Search

To explore the basis for interleukin (IL)-2 receptor (IL-2R) signaling specificity, the roles of tyrosine-based sequences located within the cytoplasmic tails of the beta and gammac chains were examined in the murine helper T cell line HT-2. Activation of the Janus kinase/signal transducer and activator of transcription (JAK/STAT) pathway, cellular proliferation, and the induction of various genes were monitored. All four of the cytoplasmic tyrosine residues as well as the distal portion of the gammac proved dispensable for the entire spectrum of IL-2R signaling responses studied. Conversely, select tyrosine residues within the beta chain were essential and differentially required for various signaling events. Specifically, activation of c-fos gene expression was found to occur exclusively through the most membrane proximal tyrosine, Tyr-338, whereas proliferation and the activation of STAT-5 were induced either through Tyr-338 or through the two C-terminal tyrosine residues, Tyr-392 and Tyr-510. These tyrosine residues mediated the induction of two different STAT-5 isoforms, which were found to form heterodimers upon receptor activation. In contrast to the tyrosine dependence of c-fos and STAT-5 induction, bcl-2 gene induction proceeded independently of all IL-2Rbeta tyrosine residues. Thus, the tyrosine-based modules present within the IL-2Rbeta cytoplasmic tail play a critical role in IL-2R signaling, mediating specificity, redundancy, and multifunctionality.

The life cycle of human immunodeficiency virus type 1 (HIV-1) is critically dependent on the transregulatory proteins Tat and Rev. Tat increases the production of HIV-specific mRNAs by direct binding to the transactivation response (TAR) element located at the 5' end of all HIV transcripts. In contrast, Rev uses a complex RNA stem loop structure, the Rev response element (RRE), which is found in full-length and singly spliced HIV transcripts. Rev is required for the cytoplasmic expression of full-length mRNAs encoding Gag, Pol, and Env structural proteins. The complex intracellular interactions between Tat, Rev, host cell factors, and their respective RNA response elements should be susceptible to interdiction by genetic therapies designed to introduce and express novel genetic information. We show that the expression of antisense RREs inhibited the cytoplasmic expression of RRE containing HIV-1 transcripts. HIV-based retroviral vectors containing either the antisense (-) or sense (+) RREs inhibited HIV replication in transient transfections. The production of full-length HIV mRNA was also decreased significantly by the expression of RREs in either orientation. Interestingly, there was a paradoxic increase in HIV p24 gag production at low levels of inhibitor; this effect may have been the result of encapsidation of RRE-containing HIV-based retroviral vectors. The data suggest that the introduction and inducible expression of RRE-containing, HIV-based retroviral vectors may have therapeutic value in HIV infection.

Interleukin 2 (IL-2) mediated signalling results from ligand binding and subsequent heterodimerization of IL-2r beta and gamma c. The high-affinity IL-2 receptor (IL-2r) is a heterotrimer comprised of the IL-2r alpha, IL-2r beta and gamma c subunits. Whereas human IL-2 effectively binds to either human or murine lymphocytes, murine IL-2 binds with markedly higher affinity to murine receptor complexes than to human complexes. Using cell lines stably expressing heterotrimeric IL-2r that vary in the species origin of individual subunits, we have demonstrated that IL-2r alpha is primarily responsible for the species specificity of IL-2 binding. Studies of ligand binding to the low affinity receptor demonstrated that IL-2r alpha displays a similar species preference to the heterotrimeric complex. Moreover, differences in ligand binding are reflected in differences in proliferation. A cell line expressing human IL-2r alpha and IL-2r beta along with murine gamma c vigorously proliferated only in response to human IL-2 at low doses, while both human and murine IL-2 stimulated proliferation of a cell line containing murine IL-2r alpha (as well as human IL-2r beta and murine gamma c). Therefore, IL-2r alpha is the chain primarily responsible for the species specificity of ligand binding.

Animal lectins play important roles in a variety of biological processes via their recognition of glycoconjugates. Galectin-3 is a beta-galactoside-binding lectin previously designated as epsilon BP (IgE-binding protein), CBP35, Mac-2, L-29, and L-34, and its expression has been associated with various physiological and pathological processes, including cell growth, tumor transformation, and metastasis. Galectin-3 is widely distributed in various tissues and cell types and is expressed in many leukocytes, with the notable exception of B and T lymphocytes. We now report that galectin-3 is abundantly expressed in a number of human T lymphotropic virus (HTLV)-I-infected human T cell lines, including F6T, HUT 102, K3T, MT-2, and SLB-I, but is not expressed in non-HTLV-I-infected T cell lines such as Jurkat, CEM, and MOLT-4. In addition, the galectin-3 level was markedly increased in human thymocytes after infection with HTLV-I as compared with uninfected thymocytes. The up-regulation of galectin-3 expression appeared to correlate well with HTLV-I gene expression, as undetectable or very low levels of galectin-3 were found in the S1T and ATL-1K cell lines, which are nonproductively infected with HTLV-I. In co-transfection experiments, the galectin-3 promoter was significantly up-regulated by expression vectors encoding the 40-kd Tax protein, a potent transactivator in HTLV-I. Analysis of various Tax mutants suggested that galectin-3 promoter induction is dependent on activation of the cyclic-AMP-responsive element binding protein/activation transcription factor family of transcription factors and, to a lesser extent, nuclear factor-kappa B/Rel induction. Transfection of human promonocytic U-937 cells with an HTLV-I Tax expression vector induced galectin-3 expression in this cell line. Functionally, galectin-3 was shown to activate interleukin-2 production in Jurkat T cells. Together, these findings raise the possibility that HTLV-I Tax production induces the transcription and subsequent synthesis and secretion of galectin-3, which in turn may further activate these T cells and contribute to the altered properties of cell growth found in adult T cell leukemia induced by HTLV-I.

Nuclear expression and consequent biological action of the eukaryotic NF-kappa B transcription factor complex are tightly regulated through its cytoplasmic retention by an ankyrin-rich inhibitory protein termed I kappa B alpha. I kappa B alpha specifically binds to and masks the nuclear localization signal of the RelA subunit of NF-kappa B, thereby effectively sequestering this transcription factor complex in the cytoplasm. Specific cellular activation signals lead to the rapid proteolytic degradation of I kappa B alpha and the concomitant nuclear translocation of NF-kappa B. However, the precise biochemical mechanisms underlying the inhibitory effects of I kappa B alpha on RelA and its inducible pattern of degradation remain unclear. By using HeLa cells transfected with various cDNAs end-coding epitope-tagged mutants of I kappa B alpha, our studies demonstrate the following: (i) sequences within the 72-amino-acid N-terminal region of I kappa B alpha are required for tumor necrosis factor alpha (TNF-alpha)-induced degradation but are fully dispensable for I kappa B alpha binding to and inhibition of RelA; (ii) serine residues located at positions 32 and 36 within the N-terminal region of I kappa B alpha represent major sites of induced phosphorylation (substitution of these serine residues with alanine abrogates TNF-alpha-induced degradation of I kappa B alpha); (iii) the C-terminal 40 residues of I kappa B alpha (amino acids 277 to 317), which include a PEST-like domain, are entirely dispensable for TNF-alpha-induced degradation and inhibition of RelA; (iv) a glutamine- and leucine-rich (QL) region of I kappa B alpha located between residues 263 and 277 and overlapping with the sixth ankyrin repeat is required for both inducible degradation and inhibition of RelA function; (v) regulation of I kappa B alpha degradation by this QL-rich region appears to occur independently of phosphorylation at serines 32 and 36. These findings thus indicate that I kappa B alpha is generally organized within distinct modular domains displaying different functional and regulatory properties. These studies have also led to the identification of a novel class of dominant-negative I kappa B alpha molecules that retain full inhibitory function on NF-kappa B yet fail to undergo stimulus-induced degradation. These molecules, which lack N-terminal sequences, potently inhibit TNF-alpha-induced activation of the human immune deficiency virus type 1 kappa B enhancer, thus indicating their possible use as general inhibitors of NF-kappa B.

