 production by mitogen-stimulated murine spleen cells. Murine ETAF is not genetically restricted and lacks species specificity since it decreases lectin-induced proliferation of human peripheral blood lymphocytes (as well as murine spleen cells) and also enhances the production of human Interleukin 2. The factor has a m.w. between 15,000 and 25,000 as determined by gel filtration and elutes as a single peak from anion exchange chromatography columns. The activity is maintained mainly at alkaline pH and is rapidly destroyed at temperatures above 60 degrees C. These observations suggest that epidermal cells may interact with the immune system by elaborating nonspecific factors that modulate lymphocyte proliferation and augment lymphokine production.     

Biochemical characterization of lymphocyte regulatory molecules. II. Purification of a class of rat and human lymphokines

Rat spleen cells activated in vitro by concanavalin A produce lymphokine molecules that possess biologic activity in a number of murine lymphocyte response assays. A single class of lymphokine most adequately described as T cell growth factor (TCGF, Interleukin-2) with a m.w. of 15,000 as estimated from gel filtration studies and with an isoelectric range of 5.4 to 5.6 stimulates i) the growth of established T cell lines in culture, ii) the proliferation of thymocytes in the presence of Con A under culture conditions where Con A alone is non-mitogenic, iii) the induction of antibody responses to heterologous erythrocyte antigens in athymic (nude) mouse spleen cell cultures, and iv) the generation of cytotoxic T lymphocytes (CTL) in thymocyte cultures and in nude mouse spleen cell cultures. We suggest that in each of the assay systems tested, this class of rat lymphokine acts directly on activated T cells. Nonactivated T cells must be stimulated by either mitogen or antigen before becoming responsive to lymphokine, but do not require antigen or mitogen for continued lymphokine-dependent proliferation. Similarly, human peripheral blood mononuclear cells activated by phytohemagglutinin (PHA) produce a class of lymphokines of identical size with an isoelectric point of 6.0 to 6.5 that possess the same biologic properties as measured in murine lymphocyte response systems.     

The four biochemically distinct species of human interleukin 1 all exhibit similar biologic activities

The supernatants of human monocytes incubated with endotoxin are able to stimulate the proliferation of murine thymocytes in the presence of PHA. This is known as LAF (lymphocyte activating factor) activity and is a characteristic activity of interleukin 1 (IL 1). The LAF activity can be resolved into four major fractions: a 15,000 dalton (pI 7), a 15,000 (pI 5.5), a 35,000 (pI 7), and a 35,000 (pI 5.5) fraction. To determine whether these four fractions shared the other biologic activities ascribed to IL 1, they were compared in a series of bioassays. When standardized with respect to their LAF activities, the four fractions did not differ significantly as mitogens for murine thymocytes, inducers of IL 2, murine or human B cell activators, human chondrocyte or synoviocyte stimulants, or inducers of acute phase proteins in vivo. On the other hand, the samples differed markedly as stimulators of porcine synoviocytes, with the 15,000 dalton (pI 5.5) fraction being the only strongly active fraction. These results are consistent with the hypothesis that all four LAF could be products of a single gene, although the porcine receptor may be able to distinguish between them. If this is the case, all four fractions can properly be termed IL 1.     

Cytolytic T lymphocyte clones that proliferate autonomously to specific alloantigenic stimulation. II. Relationship of the Lyt-2 molecular complex to cytolytic activity, proliferation, and lymphokine secretion

