Fas ligand is not only expressed in immune privileged human organs but is also coexpressed with Fas in various epithelial tissues.

AIMS: To confirm the recent data obtained in mice, showing that the Fas ligand (FasL) is involved in the phenomenon of "immune privilege" (the apparent defect of the immune system in specific anatomical sites) and to extend this finding to humans. METHODS: The expression of FasL was analysed in a panel of histologically normal human tissues by reverse transcriptase polymerase chain reaction and Western blotting. The tissues sampled were brain, breast, bone marrow, oesophagus, kidney, liver, lung, lymph node, ovary, pancreas, pituitary gland, prostate, spleen, stomach (antrum and fundus), striated muscle, testis, thyroid, and uterus. These were obtained from patients with various neoplastic and non-neoplastic disorders; placental tissue was obtained after normal obstetric delivery, and spontaneous or voluntary abortion. RESULTS: Strong FasL expression was detected in testis and placenta. FasL expression was also detectable, although it was seen to a lesser extent, in oesophagus, prostate, lung, and uterus, which also coexpressed variable amounts of Fas mRNA or protein or both. The other organs tested for FasL expression were all negative. CONCLUSIONS: FasL in humans is expressed predominantly in immune "sanctuaries" such as testis and placenta, suggesting that, similar to mice, this expression may contribute to the immune privileged status of these organs, by preventing dangerous inflammatory responses. The coexpression of FasL and Fas in particular epithelia suggests that the physiological cell turnover of some tissues may be regulated by the Fas-FasL apoptotic pathway.

Evidence for intact CD28 signaling in T cell hyporesponsiveness induced by the HIV-1 nef gene.

Infection by human immunodeficiency virus (HIV)-1 is associated with quantitative and qualitative T cell alterations that severely impair the host's immune defense system. The molecular basis for this immunosuppression remains unclear. Peripheral blood mononuclear cells (PBMC) isolated from patients show markedly decreased interleukin (IL)-2 secretion but unaffected or even increased T helper (Th)2 cytokine production. T cell functional defects were recently reported to correlate more with T cell receptor (TcR) signaling, whereas signals provided by ligation of co-receptors CD27 and CD28 appeared to be preserved. Among the various mechanisms proposed to be involved in HIV-1-induced T cell dysfunction, we and others have reported that the nef gene product exhibited significant immunosuppressive activity. By using an inducible stably integrated nef gene, we demonstrated that Nef specifically down-regulated IL-2 and interferon (IFN)-gama produced upon TcR triggering. Here, using the same experimental system, we extended our initial observations to additional mitogenic signals, and investigated the co-stimulatory function of CD28. Nef down-regulated IL-2, but not IL-4 produced upon induction by combinations of mitogens that mimicked TcR signals together with CD28 mAb or CD28's natural ligand (CD80 and CD86). However, the co-signals provided by CD28 to up-regulate IL-2 induction were unaffected by Nef, since IL-2 produced by nef-transfected cells was proportionally enhanced to the same extent as that of control cells, either upon stimulation by the CD28 mAb or CD80 and CD86. In addition, phosphatidylinositol-3 kinase recruitment induced upon CD28 triggering was also found to be unaltered by nef expression. Together with the observation that similar levels of the Nef protein were detected in nef-transfected cells and upon infection of PBMC, these data suggest a selective immunosuppression induced by nef in human T cells by altering TcR signaling without detectable impact on CD28 co-receptor function. These data agree with the T cell defects observed in PBMC isolated from HIV-infected individuals.

Specific Th1 cytokine down-regulation associated with primary clinically derived human immunodeficiency virus type 1 Nef gene-induced expression.

HIV-1 infection is associated with a progressive and functional decline in the CD4+ lymphoid Th1 subset. Here, we propose that the HIV nef gene product may function as a specific regulator of Th1 cytokine production. By use of a T cell-specific inducible expression system, we show that upon T cell activation, induced nef expression down-regulated both IL-2 and IFN-gamma production in a dose-dependent manner, whereas IL-4, IL-9, IL-13, IL-8, and TNF-alpha production remained unaffected. In addition to this, independent transfected clones expressing various nef genes, including nef sequences amplified directly from an HIV-1 primary clinical isolate, displayed a similar pattern of cytokine expression. The specific Th1 impairment induced by nef, therefore, seems to be an important and conserved feature of HIV-1 infection and may represent a significant function of this viral gene in AIDS pathogenesis.