The specific signal transduction function of the gamma c subunit in the interleukin (IL) 2, IL-4, IL-7, IL-9, and IL-15 receptor complexes remains undefined. The present structure-function analyses demonstrated that the entire cytoplasmic tail of gamma c could be functionally replaced in the IL-2 receptor (IL-2R) signaling complex by a severely truncated erythropoietin receptor cytoplasmic domain lacking tyrosine residues. Heterodimerization of IL-2R beta with either gamma c or the truncated erythropoietin receptor chain led to an array of specific signals normally derived from the native IL-2R despite the substitution of Janus kinase JAK2 for JAK3 in the receptor complex. These findings thus suggest a model in which the gamma c subunit serves as a common and generic "trigger" chain by providing a nonspecific Janus kinase for signaling program initiation, while signal specificity is determined by the unique "driver" subunit in each of the gamma c- containing receptor complexes. Furthermore, these results may have important functional implications for the asymmetric design of many cytokine receptor complexes and the evolutionary design of receptor subfamilies that share common trigger or driver subunits.

The human T-cell leukemia virus type I (HTLV-I) has been etiologically associated with the development of the adult T-cell leukemia (ATL) as well as degenerative neurologic syndrome termed tropical spastic paraparesis (TSP). HTLV-I encodes a potent transactivator protein termed Tax that appears to play an important role in the process of T-cell immortalization. Even though the mechanisms by which Tax induces transformation are still unknown, it seems likely that the ability of Tax to alter the expression of many cellular genes plays an important part in this process. Tax does not bind directly to DNA but rather deregulates the activity of cellular transcription factors. One family of host transcription factors whose activity is altered by Tax includes NF-kappa B/Rel. These transcription factors are post-transcriptionally regulated by their assembly with a second family of inhibitory proteins termed I kappa B that serve to sequester the NF-kappa B/Rel complexes in the cytoplasm. Upon cellular activation, I kappa B alpha is phosphorylated, polyubiquitinated, and degraded in the proteasome. This proteolytic event liberates NF-kappa B, permitting its rapid translocation into the nucleus where it binds to its cognate enhancer elements. Similarly, the p105 precursor of the NF-kappa B p50 subunit is also post-translationally processed in the proteasome. The mechanisms by which Tax activates NF-kappa B remain unclear, and findings presented in the literature are often controversial. We identified a physical interaction between Tax and the HsN3 subunit of the human proteasome. This raises the intriguing possibility that physical association of the HsN3 proteasome subunit with HTLV-I Tax coupled with the independent interaction of Tax with either p100 or p65-I kappa B alpha targets these cytoplasmic NF-kappa B/Rel complexes to the proteasome for processing.

The interleukin (IL)-2 receptor system has previously been shown to signal through the association and tyrosine phosphorylation of Shc. This study demonstrates that the IL-2 receptor beta (IL-2R beta) chain is the critical receptor component required to mediate this effect. The use of IL-2R beta chain deletion mutants transfected into a Ba/F3 murine cell model describes a requirement for the IL-2R beta "acid-rich" domain between amino acids 315 and 384 for Shc tyrosine phosphorylation and receptor association. COS cell co-transfection studies of IL-2R beta chain constructs containing point mutations of tyrosine to phenylalanine along with the tyrosine kinase Jak-1 and a hemagglutinin-tagged Shc revealed that the motif surrounding phosphorylated tyrosine 338 within the acid-rich domain of the IL-2R beta is a binding site for Shc. Deletion of this domain has previously been shown to abrogate the ability of IL-2 to activate Ras but does not affect IL-2-dependent mitogenesis in the presence of serum. Proliferation assays of Ba/F3 cells containing IL-2R beta chain deletion mutants in serum-free medium with or without insulin shows that deletion of the acid-rich domain does not affect IL-2-driven mitogenesis regardless of the culture conditions. This study thus defines the critical domain within the IL-2R beta chain required to mediate Shc binding and Shc tyrosine phosphorylation and further shows that Shc binding and phosphorylation are not required for IL-2-dependent mitogenesis. Neither serum nor insulin is required to supplement the loss of induction of the Shc adapter or Ras pathways, which therefore suggests a novel mechanism for mitogenic signal transduction mediated by this hematopoietin receptor.

Interleukin-2 (IL-2) regulates numerous biological events, including T lymphocyte proliferation. Interleukin-2 receptor (IL-2R)-mediated signaling is triggered by ligand-induced heterodimerization of the IL-2R beta and gamma c subunits, which results in the activation of signaling intermediates that are associated with either IL-2R beta or gamma c. Previous mutagenesis studies of the IL-2R beta cytoplasmic tail demonstrated that the partially conserved box 1 and box 2 motifs and specific tyrosine residues are critical for growth signaling. By deletion and alanine scanning mutagenesis, another set of residues that are critical for IL-2R-mediated signaling has now been identified. These residues lie within the divergent 35-amino acid "spacer" region separating box 1 and box 2. The role of this receptor subregion in early phases of IL-2R signaling was evaluated using BA/F3 stable cell lines expressing three functionally impaired mutants from this region. All three cell lines displayed substantially diminished growth responsiveness to IL-2. Receptor-mediated STAT factor activation, IL-2R beta phosphorylation, and Janus kinase activation were also markedly impaired. These findings indicate that this variable spacer region, which we have termed the V-box, is essential for the initiation of IL-2R-mediated signal transduction.

To evaluate the possible role for receptor-based tyrosine phosphorylation in growth signaling induced by interleukin-2 (IL-2), a series of substitution tyrosine mutants of the IL-2 receptor beta and gamma c chains was prepared and analyzed. Concurrent mutation of all six of the cytoplasmic tyrosines present in the beta chain markedly inhibited IL-2-induced growth signaling in both pro-B and T cell lines. Growth signaling in a pro-B cell line was substantially reconstituted when either of the two distal tyrosines (Tyr-392, Tyr-510) was selectively restored in the tyrosine-negative beta mutant, whereas reconstitution of the proximal tyrosines (Tyr-338, Tyr-355, Tyr-358, Tyr-361) did not restore this signaling function. Furthermore, at least one of the two cytoplasmic tyrosines that is required for beta chain function was found to serve as a phosphate acceptor site upon induction with IL-2. Studies employing a chimeric receptor system revealed that tyrosine residues of the beta chain likewise were important for growth signaling in T cells. In contrast, although the gamma c subunits is a target for tyrosine phosphorylation in vivo, concurrent substitution of all four cytoplasmic tyrosines of this chain produced no significant effect on growth signaling by chimeric IL-2 receptors. However, deletion of either the Box 1, Box 2, or intervening (V-Box) regions of gamma c abrogated receptor function. Therefore, tyrosine residues of beta but not of gamma c appear to play a pivotal role in regulating growth signal transduction through the IL-2 receptor, either by influencing cytoplasmic domain folding or by serving as sites for phosphorylation and subsequent association with signaling intermediates. These findings thus highlight a fundamental difference in the structural requirements for IL-2R beta and gamma c in receptor-mediated signal transduction.