Previous analyses of the inhibitory effects of anti-Lyt-2 monoclonal antibodies (mAb) on cytolytic activity suggested that Lyt-2/3 antigens expressed on the surface of murine cytolytic T lymphocytes (CTL) are involved in antigen recognition. In the present study, we investigated the effects of anti-Lyt-2 mAb (in the absence of complement) on the functional activities of H-2K/D-specific Lyt-2+ CTL clones that proliferate to antigenic stimulation in the absence of helper T cells or added interleukin 2 (IL 2) and secrete lymphokines. For those clones that were inhibited in cytolysis by anti-Lyt-2 mAb, a parallel inhibition of antigen-dependent proliferation and lymphokine secretion (interferon, macrophage-activating factor) was observed. Inhibition of proliferation or lymphokine secretion could be overcome by the addition of IL 2 or lectin, respectively. Collectively, these results would strongly suggest that anti-Lyt-2 mAb were inhibiting CTL antigen recognition. Not all CTL clones, however, were inhibited in cytolysis by anti-Lyt-2 mAb, in which case proliferation and lymphokine secretion were similarly unaffected. This heterogeneity of Lyt-2+ CTL clones in their susceptibility to inhibition of cytolytic activity, proliferation, and lymphokine secretion by anti-Lyt-2 mAb is discussed in the context of a model proposing that Lyt-2/3 molecules function to stabilize the interaction between CTL receptors and the corresponding target/stimulating cell antigens. Such a stabilization may be required by CTL possessing few and/or low affinity receptors.     

A mouse T cell product that preferentially enhances IgA production. II. Physicochemical characterization

Certain subsets of helper T cells, following stimulation with concanavalin A, secrete factors that specifically enhance the production of IgG1, IgE, and IgA by lipopolysaccharide-stimulated B cells. In the previous report, we describe a factor from the helper T cell line MB2-1 which enhances IgA production. IgA-enhancing factor has been purified from serum-free supernatants of this cell line. The purified lymphokine is a family of microheterogeneous polypeptides presumably modified post-translationally. IgA-enhancing factor has a native m.w. of 45,000 to 60,000 with subunits of between 24,000 and 28,000 under reducing conditions. Upon Edman degradation, a single amino-terminal sequence is detected which is identical to that of the lymphokine interleukin 5. IgA-enhancing factor activity is thus mediated by the same polypeptide that has been characterized as type II B cell growth factor, T cell-replacing factor, and eosinophil-differentiation factor.     

Human IL-3 (multi-CSF): identification by expression cloning of a novel hematopoietic growth factor related to murine IL-3

A cDNA clone encoding a novel hematopoietic growth factor activity produced by a gibbon T cell line has been identified using a mammalian cell expression cloning system. The sequence of this cDNA proved to have significant homology to the sequence encoding murine interleukin 3 (IL-3). The human gene, which was readily identified because of its high degree of homology to the gibbon sequence, also displayed significant homology with the murine IL-3 sequence. The recombinant gibbon IL-3 protein proved to have multipotent colony stimulating activity when tested with normal human bone marrow cells, proving that this primate hematopoietin is not only structurally but also functionally related to murine IL-3.     

Regulation of class II gene expression in a murine pre-B cell line

Class II major histocompatibility complex (MHC) gene expression has been studied in an Abelson virus-transformed pre-B cell line R8, and its Ia-negative variant R8205. These variant cells contained barely detectable levels of RNA specific for all class II genes, including the nonpolymorphic invariant chain gene (Ii), and did not express cell surface Ia. Fusion of this murine Ia-negative cell line to the human Ia-positive Raji cell produced an interspecies hybridoma that expressed the murine Ia. These data are further evidence for the existence of trans-acting factors that can regulate class II gene expression. Furthermore, the T cell-derived lymphokine B cell stimulatory factor 1 (BSF-1) induced expression of class II genes in the R8205 cells. Exposure of R8205 cells to an antibody that has been shown to mimic BSF-1 activity on normal B cells also resulted in expression of class II genes. These data demonstrate that three distinct signals--a lymphokine, an alloantibody binding to membrane structures, and an interspecies trans-acting factor--can induce expression of class II genes.     