CD28 costimulation up-regulates long-term IL-2R beta expression in human T cells through combined transcriptional and post-transcriptional regulation.

Costimulatory molecules such as CD28 are required for induction of T cell clonal expansion and for prevention of T cell unresponsiveness. In combination with either CD3 or CD2 triggering, CD28 was shown to enhance T cell proliferation, cytolytic activity, production of cytokines and especially of IL-2, and expression of the IL-2R alpha-chain (IL-2R alpha). We and others have demonstrated that the costimulatory effect of CD28 on both IL-2 and IL-2R alpha expression results from a coordinated transcriptional activation of their genes and transcript stabilization. We show here that the CD28 stimulation, together with CD2, leads to a prolonged up-regulation of the constitutive expression of the IL-2R beta-chain in human peripheral T cells. As for IL-2R alpha, the increase in IL-2R beta gene expression seems to result from both transcriptional activation and transcript stabilization. In addition, IL-2 differentially regulates its own receptors, as only expression of IL-2R alpha, but not of IL-2R beta, is largely inhibited, at both the mRNA and protein levels, by blocking IL-2R mAbs. We propose that the long lasting T cell proliferation mediated by the CD2 and CD28 costimulation is mainly the consequence of the high and prolonged expression of both the IL-2R alpha- and beta-chains.

Regions of the CD8 molecule involved in the regulation of CD2-mediated activation.

The CD8 molecule regulates T cell activation mediated via the CD3 T cell receptor and the adhesion molecule CD2. CD8 mAbs have been found to inhibit early (Ca2+ rise) as well as late events (cytotoxicity, proliferation, and lymphokine secretion) mediated via the CD2 pathway. A panel of eight anti-human CD8 mAbs was tested for inhibition of CD2-mediated Ca2+ rise in a cytotoxic T cell clone. The inhibition ranged from 5 to 53% independently of mAb isotype and affinity measured by half saturation binding. We then characterized these mAbs for their reactivity toward three mutants of the human CD8 alpha carrying amino acid sequence changes in the surface-exposed loops homologous to the immunoglobulin CDR1, 2, and 3. The mutations included replacement of the human CD8 alpha CDR1- and CDR2-like loops by the homologous mouse sequences and the insertion of a glycine in the middle of the CDR3-like loop. Thus, five mAbs were found to be affected by the mutation in the CDR2-like loop but not by alterations in the other CDR-like loops. Conversely, the other two mAbs (8E1.7 and B9.8) were affected only by mutations in the CDR1- and CDR3-like loops, respectively. Cross-inhibition experiments were essentially in agreement with these results. Interestingly, all the mAbs directed against the CDR2-like loop were potent inhibitors of CD2-mediated Ca2+ rise, with one exception probably due to poor affinity. Thus, in addition to being a site of interaction with major histocompatibility complex Class I as recent data have indicated, this region of the CD8 alpha subunit may play a role in regulating T cell activation.

Modifications of leukemic blast cells induced by in vivo high-dose recombinant interleukin-2.

High-dose recombinant human Interleukin-2 was given to 21 patients with acute myeloid (n = 11) or lymphoid (n = 10) leukemia in relapse. A rapid decrease in the peripheral leukemic blasts numbers was observed in six patients. We were unable to demonstrate at the bone marrow level a diminution in the percentage of leukemic blasts. However an increase in the expression of the adhesion molecule CD54/ICAM-1(LFA-1 ligand) affected the leukemic bone marrow blasts of these six patients. This increase in CD54 was found in eight of the 11 (73%) AML and four out of the ten (40%) ALL blasts and CD58/LFA-3 (CD2 ligand) to a lesser extent. This increased expression was not associated with modifications in the expression of MHC class II molecules. In vivo IL-2 also dramatically modified the bone marrow T-cell subsets via the increase of CD3+ cells expressing the CD45RO 'memory' marker (six out of the eight tested patients) or CD54 (seven out of the eight tested patients). Altogether these results demonstrate that leukemic blasts can be affected by in vivo IL-2 via mechanisms that could involve T cells.

Functional expression of human CD28 in murine T cell hybridomas.

