A phase I study of menogaril in patients with advanced cancer.

Menogaril (7-con-O-methylnogarol) is a semisynthetic anthracycline analogue of nogalamycin that has shown good activity against a variety of experimental tumor systems as well as decreased cardiac toxicity when compared with doxorubicin in preclinical studies. Forty-one patients with refractory solid tumors received menogaril during a phase I trial at The Johns Hopkins Oncology Center (Baltimore). Menogaril was administered as an intravenous (IV) infusion on days 1 and 8 of a 28-day cycle in doses of 8 to 140 mg/m2. Eastern Cooperative Oncology Group (ECOG) grade 3 and 4 leukopenia was the principle dose-limiting toxicity and was occasionally accompanied by thrombocytopenia. Both WBC and platelet nadirs occurred between days 15 and 22. Anemia requiring transfusion was occasionally seen. Nonhematologic toxicities observed included frequent anorexia and malaise that was not dose related and postinfusion phlebitis that was dose related and occasionally dose limiting. Gastrointestinal toxicity and alopecia were infrequent and mild in severity. Three patients with cumulative doses of menogaril greater than 1,400 mg/m2 had no significant changes in ejection fractions as determined by serial gated blood pool scans. Two patients had greater than 10% decrements in ejection fractions without clinical changes at total doses of 128 and 288 mg/m2. One patient with prior anthracycline therapy and chest irradiation decreased her left ventricular ejection fraction from 52% to 30% and developed respiratory failure after two cycles of therapy in the setting of disease progression. No responses to menogaril therapy were observed. The recommended phase II dose for menogaril on this day 1 and 8 schedule is 140 mg/m2. A starting dose of 90 mg/m2 should be considered for heavily pretreated patients. In comparing results of this phase I schedule with those of other schedules, evidence for schedule-dependent toxicity differences should be sought.

Isolation and characterization of the human cyclin T1 promoter.

Cyclin T1 (CycT1) is a regulatory subunit of a general RNA polymerase II (RNAP II) elongation factor termed P-TEFb. The human immunodeficiency virus Tat protein directly associates with CycT1 to utilize CycT1/P-TEFb (also called TAK) for activation of RNAP II elongation of the integrated proviral genome. CycT1 mRNA and protein levels are induced in activated human peripheral blood lymphocytes and CycT1 protein levels are induced by a post-transcriptional mechanism when human U937 promonocytic cells are stimulated to differentiate into macrophage-like cells. To investigate mechanisms that regulate CycT1 RNA expression, we isolated the CycT1 promoter. Multiple transcription start sites were identified within 330 nucleotides upstream of the ATG initiation codon at +1. The CycT1 promoter lacks a TATA element and possesses high constitutive activity in plasmid transfection assays. Two distinct regions of the promoter were identified upstream of +1 that contain critical regulatory elements for CycT1 promoter function.

Genomic organization and characterization of promoter function of the human CDK9 gene.

CDK9 is the catalytic subunit of a general RNA polymerase II (RNAP II) elongation factor termed p-TEFb which is targeted by the human immunodeficiency virus (HIV) Tat protein to activate elongation of the integrated proviral genome. CDK9 mRNA and protein levels have been observed to be induced in activated peripheral blood lymphocytes, a cell type relevant to HIV infection. To investigate mechanisms that regulate CDK9 RNA expression, we isolated genomic sequences containing the human CDK9 gene and found that CDK9 coding sequences are interrupted by six introns. There is a major transcriptional start site located 79 nucleotides upstream of the ATG initiator codon at nucleotide +1. Nucleotides -352 to -1 contain all the transcriptional regulatory elements needed for full promoter activity in transient expression assays. The CDK9 promoter contains features characteristic of a housekeeping gene, including GC-rich sequences and absence of a functional TATA element. The CDK9 promoter possesses high constitutive activity and may therefore have utility in expression vectors or gene therapy vectors.

Mutational analysis of the amino and carboxy termini of the HIV-2 Tat protein.