A mouse T cell product that preferentially enhances IgA production. I. Biologic characterization

Supernatants from a subset of helper T cell clones can enhance IgA, IgE, and IgG1 production in cultures of lipopolysaccharide-stimulated, T cell-depleted spleen cells. The lymphokine interleukin (IL)-4 has been shown to cause the IgE and IgG1 enhancement produced by these supernatants. IgA enhancement, however, is mediated by a factor distinct from IL-4, although IL-4 can potentiate the effect of the IgA-enhancing factor. IgA-enhancing factor is also distinct from IL-1, IL-2, IL-3, granulocyte-macrophage colony-stimulating factor, and interferon-gamma and acts directly on B cells. Purified IgA-enhancing factor enhances IgA production three- to sixfold yet causes less than a twofold increase in other isotypes. The IgA enhancing activity is not inhibited by concentrations of interferon-gamma that inhibit IL-4 activities. In the accompanying article, we show that this IgA enhancing activity is a novel property of the lymphokine IL-5.     

Characterization of monoclonal nonspecific suppressor factor (MNSF) with the use of a monoclonal antibody

The secretion of immunoglobulin (Ig) from cultured mononuclear cells by lipopolysaccharide (LPS) stimulation is inhibited by monoclonal nonspecific suppressor factor (MNSF), a lymphokine produced by murine T cell hybridoma. In an attempt to develop a murine monoclonal antibody (MAb) with specific reactivity against MNSF, a cell fusion technique that incorporated immune murine splenocytes and HAT-sensitive murine myeloma cells was used. Cross-reactivity experiments confirmed that the MAb (MO6) does not bind to unrelated proteins such as bovine serum albumin, mouse IgG, and murine interferon-gamma (IFN-gamma). There are no effects when anti-IFN-gamma antibodies are used with MNSF. As far as biological activity is concerned, MO6 inhibits in vitro the activity of MNSF in terms of the Ig secretion from cultured lymphocytes. By using MO6, affinity chromatography and immunoblotting were performed. The MNSF on the SDS-PAGE showed a band with m.w. of approximately 70,000, indicating the formation of an aggregate in saline; but after treatment with 0.4 M pyridine-acetic acid buffer, separate bands of 24,000 and 16,000 daltons were evident. Therefore MO6 recognizes 70,000 and both 24,000 and 16,000 daltons. Thus we confirmed by using this MAb and affinity chromatography, the existence of human counterpart, human nonspecific suppressor factor (hNSF), in supernatant from concanavalin A-stimulated T cells. When hNSF was fractionated by high pressure liquid chromatography (HPLC), the activity was found in a region corresponding to 70,000 daltons. However, when fractionated in pyridine-acetic acid buffer, hNSF activity was distributed in a slightly wider range of 15,000 to 30,000 daltons. Physicochemical analysis showed that the purified hNSF is resistant to either heating at 56 degrees C or to 2-mercaptoethanol treatment; however, it is labile to acidification at pH 2.0 and is also sensitive to protease treatment, the characteristics of which were similar to those of murine MNSF. Thus MO6 was confirmed to be a pertinent tool for isolation of hNSF, as well as for murine MNSF.     

Differential activity of granulocyte-macrophage and macrophage colony stimulating factors on bone resorption in fetal rat long bone organ cultures.

In this study, the ability of recombinant human macrophage (M) and murine granulocyte-macrophage (GM) colony stimulating factor (CSF) to affect both basal and stimulated bone resorption in fetal rat long-bone organ cultures was assessed. It was found that M-CSF does not affect basal bone resorption or bone resorption stimulated by parathyroid hormone, recombinant human interleukin 1 beta, prostaglandin E2 (PGE2), and 1,25 dihydroxy vitamin D3. Specifically, M-CSF at concentrations as high as 30 nM (1 microgram/mL) did not modulate 45Ca release from fetal rat long bones stimulated by these agents. The addition of recombinant murine GM-CSF (at equal molar concentration to M-CSF) also did not affect bone resorption stimulated by parathyroid hormone and interleukin 1 beta. On the other hand, GM-CSF stimulated basal bone resorption over a 120-h period and augmented the resorption mediated by exogenous PGE2 over a 48-h incubation. In addition, GM-CSF was shown to stimulate production of endogenous PGE2 in cultures of bone rudiments. These effects on bone resorption were blocked by the addition of prostaglandin synthesis inhibitors and specific antibodies to murine GM-CSF. These data indicate that M-CSF does not act as a regulator of bone turnover, but GM-CSF may cause bone resorption by stimulating the synthesis of PGE2 in bone.     