To explore the possible involvement of STAT factors ("signal transducers and activators of transcription") in the interleukin 2 receptor (IL-2R) signaling cascade, murine HT-2 cells expressing chimeric receptors composed of the extracellular domain of the erythropoietin receptor fused to the cytoplasmic domains of the IL-2R beta or -gamma c chains were prepared. Erythropoietin or IL-2 activation of these cells resulted in rapid nuclear expression of a DNA-binding activity that reacted with select STAT response elements. Based on reactivity with specific anti-STAT antibodies, this DNA-binding activity was identified as a murine homologue of STAT-5. Induction of nuclear expression of this STAT-5-like factor was blocked by the addition of herbimycin A, a tyrosine kinase inhibitor, but not by rapamycin, an immunophilin-binding antagonist of IL-2-induced proliferation. The IL-2R beta chain appeared critical for IL-2-induced activation of STAT-5, since a mutant beta chain lacking all cytoplasmic tyrosine residues was incapable of inducing this DNA binding. In contrast, a gamma c mutant lacking all of its cytoplasmic tyrosine residues proved fully competent for the induction of STAT-5. Physical binding of STAT-5 to functionally important tyrosine residues within IL-2R beta was supported by the finding that phosphorylated, but not nonphosphorylated, peptides corresponding to sequences spanning Y392 and Y510 of the IL-2R beta tail specifically inhibited STAT-5 DNA binding.

Recent evidence indicates that the nef gene of human immunodeficiency virus type 1 augments rather than inhibits viral replication in both cell culture and in vivo models. In addition, nef alters various normal cellular processes, including the display of CD4 on the cell surface. However, it remains unknown whether the enhancement of infectivity and the downregulation of CD4 represent linked or independent biologic properties of this single protein. In the present studies, mutational analyses were performed to define structure-function relationships within the Nef protein that mediate these effects. To assess the functional consequences of these mutations, sensitive and reliable assays were developed to quantitate the viral infectivity enhancement and CD4 downregulation functions of Nef. The results indicate that membrane-targeting sequences at the N terminus of Nef are important for both functions of Nef, while certain other conserved regions are dispensable for both functions. A conserved proline-X-X repeat segment in the central core of the protein, which is reminiscent of an SH3-binding domain, is critical for the enhancement of infectivity function but is dispensable for CD4 downregulation. However, the downregulation of CD4 by Nef appears to involve a two-step process requiring the initial dissociation of p56lck from CD4 to permit engagement of the endocytic apparatus by CD4. Together, these findings demonstrate that the infectivity enhancement and CD4 downregulation activities of human immunodeficiency virus type 1 Nef can be dissociated. Thus, these processes may be independent of one another in the viral replication cycle.

The biologic actions of IL-2 are mediated by the IL-2R, a multisubunit receptor complex displayed on the surface of lymphocytes and select other hematopoietic lineages. The IL-2R exhibits multiple affinities for IL-2 that result from the monomeric (alpha), heterodimeric (alpha beta and beta gamma), and heterotrimeric (alpha beta gamma) assembly of different receptor subunits. In the present studies, we have used a series of IL-2R mutants in a transient mammalian expression system to investigate the potential role of intracellular receptor regions in the ligand-binding functions of the IL-2R. Analyses of chimeric and deletion mutants of the IL-2R beta subunit have revealed that its intracellular domain critically and selectively influences high affinity ligand binding mediated through the extracellular domains of the alpha beta-heterodimeric receptor. In contrast, intermediate affinity binding of IL-2 by beta gamma-heterodimeric receptors exhibits no dependence on the cytoplasmic domain of IL-2 R beta. Further, co-expression of either a full-length or severely truncated form of IL-2 R gamma to generate an alpha beta gamma-heterotrimeric complex also overcomes the functional dependence upon the cytoplasmic tail of IL-2 R beta. Collectively, our findings suggest that the cytoplasmic domain of IL-2R beta produces intrasubunit transmembrane conformational changes in this receptor subunit that promote extracellular IL-2 binding in combination with IL-2R alpha. These findings have important implications for the receptor dynamics involved in both ligand binding and signal transduction as well as for clinical applications pertaining to altering IL-2R function.

The nef gene product of human immunodeficiency virus type 1 (HIV-1) promotes more-rapid kinetics of viral replication in primary peripheral blood mononuclear cells. We have previously shown that these enhancing effects of Nef on HIV-1 replication reflect an increase in viral infectivity detectable both in limiting dilution assays and through a single-cycle infection of the HeLa-CD4-long terminal repeat-beta-galactosidase indicator cell line. We now demonstrate that nef-defective HIV-1 can be rescued to near wild-type levels of infectivity by coexpressing Nef in trans in the cell line producing the virus. This observation indicates that HIV-1 virions produced in the presence of Nef are intrinsically different. However, we show that the major viral structural proteins are quantitatively similar in purified viral preparations. We also demonstrate the functional equivalence of the gp120-gp41 envelope glycoprotein complexes of Nef+ and Nef- HIV-1 through an assay for viral entry. Finally, we show that env-defective Nef+ HIV-1 pseudotyped with an amphotropic envelope is also more infectious than similarly pseudotyped Nef- HIV-1. Thus, the production of HIV-1 in the presence of Nef results in viral particles that are more infectious, and this increased infectivity is manifested at a stage after viral entry but prior to or coincident with HIV-1 gene expression.

The tax gene product of human T-cell leukemia virus type I (HTLV-I) is a potent transcriptional activator that both stimulates viral gene expression and activates an array of cellular genes involved in T-cell growth. Tax acts indirectly by inducing or modifying the action of various host transcription factors, including members of the NF-kappa B/Rel family of enhancer-binding proteins. In resting T cells, many of these NF-kappa B/Rel factors are sequestered in the cytoplasm by various ankyrin-rich inhibitory proteins, including I kappa B alpha. HTLV-I Tax expression leads to the constitutive nuclear expression of biologically active NF-kappa B and c-Rel complexes; however, the biochemical mechanism(s) underlying this response remains poorly understood. In this study, we demonstrate that Tax-stimulated nuclear expression of NF-kappa B in both HTLV-I-infected and Tax-transfected human T cells is associated with the phosphorylation and rapid proteolytic degradation of I kappa B alpha. In contrast to prior in vitro studies, at least a fraction of the phosphorylated form of I kappa B alpha remains physically associated with the NF-kappa B complex in vivo but is subject to rapid degradation, thereby promoting the nuclear translocation of the active NF-kappa B complex. We further demonstrate that Tax induction of nuclear c-Rel expression is activated by the RelA (p65) subunit of NF-kappa B, which activates transcription of the c-rel gene through an intrinsic kappa B enhancer element. In normal cells, the subsequent accumulation of nuclear c-Rel acts to inhibit its own continued production, indicating the presence of an autoregulatory loop. However, the pathologic action HTLV-I Tax leads to the deregulated and sustained nuclear expression of both NF-kappa B and c-Rel, a response that may contribute to HTLV-I-induced T-cell transformation.

