The binding of granulocyte-macrophage colony stimulating factor (GM-CSF) to its receptor stimulates JAK2 protein kinase activation, protein phosphorylation, and JAK2 association with the beta c chain of the GM-CSF receptor. To better understand how different domains of the JAK2 function to regulate association and phosphorylation of the beta c receptor, the minimal portion of the beta c receptor necessary for JAK2 binding has been determined. Using glutathione S-transferase (GST) fusion proteins expressing different portions of the membrane-proximal domain of the beta c chain, we demonstrate that JAK2 binds to amino acids 458-495, but showed little binding to fusion proteins containing amino acids 483-559, 483-530, or 458-484. The GST-beta c 458-495 bound equally well to the wild type (WT) JAK2, a carboxyl-terminal deletion of JAK2 removing the protein kinase domain (amino acids 1000-1129), and a deletion of the kinase-like domain (amino acids 523-746). However, an amino-terminal JAK2 deletion (amino acids 2-239) markedly reduced binding to this GST-beta c. Far Western blotting demonstrated that a GST fusion protein containing amino acids 1-294 of JAK2, but not fusion proteins containing amino acids 295-522, 523-746, or 747-1127, bound GST-beta c 458-559. When the JAK2 WT and deletions were transiently expressed along with the alpha and beta c subunits of the GM-CSF receptor and the cells were treated with GM-CSF, the following results were obtained: 1) WT JAK2 phosphorylated the beta c subunit in a GM-CSF-dependent manner, 2) the kinase-like domain deletion phosphorylated the beta c subunit, and 3) both the kinase domain deletion and the amino-terminal deletion failed to stimulate phosphorylation of the beta c subunit. Therefore, phosphorylation of the beta c subunit requires the binding of JAK2 through its amino terminus.

Single cell studies have identified intraclonal heterogeneity of cytokine production by activated T cells. To investigate implications of cytokine heterogeneity for cell fate, an interleukin (IL)-2 promoter-green fluorescent protein (GFP) reporter transgenic model was developed to track IL-2+ and IL-2- T cells during differentiation from naive precursors. Antigen-activated IL-2+ and IL-2- cells had comparable proliferative capacities in primary responses. However, T cells that expressed IL-2 in primary responses demonstrated enhanced antigenic sensitivity and increased expression of effector cytokines in secondary responses in vitro and in vivo. Thus, heterogeneity of activation during a primary response translates into heterogeneous secondary responses, in which enhanced memory/effector function is linked to cells that previously exceeded an activation threshold associated with IL-2 gene transcription.

Phorbol ester tumor promoters activate gene transcription by regulating both the synthesis and posttranslational modification of the activator protein 1 (AP-1) transcription factor. c-Jun and JunB are components of the mammalian AP-1 complex. Here we demonstrate that in U-937 human leukemic cells, phorbol esters stimulate the phosphorylation of the amino terminus of human c-Jun (JUN) but not human JunB (JUNB). Mutational analysis indicates that serine-63 and -73, which reside within the putative regulatory domain of JUN, are required for both constitutive and phorbol 12-myristate 13-acetate-inducible N-terminal JUN phosphorylation. To determine the functional role of this N-terminal phosphorylation, we prepared several chimeric proteins containing the N-terminal 84 amino acids (positions 5-89) of human JUN or murine JUNB fused to the yeast GAL4 DNAbinding domain. This region was found to be sufficient for the phorbol ester-inducible transcriptional activity ofJUN, but not JUNB. This induction was abolished by the mutation of serine-63 and -73 to leucine residues. Thus, we propose that phorbol esters enhance the trans-activation potential of JUN, but not JUNB, by the phosphorylation of the N-terminal

Analysis of the IFN-γ promoter has primarily been conducted by transient expression of reporter constructs in transformed cells. However, the activity of cis elements may differ when expressed transiently compared with their activity within native chromatin. Furthermore, the transcription factors and signaling mechanisms in transformed cells may differ from those in normal T cells. To analyze IFN-γ promoter regulation in normal T cells, we developed a novel retroviral bottom-strand reporter system to allow the chromatin integration of promoter regions in primary developing T cells. As controls, both the IL-2 and IL-4 promoters were inducible in this system, with the IL-4 reporter having Th2-specific activity. Strikingly, the IFN-γ promoter exhibited constitutive activity in both Th1 and Th2 subsets, in contrast to the behavior of the endogenous IFN-γ gene, which is inducible only in Th1 cells. In mapping this activity, we found that the AP-1/GM-CSF site in the distal promoter element is the most critical element for the constitutive activity. Transgenic reporter lines for the IFN-γ promoter confirmed the constitutive behavior of the isolated IFN-γ promoter. This constitutive activity was resistant to inhibition by cyclosporin A and was independent of Stat4 and p38 mitogen-activated protein kinase. These results suggest that IFN-γ promoter regulation may require cis elements residing either downstream or >3.4 kb upstream of the transcriptional start site, involving repression of constitutive activity.