CD28 is a 44 kDa Ig superfamily cell surface molecule expressed on most mature T cells. Through its interaction with the recently identified B7/BB1 counter-receptor, it is believed to play an important role as a co-stimulator of T cells along with the TCR-CD3 complex. Activation of T cells with CD28 mAbs synergizes with TCR-CD3 and CD2 stimulation, resulting in long term T cell proliferation, differentiation of cytotoxic T cells and production of large amounts of cytokines. In order to further delineate the role of CD28 in signal transduction and T cell activation, human CD28 was transfected into CD3+ murine T cell hybridomas. High levels of cell surface CD28 expression was achieved by protoplast fusion. The transfected molecule retained all the native CD28 mAb epitopes found on human T cells. In these transfectants, CD28 mAbs, similarly to CD3 mAbs, were able to induce Ca2+ mobilization, IL-2 promoter induction (measured as beta-galactosidase activity in T cells hybridomas pre-transfected with the IL-2-lac Z reporter gene), IL-2 secretion, TNF alpha production and apoptosis (observed as growth arrest and genome fragmentation). The parental host cells, or cells transfected with vector alone, responded only to mAbs to CD3. IL-2 secretion in the transfectants was obtained using either an IgM mAb to CD28 or IgG mAbs presented on the surface of IgG-FcR+ B lymphoma cells. Optimal activation via CD28 was inhibited by suboptimal concentrations of soluble CD3 mAb, suggesting an interaction between the two pathways. The immunosuppressive drugs Cyclosporin A and FK506 completely blocked CD28 and CD3 mediated IL-2 production in these transfectants whereas rapamycin had only a partial inhibitory effect. Finally, since the transfected human CD28 molecule confers full functional responsiveness to the murine T cell hybridomas without the need for costimulators such as PMA, this model is ideal for studying the structure-function relationships of the CD28 molecule as well as the transmembrane and cytoplasmic associations implied in CD28 signaling.

Regulation of CD2-mediated human T cell activation: anti-CD8 monoclonal antibodies inhibit CD2-mediated rise in intracellular calcium.

The human CD8 glycoprotein regulates the function of cytotoxic T cells activated by antigenic peptide as well as via CD3 or CD2 mAbs. Activation of T cells by CD2 mAbs requires two mAbs directed against distinct CD2 epitopes and induces tyrosine phosphorylation, PI-PLC activity generating the second messengers, IP3 and DAG, and finally lymphokine secretion. We have investigated the role of the CD8 alpha molecule in CD2-mediated activation of human cytotoxic T cell clones and CD8+ resting T cells. CD8 alpha-specific mAb inhibited 60% of the allospecific cytotoxicity of the CD8+ clone against its target cell and 86% of the CD2-redirected killing against the HLA Class I-negative Daudi target cell. In addition, CD8 alpha-specific mAb inhibited CD2-mediated TNF alpha and IL2 secretion by the CD8+ clone. Furthermore, CD8 alpha-specific mAb inhibited the increase in intracellular ionized calcium mediated by CD2 mAbs in the CD8+ clone and in purified T cells. Since the [Ca2+]i recruitment from internal stores induced by CD2 mAbs was inhibited, the inhibitory effect induced by the CD8 alpha-specific mAb probably acts on the PI-PLC activation pathway. This inhibition mechanism involves neither a decrease in affinity of CD2 mAb for its target nor a decrease in CD2 cell surface expression or a rise in cAMP known as an inhibitor of the CD2-mediated PI-PLC activity. These results suggest that the inhibitory mechanism induced by the CD8 mAb may prevent the activation of the PI-PLC activity, probably through the CD8 alpha-associated protein tyrosine kinase p56lck.

Anti-T11.1 and -T11.2 monoclonal antibodies play a different role in CD2-mediated signal transduction.

We comparatively evaluated (Ca2+)i mobilization after triggering with a stimulatory pair of CD2 (CD2.9, anti-T11.1 + CD2.1, anti-T11.2) or CD3 mAbs in the differentiated T-cell line Jurkat, using INDO-1 labeling and cytofluorimetry. The results obtained showed different (Ca2+)i mobilization kinetics following CD2 or CD3 stimulation (the former being slower than the latter), not due to different association kinetics of mAbs. In a nonreciprocal manner, however, preliminary interaction with CD2.1 (anti-T11.2) followed by CD2.9 (anti-T11.1) induces a rapid (Ca2+)i rise, similar to CD3 stimulation, as shown by preincubation experiments. There is no interference between CD2.9 and CD2.1 mAb binding. CD2.1 mAb by itself is unable to induce (Ca2+)i mobilization; in addition, preincubation with CD2.1 mAb did not modify the CD2, CD3, CD45, or CD28 immunoprecipitation patterns. Triggering of the epitope recognized by CD2.1 mAb may favor, possibly via conformational changes of CD2 molecule or (Ca2+)i-unrelated metabolic effect(s), optimal signal transduction.