Purification and partial sequence of human osteoclast-activating factor: identity with interleukin 1 beta

The lymphokine osteoclast-activating factor (OAF) was purified to homogeneity. OAF was produced by human peripheral blood mononuclear cells stimulated with concanavalin A and phorbol myristate acetate under serum-free culture conditions. OAF was purified by sequential gel filtration, ion-exchange, and reverse-phase HPLC by following bone resorptive activity. Homogeneity was indicated by the criteria of a single 17,800-dalton band on silver-stained polyacrylamide gels, a single pI 6.8 band on isoelectric focusing, and a single aminoterminal sequence. Purified OAF stimulated half-maximal release of calcium from fetal rat long bones at a concentration of approximately 0.66 ng/ml. The amino-terminal sequence of OAF was determined and found to be identical to that of interleukin 1 beta. Homogeneous OAF possessed an activity of 8.2 X 10(6) U/mg in the thymocyte proliferation assay. Because the m.w., isoelectric point, amino-terminal sequence, and specific activity in the thymocyte proliferation assay are the same for homogeneous OAF and interleukin 1 beta, we conclude that they are the same molecule, and that interleukin 1 beta is the major protein with OAF activity produced by lectin-stimulated peripheral blood mononuclear cells.     

Activity of a partially purified human BCGF on murine assays for B-cell stimulatory factors. I. BCGF II-like activity of human BCGF

To further characterize a human B-cell growth factor (BCGF) produced by phytohemagglutinin (PHA) P-stimulated peripheral blood T cells, a partially purified preparation of this material was tested in a number of murine assays for B-cell stimulatory factors (BSF). Human BCGF lacked murine BSF-1 activity as assessed via the induction of polyclonal proliferation of anti-IgM-stimulated murine B cells; however, this material consistently augmented the proliferative response of murine B cells to anti-IgM and a saturating dose of murine BSF-1. Human BCGF also induced proliferation in unstimulated murine B cells, and augmented the proliferative response of dextran sulfate activated murine B cells. Human BCGF is therefore capable of causing proliferation of unstimulated and activated murine B cells, and by these criteria closely resembles murine BCGF II. In contrast to murine BCGF II, however, human BCGF failed to stimulate proliferation or immunoglobulin (Ig) secretion by murine BCL1 B lymphoma cells. A murine analog of this human BCGF showing the same pattern of biological responses was found in concanavalin A-stimulated supernatants of the murine MB2.1 T-cell line and D9-Cl T-cell hybridoma. The active component of the human BCGF preparation was not due to contaminating PHA, interleukin 1, interleukin 2; interferon-gamma, or endotoxin. Comparison between the above human BCGF and a commonly used source of murine BCGF II, i.e., supernatant from antigen-stimulated D10.G4.1 T cells, provided information suggestive of BCGF II heterogeneity. Both human BCGF and D10.G4.1 supernatant caused proliferation of unstimulated and dextran sulfate-stimulated murine B cells; however, only the human BCGF preparation augmented the proliferative response of murine B cells to anti-IgM and a saturating dose of murine BSF-1, and only the D10.G4.1 supernatant stimulated BCL1 cell proliferation and immunoglobulin secretion. The data therefore indicate that the different assays for BCGF II used in this study respond to different factors, and suggest the existence of two BCGF II-like activities.     

Human Macrophage-Activating Factors for Cytotoxicity

Two different factors (MAF-C I and MAF-C II) were obtained by anion exchange chromatography of the culture supernatant of a human T-cell hybridoma, H3-E9–6, which produces macrophage-activating factors for cytotoxicity (MAF-C). These 2 factors induced the cytotoxicity of monocytes synergistically as a priming signal (MAF-C I) and a triggering signal (MAF-C II), respectively. On gel filtration on a column of Superose 12, MAF-C II was eluted mainly at the void volume, whereas MAF-C I was eluted in the fractions corresponding to approximate molecular weights of 30–300 K. On the other hand, gel filtration in the presence of sodium deoxycholate revealed that MAF-C II has an approximate molecular weight of 40,000, but MAF-C I was unstable under these conditions. When the activity for mouse macrophages (MAF-Cm activity) was tested, the MAF-C II fraction showed high MAF-Cm activity in the presence of murine recombinant interferon gamma (rIFN-γ), but the MAF-C I fraction did not show MAF-Cm activity even in the presence of lipopolysaccharide (LPS). These results suggest that MAF-C I (priming lymphokine) has species specificity but MAF-C II (triggering lymphokine) does not.     