The gene product of the nef gene of HIV-1 acts both to increase the infectivity of viral particles and to reduce the expression of the CD4 receptor molecule on the cell surface. These two functions of Nef may be related, in that downregulation of CD4 may promote the production of HIV that has greater infectivity.

Binding of interleukin-2 (IL-2) to the IL-2 receptor (IL-2R) triggers a series of intracellular events culminating in lymphocyte proliferation and differentiation. A novel transient assay of signal transduction leading to proliferation is now described which allows the rapid functional assessment of wild type and mutant receptors, including the IL-2R and other members of the cytokine receptor superfamily. This assay has been used to define domains and specific residues within the IL-2R beta intracellular region that contribute to growth signal transduction. In these studies, internal deletion of either the conserved "Box 1" or "Box 2" proximal cytokine receptor homology segments significantly impaired receptor function. Similarly, mutation of specific key residues within or between Box 1 and Box 2, or deletion of the C-terminal 94 residues of the IL-2R beta chain, impaired growth signaling. In contrast, either replacement of the transmembrane domain with that of the CD4 molecule or internal deletion of the 119 amino acids immediately downstream of Box 2 had no impact on growth signaling competence. These studies thus further define the functional architecture of the intracellular region of IL-2R beta, and reveal specific receptor domains that are dispensable, unique, or functionally redundant.

The signal transduction mechanism of the interleukin-2 receptor (IL-2R) remains largely undefined. The cytosolic domain of the IL-2R beta subunit is known to be essential for coupling to intracellular signaling pathways such as protein tyrosine kinase (PTK) and for control of IL-2 dependent cellular proliferation. A panel of cell lines that express IL-2R beta chains that contain sequential truncation mutations within the cytosolic domain were constructed. These cell lines were used to map the interaction of IL-2R with PTK activation, and the linkage of PTK function to activation of the enzyme phosphatidylinositol-3-kinase (Pl3K). The data show that the amino terminal segment of the acidic region (residues 314-350) within IL-2R beta is a critical site for PTK activation, and that activation of Pl3K is linked to IL-2 dependent tyrosine phosphorylation.

RelA (p65) functions as the critical transactivating component of the heterodimeric p50-p65 NF-kappa B complex and contains a high-affinity binding site for its cytoplasmic inhibitor, I kappa B alpha. After cellular activation, I kappa B alpha is rapidly degraded in concert with the induced nuclear translocation of NF-kappa B. The present study demonstrates that tumor necrosis factor alpha-induced degradation of I kappa B alpha in human T cells is preceded by its rapid phosphorylation in vivo. However, these effects on I kappa B alpha result in nuclear mobilization of only a fraction of the entire cytoplasmic pool of RelA. Subsequent studies have revealed that (i) cytoplasmic RelA is stably associated not only with I kappa B alpha but also with other ankyrin motif-rich proteins including the products of the NF-kappa B2 (p100) and NF-kappa B1 (p105) genes; (ii) in contrast to RelA-I kappa B alpha, RelA-p100 cytoplasmic complexes are not dissociated following tumor necrosis factor alpha activation; (iii) p100 functions as a potent inhibitor of RelA-mediated transcription in vivo; (iv) the interaction of RelA and p100 involves the conserved Rel homology domain of both proteins but not the nuclear localization signal of RelA, which is required for I kappa B alpha binding; (v) p100 inhibition of RelA function requires the C-terminal ankyrin motif domain, which mediates cytoplasmic retention of RelA; and (vi) as observed with I kappa B alpha, nuclear RelA stimulates p100 mRNA and protein expression. These findings thus reveal the presence of a second inducible autoregulated inhibitory pathway that helps ensure the rapid but transient action of nuclear NF-kappa B.

Human T-cell leukemia virus type I (HTLV-I) is the etiologic agent of the adult T-cell leukemia, an aggressive and often fatal malignancy of activated human CD4 T cells. HTLV-I encodes an essential 40-kDa protein termed Tax that not only transactivates the long terminal repeat of this retrovirus but also induces an array of cellular genes. Tax-mediated transformation of T cells likely involves the deregulated expression of various cellular genes that normally regulate lymphocyte growth produced by altered activity of various endogenous host transcription factors. In particular, Tax is capable of modulating the expression or activity of various host transcription factors, including members of the NF-kappa B/Rel and CREB/ATF families, as well as the cellular factors HEB-1 and p67SRF. An additional distinguishing characteristic of HTLV-I infection is the profound state of viral latency that is present in circulating primary leukemic T cells. In this study, we demonstrate that HTLV-I Tax can physically associate with p100, the product of the Rel-related NF-kappa B2 gene, both in transfected cells and in HTLV-I-infected leukemic T-cell lines. Furthermore, the physical interaction of Tax with p100 leads to the inhibition of Tax-induced activation of the HTLV-I and human immunodeficiency virus type 1 long terminal repeats, reflecting p100-mediated cytoplasmic sequestration of the normally nuclearly expressed Tax protein. In contrast, a mutant of Tax that selectively fails to activate nuclear NF-kappa B expression does not associate with p100. Together, these results suggest that the cytoplasmic interplay of Tax and p100 may play an important role in the initiation and maintenance of HTLV-1 latency observed in adult T-cell leukemia.

The predominant inducible form of the NF-kappa B transcription factor is a heteromeric complex containing two Rel-related DNA-binding subunits, termed p65 and p50. Prior transfection studies have shown that when these p65 and p50 subunits are expressed independently as stable homodimers, p65 stimulates kappa B-directed transcription, whereas p50 functions as a kappa B-specific repressor. While authentic p50 homodimers (previously termed KBF1) have been detected in nuclear extracts from nontransfected cells, experimental evidence supporting the existence of p65 homodimers in vivo was lacking. We now provide direct biochemical evidence for the presence of an endogenous pool of inducible p65 homodimers in intact human T cells. As with the prototypical NF-kappa B p50-p65 heterodimer, this novel p65 homodimeric form of NF-kappa B is functionally sequestered in the cytoplasm but rapidly appears in the nuclear compartment following cellular stimulation. Site-directed mutagenesis studies indicate that the homodimerization function of p65 is dependent upon the presence of cysteine 216 and a conserved recognition motif for protein kinase A (RRPS; amino acids 273 to 276), both of which reside within a 91-amino-acid segment of the Rel homology domain that mediates self-association. In contrast, mutations at these two sites do not affect heterodimerization of p65 with p50 or its functional interaction with I kappa B alpha. These later findings indicate that neither homo- nor heterodimer formation is an absolute prerequisite for I kappa B alpha recognition of p65. Taken together with prior in vivo transcription studies, these results suggest that the biological activities of p65 and p50 homodimers are independently regulated, thereby providing an integrated and flexible control mechanism for the rapid activation and repression of NF-kappa B/Rel-directed gene expression.