Interleukin-7 is a potent co-stimulus of the adhesion pathway involving CD2 and CD28 molecules.

Co-stimulation of highly purified peripheral T lymphocytes from healthy blood donors with the adhesion molecules CD2 and CD28 in association with recombinant interleukin-7 (rIL-7) induced T-cell proliferation, multiple cytokine secretion and IL-2 receptivity. We demonstrated that rIL-7 is as potent as rIL-2 in inducing the proliferation of unseparated, CD4+ and CD8+ T cells. In contrast to low or undetectable levels of IL-1 alpha, IL-6 and IL-2, high levels of tumour necrosis factor-alpha (TNF-alpha), IL-4 and granulocyte-macrophage colony-stimulating factor (GM-CSF) were secreted. Experiments using blocking antibodies suggested a direct mechanism for rIL-7 co-stimulatory effect, although induction of the CD25/IL-2 receptor alpha-chain (CD25/IL-2R alpha) was observed. Monoclonal antibodies (mAb) against the adhesion molecules CD2 and CD28 are likely to mimic the interaction with their respective physiological ligands [lymphocyte function-associated antigen-3 (LFA-3)/CD58, CD59 and CD48 for CD2, B7/BB1 for CD28]. Taken together, these in vitro data suggest that IL-7 could participate in paracrine interactions between T lymphocytes and thymic stromal cells or dendritic cells, via its potent co-stimulatory activity with CD2 and CD28 adhesion molecules.

Signalling through CD28 T-cell activation pathway involves an inositol phospholipid-specific phospholipase C activity.

Stimulation of the human T-cell line, Jurkat, by a monoclonal antibody (mAb) directed against the CD28 molecule leads to sustained increases in intracellular levels of Ca2+ ([Ca2+]i); the initial rise in Ca2+ comes from internal stores, followed by Ca2+ entry into the cells. The CD28 molecule also appears to activate polyphosphoinositide (InsPL)-specific phospholipase C (PLC) activity in Jurkat cells, as demonstrated by PtdInsP2 breakdown, InsP3 and 1,2-diacylglycerol generation and PtdIns resynthesis. We also observed that interleukin-2 (IL2) production induced via CD28 triggering was sensitive to a selective protein kinase C inhibitor. Of the four other anti-CD28 mAbs (CD28.2, CD28.4, CD28.5, CD28.6) tested, only one (CD28.5) was unable to generate any InsPL-specific PLC or IL2 secretion. However, the cross-linking of cell-bound CD28.5 with anti-mouse Ig antibodies led to an increase in [Ca2+]i. CD28-molecule clustering in itself appears to be a sufficient signal for induction of PLC activity.

The CD2 and CD28 adhesion molecules induce long-term autocrine proliferation of CD4+ T cells.

In vitro human T lymphocyte activation requires two-signal triggering delivered by lectins, phorbol esters or antibodies directed against surface molecules. Stimulation of adhesion molecules by CD2 and/or CD28 antibodies defines alternative activation pathways. Activation by CD2 + CD28 monoclonal antibodies induces high-level, long-lasting and monocyte-independent proliferation of highly purified T cells. Limiting dilution cultures showed that CD28 in association with CD2 or CD3, without addition of exogenous cytokines, induced single-cell proliferation. CD2 + CD28 stimulation induced long-term interleukin (IL)-2-dependent autocrine proliferation of CD4+ T cell clones. We tried to elucidate this long-term proliferation by evaluating cytokine secretion and cytokine dependency. CD28 associated to CD3 or CD2 induced high levels of IL-2, tumor necrosis factor (TNF) and IL-4 secretion for 10 days, in contrast to CD3 alone which induced only TNF secretion. Cytokines of the monocytic lineage were also secreted, such as colony-stimulating factor-1, granulocyte-macrophage colony-stimulating factor or IL-1, the latter being more specific of CD2 + CD28 activation. Blocking antibodies confirmed the crucial role of IL-2 in CD2 + CD28 activation. Anti-IL-4, anti-IL-7 receptor or anti-TNF antibodies had no effect on proliferation. Stimulation with CD2 + CD28 induced long-term autocrine (at least for IL-2) proliferation for CD4+ T cells, with no evidence for the implication of another cytokine among those tested other than IL-2. This represents a model for long-term autocrine growth for non-leukemic cells.