Alloreactive cloned T cell lines. VI. Multiple lymphokine activities secreted by helper and cytolytic cloned T lymphocytes

Culture supernatants generated by alloantigenic or lectin stimulation of a cloned helper T lymphocyte, designated L2, contain interleukin 2 (IL 2), granulocyte/macrophage colony-stimulating factor (CSF), B cell stimulating factor (BCSF), macrophage (Ia+)-recruiting factor (MIRF), (Ia+)-inducing activity, gamma-interferon, Fc receptor-enhancing activity, macrophage migration inhibitory factor (MIF), macrophage activation factor (MAF), interleukin 3 (IL 3), and a factor responsible for prolonging the synthesis and secretion of the fourth and second components of complement by guinea pig peritoneal macrophages. Erythropoietin was not detected. A spontaneously arising variant of L2, designated L2V, produces much lower quantities of macrophage-stimulating activities, IL 2, and interferon. However, when compared to L2, L2V produces much higher levels of BCSF, equivalent amounts of IL 3, and slightly smaller amounts of CSF. Unlike L2V, a cytolytic clone, designated L3, secretes lymphokines that primarily affect macrophage function. The time course of lymphokine production by L2 cells indicates that for the six lymphokine activities studied there are three different times at which maximal or near maximal levels are reached, as follows: 1) IL 2, 12 to 24 hr; 2) IL 3 and CSF, 24 to 48 hr; and 3) (Ia+)-inducing activity, MAF, and interferon, 48 hr or later. Only IL 2 activity disappears during the 8-day culture cycle. The time course data and the differential production of activities by the three types of lymphocyte clones suggest that at least four terminal effector lymphokine molecules account for the ten biologic activities tested.     

Requirement for three distinct lymphokines for the induction of cytotoxic T lymphocytes from thymocytes

The induction of cytotoxic T lymphocytes (CTL) from CTL precursors requires a combination of antigen and lymphokine signals. To investigate lymphokine requirements for CTL generation, we used an assay in which helper T cell and accessory cell-depleted spleen cells or whole thymocytes were cultured with lectin (Con A) and lymphokines. This culture was followed by assessment of lectin-dependent cytolysis. High concentrations of recombinant interleukin 2 (R-IL 2) (100 U/ml) alone were not sufficient for lectin-mediated CTL induction from thymocytes, whereas 20 to 100 U/ml of R-IL 2 alone could induce a significant lectin-mediated CTL response from accessory cell-depleted spleen cells. Using thymocytes as responders, we found purified or recombinant interferon-gamma (IFN-gamma) did not cause cytolytic activity either in the absence of or in the presence of R-IL 2. However, supernatant from Con A-stimulated rat spleen cells (rat Con A SN) in combination with R-IL 2 could induce cytolytic activity, suggesting that several factors are required for CTL induction. Con A SN was fractionated by gel filtration and the fractions were tested for ability to induce CTL. In the presence of a low level of R-IL 2 (5 U/ml), fractions with a Mr of approximately 31,000 could induce CTL, and this activity was referred to as CTL differentiation factor (CDF). The peak fractions containing CDF activity did not have detectable IL 1, IL 2, IFN-gamma, or CSF activity. However, by add-back experiments and the use of blocking antibodies, a monoclonal antibody against the IL 2 receptor or antibodies against murine IFN-gamma, we demonstrated that CTL induction from mature thymocytes (L3T4-, Lyt-2+) requires CDF activity in addition to IL 2 and IFN-gamma.