NF-kappa B and AP-1 represent distinct mammalian transcription factors that target unique DNA enhancer elements. The heterodimeric NF-kappa B complex is typically composed of two DNA binding subunits, NF-kappa B p50 and NF-kappa B p65, which share structural homology with the c-rel proto-oncogene product. Similarly, the AP-1 transcription factor complex is comprised of dimers of the c-fos and c-jun proto-oncogene products or of closely related proteins. We now demonstrate that the bZIP regions of c-Fos and c-Jun are capable of physically interacting with NF-kappa B p65 through the Rel homology domain. This complex of NF-kappa B p65 and Jun or Fos exhibits enhanced DNA binding and biological function via both the kappa B and AP-1 response elements including synergistic activation of the 5' long terminal repeat of the human immunodeficiency virus type 1. These findings support a combinatorial mechanism of gene regulation involving the unexpected cross-coupling of two different classes of transcription factors to form novel protein complexes exhibiting potentiated biological activity.

Human T-cell leukemia virus type I (HTLV-I) Rex and human immunodeficiency virus type 1 (HIV-1) Rev are essential gene products required for the replication of these two pathogenic human retroviruses. Both Rex and Rev act at a posttranscriptional level by binding to highly structured RNA-response elements, the Rex-response element in HTLV-I and the Rev-response element in HIV-1. Using a sensitive in vivo assay of protein-protein interaction, we now demonstrate that the HTLV-I Rex and HIV-1 Rev proteins readily form homomultimeric complexes in the absence of their cognate RNA-response elements yet fail to form heteromultimeric complexes with each other. Dominant negative mutations have been identified in both the rex and rev genes which presumably specify a critical activation or effector domain in each of these viral transactivators. Surprisingly, these dominant negative mutants of Rex and Rev fail to interact in vivo. These findings raise the possibility that the binding of nonfunctional monomers rather than functional multimers underlies the transdominant phenotype of these Rex and Rev mutants. Further, it seems likely that the assembly of functional and stable multimers of Rex and Rev in vivo may depend not only on the intrinsic multimerization domains of these proteins but also on the binding of a bridging cellular cofactor to the related activation domains present in each viral transactivator.

The eukaryotic transcription factor nuclear factor-kappa B (NF-kappa B) participates in many parts of the genetic program mediating T lymphocyte activation and growth. Nuclear expression of NF-kappa B occurs after its induced dissociation from its cytoplasmic inhibitor I kappa B alpha. Phorbol ester and tumor necrosis factor-alpha induction of nuclear NF-kappa B is associated with both the degradation of performed I kappa B alpha and the activation of I kappa B alpha gene expression. Transfection studies indicate that the I kappa B alpha gene is specifically induced by the 65-kilodalton transactivating subunit of NF-kappa B. Association of the newly synthesized I kappa B alpha with p65 restores intracellular inhibition of NF-kappa B DNA binding activity and prolongs the survival of this labile inhibitor. Together, these results show that NF-kappa B controls the expression of I kappa B alpha by means of an inducible autoregulatory pathway.

The biological activity of interleukin-2 receptors (IL-2R) is dependent on the functional coupling of IL-2R beta molecules to intracellular enzymes such as protein tyrosine kinase. The serine-rich, S-domain within the cytosolic portion of IL-2R beta plays an essential role in transduction of the IL-2 proliferative signal. Cells bearing either wild type IL-2R beta (Baf alpha/beta) or deletions within the S-domain (Baf alpha/beta SD1) were used to evaluate the importance of the S-domain in linking ligand binding to protein tyrosine kinase induction. Multiparameter, side by side comparisons showed that only those cells that expressed wild type IL-2R beta responded to IL-2 by increased cellular proliferation, accumulation of the c-myc proto-oncogene, and activation of protein tyrosine kinase. Activation of protein tyrosine kinase was, in turn, linked to increased tyrosine phosphorylation and activation of phosphatidylinositol-3-kinase. These findings indicate that the S-domain of the IL-2R beta chain is an essential component in the signal transduction cascade that links IL-2 binding to tyrosine kinase and phosphatidylinositol-3-kinase activation.

The Rex protein of the type I human T-cell leukemia virus (HTLV-I) is essential for viral replication, acting post-transcriptionally to enhance the expression of unspliced and singly spliced viral mRNAs that encode the Gag, Pol, and Env virion proteins. Rex function involves its direct interaction with a complex stem-loop structure termed the Rex RNA response element (RexRE), which is located within the 3' retroviral long terminal repeat. Binding of Rex to the RexRE involves a positively charged arginine-rich domain located near the N-terminus which also functions as a nuclear localization signal. Strikingly, substitution of all seven of the arginine residues present within this domain with positively charged lysine residues exerted no adverse effect on the nuclear targeting of Rex. However, these lysine substitutions completely abrogated both Rex binding to the RexRE and Rex function. Reversion of multiple substituted lysines to arginines at specific locations within this domain was required to restore both RexRE binding and biological function to the Rex protein. Thus, while the presence of positive charge alone in this domain appears sufficient for nuclear localization of Rex, multiple arginine residues at specific sites are essential for the full expression of RNA binding and functional activity of this retroviral trans-regulatory protein.

The long terminal repeat (LTR) of the type 1 human immunodeficiency virus (HIV-1) and the 5' regulatory region of the gene encoding the interleukin 2 receptor alpha subunit (IL-2R alpha) share functional kappa B enhancer elements involved in the regulation of these inducible transcription units during T-cell activation. These kappa B enhancer elements are recognized by a structurally related family of interactive proteins that includes p50, p65, and the product of the c-rel protooncogene (c-Rel). Recent biochemical studies have shown that p65 and p50 form the prototypical NF-kappa B complex, which is rapidly translocated from the cytoplasm to the nucleus during T-cell activation. This intracellular signaling complex potently stimulates kappa B-directed transcription from either the HIV-1 LTR or the IL-2R alpha promoter via the strong transactivation domain present in p65. We now demonstrate that nuclear expression of human c-Rel, which is induced by either phorbol ester or tumor necrosis factor alpha with delayed kinetics relative to p65, markedly represses p65-mediated activation of these transcription units. These inhibitory effects of c-Rel correlate with its DNA-binding activity but not with its ability to heterodimerize with p50, suggesting that c-Rel inhibition involves competition with p50/p65 for occupancy of the kappa B enhancer element. Together, these findings suggest that one function of c-Rel is as a physiologic repressor of the HIV-1 LTR and IL-2R alpha promoters, serving to efficiently counter the strong transcriptional activating effects of p65.

The active nuclear form of the NF-kappa B transcription factor complex is composed of two DNA binding subunits, NF-kappa B p65 and NF-kappa B p50, both of which share extensive N-terminal sequence homology with the v-rel oncogene product. The NF-kappa B p65 subunit provides the transactivation activity in this complex and serves as an intracellular receptor for a cytoplasmic inhibitor of NF-kappa B, termed I kappa B. In contrast, NF-kappa B p50 alone fails to stimulate kappa B-directed transcription, and based on prior in vitro studies, is not directly regulated by I kappa B. To investigate the molecular basis for the critical regulatory interaction between NF-kappa B and I kappa B/MAD-3, a series of human NF-kappa B p65 mutants was identified that functionally segregated DNA binding, I kappa B-mediated inhibition, and I kappa B-induced nuclear exclusion of this transcription factor. Results from in vivo expression studies performed with these NF-kappa B p65 mutants revealed the following: 1) I kappa B/MAD-3 completely inhibits NF-kappa B p65-dependent transcriptional activation mediated through the human immunodeficiency virus type 1 kappa B enhancer in human T lymphocytes, 2) the binding of I kappa B/MAD-3 to NF-kappa B p65 is sufficient to retarget NF-kappa B p65 from the nucleus to the cytoplasm, 3) selective deletion of the functional nuclear localization signal present in the Rel homology domain of NF-kappa B p65 disrupts its ability to engage I kappa B/MAD-3, and 4) the unique C-terminus of NF-kappa B p65 attenuates its own nuclear localization and contains sequences that are required for I kappa B-mediated inhibition of NF-kappa B p65 DNA binding activity. Together, these findings suggest that the nuclear localization signal and transactivation domain of NF-kappa B p65 constitute a bipartite system that is critically involved in the inhibitory function of I kappa B/MAD-3. Unexpectedly, our in vivo studies also demonstrate that I kappa B/MAD-3 binds directly to NF-kappa B p50. This interaction is functional as it leads to retargeting of NF-kappa B p50 from the nucleus to the cytoplasm. However, no loss of DNA binding activity is observed, presumably reflecting the unique C-terminal domain that is distinct from that present in NF-kappa B p65.