The transactivator proteins of HIV-1 and HIV-2, Tat-1 and Tat-2, are highly homologous in the center of each molecule but are divergent in the amino and carboxy termini. The structure of Tat-1 has been extensively characterized by mutagenesis studies, whereas little is as yet known specifically about the structure of Tat-2. To characterize the Tat-2 protein, we performed a mutational analysis of the amino and carboxy termini of the fully functional first exon (99 residues) of the Tat-2 protein. We found that deletion of residues 8 through 33 in the amino terminus drastically reduced transactivation activity, whereas deletion of residues 8 through 47 largely abolished transactivation activity. We also analyzed chimeric proteins in which the amino termini of the Tat-1 and Tat-2 proteins were exchanged precisely at the first cysteine in the cysteine-rich regions. Both chimeric proteins possessed very low levels of transactivation activity, indicating that the amino termini of Tat-1 and Tat-2 are not interchangeable. Truncation mutants in the carboxy terminus were analyzed and amino acid 90 at the end of the basic domain was found to be at or near the limit of carboxy residues that can be deleted without abolishing Tat-2 function. A Tat-2 mutant truncated after residue 84 within the basic domain was found to be a transdominant mutant able to inhibit wild-type Tat-1 and wild-type Tat-2 activities. Additionally, the results of immunoprecipitations suggested that deletions in the Tat-2 amino terminus can reduce protein stability.

Limited proteolytic digestions identify common structural features of HIV-1 Tat proteins expressed during infection from alternatively spliced mRNAs.

Although multiple forms of the HIV-1 Tat protein are synthesized during infection from alternatively spliced mRNAs, only 72 amino acid residues encoded in the first Tat exon are necessary for full transactivation activity. We used limited proteolytic digestions of proteins expressed in vitro to study the structures of three Tat proteins from isolate HXB2: 72R Tat (first Tat exon); 86R Tat (first and second Tat exons), Tev or TNV (first Tat exon plus Env and Rev exons). For the 86R and Tev proteins, either trypsin or chymotrypsin cleaved the majority of carboxyl residues from the first exon. Moreover, when released from carboxyl residues, the first exon of 86R and Tev was relatively resistant to subsequent proteolysis. The entire 72R Tat protein was relatively resistant to proteolysis. The protease-resistant first exon in all Tat proteins was abolished by EDTA treatment, suggesting that divalent cations are required for its conformation. Our results suggest that the first exon in the 86R, 72R, and Tev proteins is folded into a similar structure which, as defined by partial proteolysis, acts as a single biochemical domain.

Wild-type and mutant HIV-1 and HIV-2 Tat proteins expressed in Escherichia coli as fusions with glutathione S-transferase.

Human immunodeficiency virus type 1 (HIV-1) and HIV-2 encode related transcriptional activators known as Tat-1 and Tat-2, respectively, that are required for efficient viral replication. The Tat proteins have been studied extensively, and it appears that their mechanism of action is unique to the primate immunodeficiency viruses or a few distantly related lentiviruses. Here we describe a collection of 24 wild-type and mutant Tat-1 and Tat-2 proteins that are expressed in Escherichia coli as fusions with glutathione S-transferase (GST). The GST-Tat fusions can be used for biochemical studies after simple purification from E. coli lysates in a single step under nondenaturing conditions. The availability of these GST-Tat fusions should be useful to investigators examining biochemical properties of Tat-1 and Tat-2 proteins. E. coli cultures harboring GST-Tat fusions described here are available through the National Institute of Health AIDS Research and Reference Reagent Program.

Construction and characterization of a potent HIV-2 Tat transdominant mutant protein.