CD28 mAbs with distinct binding properties differ in their ability to induce T cell activation: analysis of early and late activation events.

A panel of eight different CD28 mAbs was used to analyse the structure-function relationships of the CD28 molecule. The results of binding inhibition experiments show a complex and heterogeneous pattern of inhibition; however a subgroup of mAbs was identified, namely CD28.1, CD28.3, and CD28.5, which exhibited almost identical inhibition profiles. To test the hypothesis that the different binding specificities are related to functionally distinct subregions of the CD28 molecule, the ability of each mAb to (i) induce IL-2 release and (ii) increase intracellular calcium [(Ca2+)i] in Jurkat T cells was analysed. The results show that the mAbs CD28.1, CD28.3, and CD28.5 are almost totally unable to induce IL-2 release, and their ability to increase (Ca2+)i is relatively low. All other mAbs are able to induce a marked (Ca2+)i rise, however they strongly differ in their ability to induce IL-2 release. Such differences cannot be explained by differences in the isotypes or binding kinetics of the mAbs. These results imply the existence of functionally distinct subregions on the CD28 molecule. In addition, the (Ca2+)i rise may be associated with either high or low IL-2 secretion following CD28 triggering.

Differential immuno-suppressive effects of metabolic inhibitors on T-lymphocyte activation.

An important challenge in the field of auto-immune diseases, bone marrow and organ transplantation is the control of T-lymphocyte activation. To gain more insight into the in vitro correlation of immunosuppression, we investigated the effects of cyclosporin A (CSA) and two other metabolic inhibitors on cytokine secretion and T-cell proliferation. Secretion of TNF-alpha and GM-CSF was much more resistant to metabolic inhibitors than proliferation or synthesis of IL-1 alpha or IL-2. Moreover, our data suggested that the regulation of IL-1 alpha production in T-cells was CSA and protein kinase C (PKC)-dependent, as opposed to monocytes regulation. The receptivity to the epithelial cell-derived cytokine IL-7, associated either with antigen-dependent or independent triggering, was almost similarly inhibited by cyclosporin A, forskolin or PKC inhibitor, in sharp contrast to IL-2 receptivity. In this latter case, CD28+ IL-2 stimulation was more sensitive to both forskolin and PKC inhibition than that of CD2 or CD3+ IL-2. With regard to CSA effects, limiting dilution analysis provided evidence for some heterogeneity at the clonal level. This strongly suggested that T-cell functional monitoring at the population level does not truly reflect the actual immunosuppression. Additional experiments are required to evaluate the sensitivity to metabolic inhibitors of T-lymphocyte activation via the natural ligands of CD2 and CD28.

Therapeutic applications of anti-CD4 antibodies.

The CD4 molecule represents a major functional T-cell surface molecule, defining an important T-cell subset, which also is expressed on monocyte, dendritic, and Langerhans cells. Various in vivo studies have demonstrated its implication in various steps of physiological T-cell activation: 1. CD4 interacts with its physiological ligand, the class II molecules, thus increasing the affinity of the conjugation between CD4+CD(3+)-TCR+ and class II+ antigen-presenting cells. 2. Through CD4, the signal transduction machinery is stimulated via its association with p56lck. In addition, CD4 has proved to be the receptor for gp120, the surface glycoprotein of HIV, that allows the virus to penetrate the CD4+ T cells and monocytes. Based on in vitro studies in various animal models, CD4 mAbs have proved to be efficient in the prevention and/or therapy of a variety of immunologically based diseases: 1. When injected early in the prodromic phase of autoimmune diseases (AID) such as diabetes, either delay or prevention is achieved with or without maintenance after therapy. 2. These mAbs have proved to be self-tolerogenic, thus allowing prolonged in vivo therapy and suppression of immunogenicity of mAb of a distinct specificity. In humans, CD4 mAbs are, or could be, used and evaluated in AID (lupus, diabetes, rheumatoid arthritis, etc.), transplantation, leukemias and lymphomas expressing CD4, and, finally, in AIDS patients, in whom CD4 mAbs can block HIV-CD4 binding and deliver a negative signal to T cell, thus blocking T-cell activation and HIV transcription. CD4 mAbs at least provide evidence that the CD4 molecules are suitable for immunomodulation and could be the target for a new pharmacological antagonist.