IL-2 is one of the principal growth factors regulating the proliferation of T lymphocytes. Although two independent IL-2-binding molecules have been molecularly cloned and shown to participate in the formation of a high affinity receptor complex, their primary structures do not suggest a specific mechanism for IL-2 growth signal transduction across the cell membrane. Neither IL-2 receptor subunit contains an intrinsic kinase domain; nevertheless, tyrosine phosphorylation of various intracellular substrates is one of the first biochemical changes observed following activation of the IL-2 receptor (IL-2R). Both serine/threonine and tyrosine kinases can be co-precipitated as part of the IL-2R complex suggesting that the IL-2 signalling may involve the activation of non-covalently associated intracellular kinases. However, controversy exists as to which kinases are involved in IL-2 signal transduction; in particular, which kinase(s) mediates the first or proximal event(s) in the signalling process. Activation of the IL-2R leads to serine and threonine phosphorylation of the SRC tyrosine kinase family member, LCK, and an increase in LCK tyrosine kinase activity. Furthermore, LCK can be co-immunoprecipitated with the beta chain of the IL-2R indicating its association with the receptor complex. IL-2 has also been reported to increase FYN kinase activity and to alter its association with the 85 kDa subunit of phosphatidylinositol-3 kinase thus suggesting a role for FYN in IL-2 signal transduction. However, in this report, we now demonstrate that neither LCK nor FYN are obligatory for IL-2-induced growth of HTLV-I-infected human T cells. Lack of expression of LCK or FYN in the HTLV-I-infected T cell lines was demonstrated by a combination of Northern blotting, polymerase chain reaction, Western blotting, and in vitro kinase activity. Despite the absence of LCK or FYN, IL-2 induced similar patterns of rapid tyrosine phosphorylation. Similar results were observed in cell lines lacking expression of the LYN, FGR, HCK, and LTK tyrosine kinases. Thus, none of these tyrosine kinases alone appears to be required for growth signalling through the IL-2R in the HTLV-I-infected T cell lines analyzed. The findings raise the possibility that an, as yet, unidentified tyrosine kinase is involved. Alternatively, this biological signalling system may exhibit remarkable redundancy whereby several different tyrosine kinases may be capable of associating with the IL-2R complex and mediating intracellular signalling.

Infection with human immunodeficiency virus type 1 leads to a persistent but progressive cytopathic process that culminates in the near complete destruction of the CD4+ subset of T cells. The levels of human immunodeficiency virus type 1 replication and virus burden increase throughout the clinical course of disease reflecting a balance between the viral and cellular regulatory influences as well as the ability of the host immune system to eliminate infected T cells. Human immunodeficiency virus type 1 replication is dependent on the state of cellular activation and involves both inducible host cell derived transcription factors and at least three virus-derived gene products. Further study of the mechanism of action of these factors, particularly those encoded by the virus, may facilitate the future development of highly specific and effective therapies for human immunodeficiency virus type 1.

The v-rel oncogene product from the avian reticuloendotheliosis virus strain T corresponds to a member of the Rel-related family of enhancer-binding proteins that includes both the mammalian 50- and 65-kDa subunits of the NF-kappa B transcription factor complex. However, in contrast to NF-kappa B, v-Rel has been shown to function as a dominant-negative repressor of kappa B-dependent transcription in many mature cell types. We now demonstrate that a highly conserved motif within the Rel homology domain of v-Rel containing a consensus protein kinase A phosphorylation site is required for DNA binding, transcriptional repression, and cellular transformation mediated by this oncoprotein. However, replacement of the serine phosphate acceptor within the protein kinase A site with an alanine did not alter any of these functions of v-Rel, suggesting that phosphorylation at this site is not central to the regulation of this oncogene product. Rather, the inactive mutations appear to identify a functional domain within v-Rel required for these various biological activities. It is notable that these same mutations do not impair the ability of v-Rel to heterodimerize with the 50-kDa subunit of NF-kappa B, suggesting that v-Rel-mediated transcriptional repression likely involves direct nuclear blockade of the kappa B enhancer rather than indirect alterations in the composition of preformed cytoplasmic NF-kappa B complexes. Paradoxically, when introduced into undifferentiated F9 cells, v-Rel functions as a kappa B-specific transcriptional activator rather than as a dominant-negative repressor. These stimulatory effects of v-Rel require both the conserved protein kinase A phosphorylation site and additional unique C-terminal sequences not needed for v-Rel-mediated repression in mature cells. Retinoic acid-induced differentiation of these F9 cells restores the repressor function of v-Rel. These opposing biological actions of v-Rel occurring in cells at distinct stages of differentiation may have important implications for the mechanism of v-Rel-mediated transformation occurring in avian splenocytes.

Using electrophoretic mobility shift assays and DNA-protein cross-linking techniques, we demonstrate that IL-2 binding to functional forms of the human IL-2R activates nuclear expression of the eukaryotic transcription factor NF-kappa B. These inductive effects of IL-2 were observed in three different cellular systems including human Jurkat T cells stably transfected with IL-2R beta cDNA, mouse pro-B BA/F3 cells stably expressing human IL-2R beta chains either alone or in combination with human or murine IL-2R alpha chains, and purified primary resting human T cells constitutively displaying small numbers of IL-2R beta molecules. IL-2 activation of nuclear NF-kappa B expression is regulated in part at a post-translational level, involving the rapid translocation of both the 50- and 65-kDa subunits of NF-kappa B from the cytoplasm to the nucleus. However, IL-2 induction also produced an increase in mRNA for NF-kappa B p105, indicating an additional pretranslational component of regulation. In contrast, IL-2 exerts only minor effects on NF-kappa B p65 mRNA expression. IL-2 induction of NF-kappa B through the IL-2R beta subunit was both correlated with activation of the endogenous IL-2R alpha gene and critically dependent upon the presence of a serine-rich cytoplasmic domain within IL-2R beta (amino acid residues 267-312). This domain has previously been shown to be essential for IL-2-induced cell growth and may correspond to a binding site for an IL-2R beta-associated tyrosine kinase. Together, these findings suggest that NF-kappa B may play an important role as one intracellular second messenger relaying signals from the plasma membrane to cell nucleus that leads to IL-2-induced activation and growth.