To understand the mechanism by which phorbol esters (PMA) stimulate c-jun transcription in human leukemic cell line U937, we have mutated specific enhancer sequences within the c-jun promoter. We find in the region of DNA from -132 to +170 containing Sp1, C-TF and AP-1 sequences that mutation of the AP-1 sequence alone is not sufficient to abrogate transcription, and mutation of the Sp1 sequence increases transcription 4-fold. Although mutation of the CTF site had no effect, CTF and AP-1 mutations together totally abrogate PMA-induced transcription. In comparison mutations of either of these sites alone or together in a construct containing -1639/+740 of the c-jun promoter had no effect on transcription. Because this data suggested the possibility of other upstream control regions, we sequenced the promoter from -142 to -1639. This sequence demonstrates a greater than 70% homology between human, and mouse c-jun promoters for the region from -142 to -441, and a second AP-1-like site in the -183 to -192 region. Mutation of this site did not influence transcription by PMA. By making constructs containing varying portions of the promoter, we have identified the region between -142 and -711 to be responsible for mediating PMA-induced c-jun transcription.

HIV-1 replicates preferentially in IL-4 producing CD4 T cells for unclear reasons. We show increased HIV-1 expression is irrespective of viral tropism for chemokine receptors as previously suggested, but rather transcription of the HIV-1 long terminal repeat (LTR) is increased in IL-4 producing CD4 T cells. Increased expression of HIV-1 message is also confirmed in IL-4 producing CD4 T cells from HIV-1 infected individuals ex vivo. In exploring a transcriptional mechanism, we identify a novel c-maf (required for IL-4 expression) transcription factor binding site just upstream of the dual NFκB/NFAT binding sites in the proximal HIV-1 LTR. We demonstrate c-maf binds this site in vivo and synergistically augments HIV-1 transcription in cooperation with NFAT2 and NFκB p65, but not NFAT1 nor NFκB p50. Conversely, siRNA inhibition of c-maf reduces HIV-1 transcription in IL-4 producing T cells. Thus, c-maf increases HIV-1 expression in IL-4 producing CD4 T cells by binding the proximal HIV-1 LTR and augmenting HIV-1 transcription in partnership with NFAT2 and NFκB p65, specifically. This has important implications for selective targeting of transcription factors during HIV-1 infection since, over the course of HIV-1 progression/AIDS, IL-4 producing T cells frequently predominate and substantially contribute to disease pathology.

Bryostatin 1, a macrocyclic natural lactone isolated from a marine Bryozoan, has undergone phase I testing in humans. Side effects of treatment have included muscle pain and joint aches, a transient decrease in platelets, and the release of tumor necrosis factor alpha (TNF alpha) and IL-6 into the blood stream. In animals, anticancer activity has been demonstrated against murine leukemias, lymphomas, melanomas, and sarcomas. The mechanism of action of this compound depends in part on its ability to activate protein kinase C. To determine the biologic activity and toxicity of other members of the family of bryostatin compounds, we studied the ability of bryostatins 5 and 8 to inhibit the growth of murine melanoma K1735-M2. Bryostatins 1, 5, and 8 induced equivalent inhibition of melanoma growth, but bryostatins 5 and 8 induced less weight loss than bryostatin 1 (P < 0.001). Neither the injection of an antimurine TNF alpha antibody nor an adenovirus, which produces a mutated TNF receptor inhibiting TNF alpha activity, into mice had any effect on either bryostatin-induced weight loss or melanoma tumor growth inhibition. Using a novel competition assay, the levels of bryostatin in the plasma were measured. The approximate half-life (t1/2) of bryostatin was 8.62 min, the clearance (Cl) 3.53 ml/min and the AUC 322.20 nmol/l min. A similar result was obtained with each bryostatin analog. These results suggest that human testing of additional bryostatin analogs may yield compounds with similar antitumor activity but decreased side effects. A novel assay to measure the level of all bryostatins in the plasma of patients undergoing treatment is described.

Bryostatin 1, a macrocyclic lactone, has undergone phase I trials as an anticancer agent. Because of the lipid solubility of this compound it must be delivered either in ethanol or in a PET formulation. During the trial, these vehicles caused a large number of treatment-related side effects. We have synthesized the triethanolamine salt of 26-succinylbryostatin 1 and find that this compound is approx. 100-fold more water soluble than bryostatin 1. Because of the potential for clinical use, we have evaluated the biologic activity of this compound. We find that in a concentration-dependent manner 26-succinylbryostatin 1 is capable of activating protein kinase C (PKC) in vitro and displacing [3H]PDBu from PKC. However, at all concentrations tested the activity was less than the parent compound bryostatin 1. Addition of bryostatin 1 but not 26-succinylbryostatin 1 to U937 leukemic cells in culture stimulated a drop in cytosolic PKC, secondary to translocation of PKC to the membrane. Although 26-succinylbryostatin 1 did not stimulate a drop in the cytosolic levels of PKC, addition to U937 cells activated transcription from an AP-1 enhancer construct and c-Jun protein phosphorylation in a similar fashion to bryostatin 1 and differentiation of U937 cells. Unlike bryostatin 1, 26-succinylbryostatin 1 was unable to cause aggregation of human platelets. Although injection of bryostatin-1 into mice carrying B16 melanoma inhibits tumor growth, there was no significant inhibition of melanoma growth when identical doses of 26-succinylbryostatin 1 were injected. Therefore, 26-succinylbryostatin 1 shares some but not all of the pharmacologic properities of bryostatin 1. This compound can activate protein phosphorylation without lowering cytosolic levels of PKC.