Ig CDR3-like region of the CD4 molecule is involved in HIV-induced syncytia formation but not in viral entry.

The HIV1 envelope glycoprotein gp120 binding site has been previously mapped by genetic studies to the CDR2-like region of the first domain of the CD4 molecule. mAb reactive with epitopes linked to this region (e.g., OKT4A) inhibit both HIV entry into CD4-positive cells and syncytia formation. A second area of this domain, the CDR3-like region, has been shown to be involved in gp120-CD4 interactions, but its role remained so far unclear. We show here that a mAb specific for the CDR3-like region of the CD4 receptor, 13B8-2, actually blocks soluble gp120 binding to CD4, inhibits HIV-induced cell-cell fusion, and prevents viral production by infected cells. However, this mAb fails to inhibit the binding of viral particles to cell-surface CD4 and their entry into CD4-positive cells. These results strongly suggest i) that soluble gp120 and virion-anchored gp120 bind CD4 in distinct manners, ii) that gp120-CD4 interactions required for viral entry and syncytia formation are different, and iii) that mAb binding to the CDR3-like region of the first domain of CD4 affects a post-entry step of the HIV replicative cycle.

Prolonged IL-2 receptor alpha/CD25 expression after T cell activation via the adhesion molecules CD2 and CD28. Demonstration of combined transcriptional and post-transcriptional regulation.

The T cell adhesion molecule CD28 provides a costimulatory signal in combination with CD2 and CD3 mAb. CD28 regulates the expression of cytokines by T cells, not only IL-2, but also IL-1 alpha and CSF-1, usually synthesized by accessory cells. We have investigated the mechanisms through which CD28 modulates the expression of the IL-2R alpha chain. Whereas activation through CD2 or CD28 alone induced no or only low IL-2R alpha chain expression, activation through CD2 plus CD28 led to both a high and prolonged (greater than 14 days) cell surface and mRNA expression. In contrast, immobilized CD3 mAb-dependent activation induced a transient expression of the IL-2R alpha chain, which was neither further increased nor prolonged by CD28 costimulation. Upon CD2 plus CD28 stimulation, the half-lives of the two IL-2R alpha transcripts increased progressively between days 1 and 4, in contrast to each pathway alone. Whereas each activation pathway alone induced either no (CD2) or low (CD28) levels of IL-2R alpha gene transcription, the CD2 plus CD28 stimulation was associated with its increased transcription, which persisted at similar rates between 5 and 96 h post stimulation. The in vitro costimulation via the CD2 and CD28 molecules thus regulates the expression of the IL-2R alpha gene both at the transcriptional and post transcriptional levels. Our results therefore demonstrate a new immunoregulatory function of the CD28 molecule on IL-2R alpha expression, which, through its increased transcription rate and stabilization, could, together with high levels of cytokines secretion, be responsible for the prolonged T cell proliferation.

Contributions of the CD2 and CD28 T lymphocyte activation pathways to the regulation of the expression of the colony-stimulating factor (CSF-1) gene.

The T cell adhesion molecule CD28 provides a costimulatory signal, in combination with either CD2 or CD3 mAb. CD28 appears to regulate the expression, by T cells, of cytokines normally produced by accessory cells, namely IL-1 alpha, TNF-alpha, and CSF-1. The CSF-1 gene is expressed as a 4.0-kb transcript by human T cells activated with mAb directed against CD2 and CD28, alone or in combination. A kinetic analysis of its expression showed low steady-state levels of the transcript after CD2 stimulation, and a transient rise after CD28 stimulation. In contrast, a mean fivefold increase in the levels of the transcript was detected in CD2 plus CD28-stimulated cells. The potential mechanisms regulating this increase were investigated. The half-life of the CSF-1 transcript was identical in cells activated with either CD28 or CD2 plus CD28. Transcription of the CSF-1 gene appeared to undergo attenuation in resting cells as well as in cells activated via a single pathway; this attenuation was relieved in part by the combined CD2 plus CD28 stimulation. Thus in vitro costimulation via the CD2 and CD28 molecules regulates the expression of the CSF-1 gene mainly at the transcriptional level.