The 27-kDa Rex trans-acting protein appears to be essential for replication of human T-cell leukemia virus type I. Mutations introduced outside of the Rex RNA-binding domain-nucleolar localization signal display either wild-type activity or, conversely, yield dominant-negative proteins. We generated missense mutations in a particular domain of the Rex protein (amino acid residues 54 to 69) which is characterized by a cluster of dominant-negative mutants. Our results indicate that amino acids 57 to 67 are critically important for Rex function mediated through the RxRE cis-acting RNA sequence. Within this domain, only amino acids 61 to 63 could be mutated without loss of function. All other missense and deletion mutants yielded dominant-negative proteins. In vitro RNA-binding studies performed with glutathione S-transferase-Rex fusion proteins demonstrated that all of the mutant Rex proteins interacted specifically with RxRE RNA. Analysis of chimeric Rex-Rev proteins suggests that this Rex domain is important for oligomerization.

The nuclear factor (NF)-kappa B transcription factor system is composed of at least four inducible nucleoprotein adducts termed p50, p55 (NF-kappa B p50), p75 (NF-kappa B p65), and p85 (c-Rel). These proteins are expressed in the nuclei of activated T cells in a distinctly biphasic fashion, with p55 and p75 induction occurring within minutes whereas the induction of p50 and p85 occurs after several hours. In contrast, p50 and p55 are constitutively expressed in the nuclei of U937 and THP-1 monocytic cells. However, cellular activation is required for the nuclear expression of p75 in these cells. Additionally, activation of monocytic cells does not result in a significant induction of p85. Tumor necrosis factor alpha induces the nuclear expression of p55 and p75 in these monocytic cells within 20 min, presumably reflecting the liberation of these proteins from I kappa B. In contrast, phorbol myristate acetate (PMA) induces the expression of these proteins with delayed kinetics, raising the possibility that PMA is incapable of mediating the efficient release of p55 and p75 from I kappa B in these cells. These findings highlight important differences in the regulation of these proteins in monocytic cells versus T cells and suggest that the induced expression of NF-kappa B p65 in monocytes may play a central role in the activation of HIV-1 gene expression.

Molecular cloning of the polypeptide component of the Rel-related human p75 nucleoprotein complex has revealed its identity with the 65-kDa (p65) subunit of NF-kappa B. Functional analyses of chimeric proteins composed of NF-kappa B p65 C-terminal sequences linked to the DNA-binding domain of the yeast GAL4 polypeptide have indicated that the final 101 amino acids of NF-kappa B p65 comprise a potent transcriptional activation domain. Transient transfection of human T cells with an expression vector encoding NF-kappa B p65, but not NF-kappa B p50, produced marked transcriptional activation of a basal promoter containing duplicated kappa B enhancer motifs from the long terminal repeat of type 1 human immunodeficiency virus. These stimulatory effects of NF-kappa B p65 were synergistically enhanced by coexpression of NF-kappa B p50 but were completely inhibited by coexpression of the v-rel oncogene product. Together, these functional studies demonstrate that NF-kappa B p65 is a transactivating subunit of the heterodimeric NF-kappa B complex and serves as one cellular target for v-Rel-mediated transcriptional repression.

The Tax trans-activator protein of the type I human T-cell leukemia virus is expressed predominantly in the nuclei of cells. However, this viral trans-activator is distinguished from most other nuclear proteins by the absence of a short highly basic nuclear localization signal. Previous mutational analyses of the tax gene revealed that many of the missense mutations involving the amino terminus of Tax resulted in a predominantly cytoplasmic pattern of expression. We now report that the amino terminal 48 residues of Tax comprise a functional nuclear localization signal as demonstrated by the ability of this region to retarget expression of a large cytoplasmic protein to the nucleus.

The Tax protein of the human T-cell leukemia virus type I (HTLV-I) serves as a potent transcriptional activator of its own long terminal repeat as well as select cellular genes, including interleukin-2 and the alpha subunit of the interleukin-2 receptor. Tax activation of these two growth-related genes appears to involve the induced nuclear expression of DNA-binding proteins that specifically engage related kappa B enhancer elements present in the 5' regulatory regions of these genes. In human T cells, kappa B enhancer-binding activity has been discerned as an unexpectedly large family of UV cross-linked nucleoprotein adducts, termed p50, p55, p75, and p85. The protein components of each of these DNA-protein adducts have been shown to share structural similarity with the v-rel oncogene product. The p55 adduct is composed of the 50-kDa subunit of NF-kappa B derived from a 105-kDa precursor polypeptide, while the p50 adduct contains a smaller protein that is closely related to NF-kappa B p50. The p75 adduct contains the 65-kDa subunit of NF-kappa B, while the p85 adduct is composed of the human c-rel proto-oncogene product. We now demonstrate that HTLV-I Tax, in the absence of other viral pX gene products, is capable of inducing the nuclear expression of all four of these kappa B-binding proteins in human T cells, with most marked effects involving c-Rel and NF-kappa B p65. Tax induction of the nuclear expression of c-Rel and NF-kappa B p50 is regulated, at least in part, at a pretranslational level involving increases in c-rel and NF-kappa B p105 mRNA expression. To study the pattern of expression of these kappa B-specific proteins in cells infected with the whole HTLV-I, seven cloned HTLV-I-infected T-cell lines were established from the peripheral blood of patients with adult T-cell leukemia. Of note, only three of these seven cell lines produced Tax, and c-rel mRNA and nuclear protein expression was confined to these three cell lines. In contrast, NF-kappa B p50 and NF-kappa B p65 were constitutively expressed in the nuclei of all seven of the HTLV-I-infected cell lines, even in the absence of detectable Tax or other viral gene expression. These findings raise the possibility of an alternate, Tax-independent pathway for the induced nuclear expression of NF-kappa B p50 and NF-kappa B p65 following HTLV-I infection.

Kappa B (kappa B) enhancer binding proteins isolated from the nuclei of activated human T cells produce two distinct nucleoprotein complexes when incubated with the kappa B element from the interleukin-2 receptor-alpha (IL-2R alpha) gene. These two DNA-protein complexes are composed of at least four host proteins (p50, p55, p75, p85), each of which shares structural similarity with the v-rel oncogene product. Nuclear expression of these proteins is induced with distinctly biphasic kinetics following phorbol ester activation of T cells (p55/p75 early and p50/p85 late). DNA-protein crosslinking studies have revealed that the more rapidly migrating B2 complex contains both p50 and p55 while the more slowly migrating B1 complex is composed of p50, p55, p75, and p85. Site-directed mutagenesis of the wild-type IL-2R alpha kappa B enhancer (GGGGAATCTCCC) has revealed that the binding of p50 and p55 (B2 complex) is particularly sensitive to alteration of the 5' triplet of deoxyguanosine residues. In contrast, formation of the B1 complex, reflecting the binding of p75 and p85, critically depends upon the more 3' sequences of this enhancer element. DNA binding by all four of these Rel-related factors is blocked by selective chemical modification of lysine and arginine residues, suggesting that both of these basic amino acids are required for binding to the kappa B element. Similarly, covalent modification of free sulfhydryl groups with diamide (reversible) or N-ethylmaleimide (irreversible) results in a complete loss of DNA binding activity. In contrast, mild oxidation with glucose oxidase selectively inhibits p75 and p85 binding while not blocking p50 and p55 interactions. These findings suggest that reduced cysteine thiols play an important role in the DNA binding activity of this family of Rel-related transcription factors.

The Tax oncoprotein of the type I human T cell leukemia virus (HTLV-I) activates transcription of cellular and viral genes through at least two different transcription factor pathways. Tax activates transcription of the c-fos proto-oncogene by a mechanism that appears to involve members of the cAMP response element binding protein (CREB) and activating transcription factor (ATF) family of DNA-binding proteins. Tax also induces the nuclear expression of the NF-kappa B family of rel oncogene-related enhancer-binding proteins. We have investigated the potential role of these CREB/ATF and NF-kappa B/Rel transcription factors in Tax-mediated transformation by analyzing the oncogenic potential of Tax mutants that functionally segregate these two pathways of transactivation. Rat fibroblasts (Rat2) stably expressing either the wild-type Tax protein or a Tax mutant selectively deficient in the ability to induce NF-kappa B/Rel demonstrated marked changes in morphology and growth characteristics including the ability to form tumors in athymic mice. In contrast, Rat2 cells stably expressing a Tax mutant selectively deficient in the ability to activate transcription through CREB/ATF demonstrated no detectable changes in morphology or growth characteristics. These results suggest that transcriptional activation through the CREB/ATF pathway may play an important role in Tax-mediated cellular transformation.

The Rex protein of the type I human T-cell leukemia virus (HTLV-I) is essential for the replication of this pathogenic retrovirus and, surprisingly, can also replace the function of the structurally distinct Rev protein of the type 1 human immunodeficiency virus (HIV-1). Rex action requires a 255-nucleotide viral RNA stem-loop structure termed the Rex RNA response element (RexRE) located in the 3' retroviral long terminal repeat. Rex function leads to the induced cytoplasmic expression of the incompletely spliced family of viral mRNAs that uniquely encode the HTLV-I structural and enzymatic proteins (Gag, Pol, and Env). Our studies now demonstrate that Rex acts by binding directly to the RexRE in a sequence-specific manner. These effects of Rex require the presence of a 10-nucleotide subregion of the RexRE that is essential for Rex function in vivo. Dominant-negative mutants of Rex also bind to the RexRE with high affinity, while a recessive-negative Rex mutant altered within its arginine-rich, positively charged domain fails to engage the RexRE. Analogously, both the wild-type and dominant-negative Rex proteins specifically bind to the structurally distinct HIV-1 Rev response element, a finding that likely underlies the respective stimulatory and inhibitory effects of these HTLV-I proteins in the heterologous HIV-1 system. However, consistent with their lack of amino acid homology, the binding sites for Rex and Rev within the HIV-1 Rev response element are distinct.

The IL-2 and the IL-2-R alpha genes are both expressed transiently in normal T lymphocytes after Ag or mitogen activation. In contrast, the human T cell line, IARC 301, expresses these two genes constitutively and we have previously demonstrated that its growth depends on the autocrine production of this T cell growth factor and high affinity IL-2R. To dissect the molecular basis for the unusual persistent expression of the IL-2 and IL-2-R alpha genes in these IARC 301 T cells, we have analyzed the interactions of constitutively expressed nuclear proteins with the 5' flanking regions of the IL-2 and IL-2-R alpha genes using both DNase I footprinting and gel retardation techniques. We have found that a region in both genes (-276 to -250 for IL-2-R alpha and -203 to -183 for IL-2), which corresponds to a kappa B enhancer element, is specifically protected by nuclear proteins from IARC 301. In agreement with this finding, both the IL-2 and IL-2-R alpha promoters are active in transient transfection assays in IARC 301 cells. In contrast, mutation of the kappa B enhancer results in markedly attenuated activities of both promoters. Two proteins binding the kappa B sequence, NF-kappa B and KBF1, are constitutively expressed in IARC 301 nuclei and induced by PMA and PHA in Jurkat. They bind to the kappa B motifs with different relative affinities that may reflect their different contribution in the expression of various promoters.

HIVEN86A is an inducible member of a set of cellular proteins that specifically bind to the kappa B enhancer (Franza et al., 1987; Franza, 1988; Franza, 1990; Ballard et al., 1989; Bohnlein et al., 1988). This enhancer motif has been detected in numerous cellular and viral transcription control domains (Boshart et al., 1985; Sen & Baltimore, 1986; Nabel & Baltimore, 1987). Recently, cDNAs have been cloned (Kieran et al., 1990; Baldwin & Sharp, 1987) that encode the 50 kD DNA binding subunit of murine NF-kappa B (for review: Leonardo & Baltimore, 1989) and the closely related human kappa binding factor (KBF-1) (Kimura et al., 1986; Baldwin & Sharp, 1987). A 350 amino acid domain at the N-terminus of these proteins was found to be homologous with the v-rel oncogene from the avian reticuloendotheliosis virus, strain T (REV-T), as well as a maternal effect gene, dorsal (Kieran et al., 1990; Ghosh et al., 1990). Dorsal is known to activate transcription of certain Drosophila genes (Rushlow et al., 1987). The v-Rel oncoprotein has been identified as a transcriptional activator (Gelinas & Temin, 1988; Hannink & Temin, 1989; Bull et al., 1990) in certain assay systems and shown to be induced by the tumor promoter, phorbol 12-myristate 13-acetate (PMA) in avian cells (for review: Rice & Gilden, 1988). HIVEN86A is also inducible by PMA (Franza et al., 1987; Franza, 1988; Franza, 1990). We now demonstrate that the protein product of the human c-rel proto-oncogene is structurally identical to HIVEN86A.

HTLV-I structural gene expression is posttranscriptionally regulated by the Rex protein and the Rex response element (RexRE), a 255 nucleotide RNA stem-loop structure located in the retroviral 3' LTR. Independent of Rex, the RexRE also plays a critical role in the polyadenylation of all HTLV-I transcripts. Folding of the RexRE serves to spatially juxtapose the widely separated AAUAAA hexamer and GU-rich elements that are essential for polyadenylation. In turn, this folding promotes the cooperative and stable binding of two nuclear factors at these elements that commits this poly(A) site to 3' processing. These findings highlight a novel mechanism of 3' end formation in the HTLV family of retroviruses and underscore the general requirement for protein-protein interactions in the polyadenylation reaction.

The Rex protein of the human T-cell leukemia virus type II (HTLV-II), Rex-II, plays a central role in regulating the expression of the structural genes of this retrovirus. Rex-II acts posttranscriptionally by inducing the cytoplasmic expression of the incompletely spliced viral mRNAs that encode the Gag and Env structural proteins and the enzymes derived from the pol gene. We now define a 295-nucleotide cis-acting regulatory element within the 3' long terminal repeat of HTLV-II that is required for the effects of Rex-II. This Rex-II response element (RexIIRE) corresponds to a predicted, highly stable RNA secondary structure and functions when present in the sense but not in the antisense orientation. The RexIIRE confers responsiveness not only to Rex-II but also to the Rex protein of HTLV-I. Deletion and substitution mutagenesis of the RexIIRE permitted identification of a small subregion within the larger element critically required for Rex-II responsiveness and further suggested that the structurally distinct RexIIREs generated from the 5' and 3' long terminal repeats of HTLV-II may differentially regulate the cytoplasmic expression of unspliced gag-pol and singly spliced env mRNAs. While the Rev protein of human immunodeficiency virus type 1 fails to function via the RexIIRE, the Rex-II protein, like Rex-I, can functionally replace the Rev protein of human immunodeficiency virus type 1 via its interaction with the Rev response element (RevRE).