The human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) Tat proteins Tat-1 and Tat-2 stimulate transcription of the viral long terminal repeat (LTR) sequences and are required for efficient viral replication. A class of mutant Tat proteins, termed "transdominant mutants," has been described that possesses relatively low transactivation activity, yet is able to inhibit the function of wild-type Tat. These mutant proteins contain a nonfunctional TAR RNA-binding domain but apparently retain a functional activation domain. A potential limitation for therapeutic use of transdominant mutants described to date is their low but significant basal level of transactivation for the HIV-1 or HIV-2 LTRs. In order to make an improved transdominant mutant, we have constructed Tat-2 proteins that contain mutations in four contiguous arginines at residues 81 to 84 in the RNA-binding domain. Using purified proteins and in vitro RNA-binding assays, we verified that these mutant Tat-2 proteins are defective for TAR RNA binding. We also verified that these mutant Tat-2 proteins bind to a cellular protein kinase in vitro that we have previously shown to bind specifically to the Tat-1 and Tat-2 activation domain. Using plasmid cotransfection assays, we compared the phenotypes of these mutant Tat-2 proteins with the most potent Tat-1 transdominant mutant described to date. One Tat-2 mutant, named "R81-84A," was found to be equivalent to the Tat-1 mutant in ability to inhibit wild-type Tat transactivation of HIV-1 and HIV-2 LTRs. Moreover, the R81-84A mutant possessed a significantly lower basal level of transactivation than the Tat-1 mutant. The R81-84A Tat-2 mutant is therefore a promising reagent for future development as an anti-HIV agent. Additionally, our results suggest that wild-type Tat-2 transactivation of the HIV-2 LTR is especially sensitive to inhibition by transdominant mutants.

RNAs Selected in vitro by the HIV-2 Tat Protein.

The Tat proteins of human immunodeficiency virus types 1 (HIV-1) and 2 (HIV-2), termed Tat-1 and Tat-2, respectively, are essential for efficient viral replication. Tat proteins activate viral transcription by binding to the TAR RNA stem-loop structure at the 5' end of viral transcripts. We used an in vitro selection procedure to identify RNAs present in a large sequence pool that are able to bind to purified Tat-2 protein. The sequences of the selected RNAs demonstrated a consensus feature: 20 of 27 RNAs contained computer-predicted loop structures that were >50% U or C nucleotides. A selected RNA was characterized for its in vitro binding properties to various Tat-2 proteins. This synthetic RNA was bound by wild-type Tat-2 proteins with an affinity that was only slightly lower than that of the natural HIV-2 TAR RNA. Tat-2 required a wild-type RNA binding domain to bind to this synthetic RNA. This study indicates that in vitro selection techniques can be used to investigate Tat protein-TAR RNA interactions. Copyright 1997 S. Karger AG, Basel

Risk management in chemical safety: some general observations relating to the state of the art.

Risk management for chemical safety begins with judgement and produces arbitrary regulations. In between, technical risk assessment has value in clarifying the behaviour of particular chemical risk systems. Risk quantification can be useful in some cases and quite misleading in others. In no case can it alone be the decisive factor. In the public's perception, chemicals are feared most for potential insidious exposure in the general environment. Risk management then must deal with high system uncertainty and information heavily influenced by value judgements. The technical contribution to risk assessment is reduced, in real terms, as the decision-making scenario expands with its increasing exposure to system 'noise', interference and misinformation. Improved communication in the system requires the use of simplified language. The suitability of risk estimation methodologies must be assessed in terms of general comprehensibility in the management for chemical safety will remain, perhaps always, more of an art than a science, concerned with an interplay of information that can be 'rationalised' with other information that defies such rationalisation.

TAR RNA binding properties and relative transactivation activities of human immunodeficiency virus type 1 and 2 Tat proteins.

Using gel shift assays, we found that the human immunodeficiency virus type 1 (HIV-1) Tat protein (Tat-1) bound both HIV-1 and HIV-2 TAR RNAs with similar high affinities. In contrast, the HIV-2 Tat protein (Tat-2) bound only TAR-2 RNA with high affinity. We conclude that the weak in vivo activity of Tat-2 on the HIV-1 long terminal repeat that has been observed previously is likely the result of low affinity for TAR-1 RNA. Additionally, TAR-2 RNA was found to contain multiple specific binding sites for Tat proteins. GAL4-Tat fusion proteins were analyzed to compare the relative transactivation activities of Tat-1 and Tat-2 in the absence of requirements for binding to TAR RNAs. The GAL4-Tat-2 protein was found to transactivate synthetic promoters containing GAL4 binding sites at levels severalfold higher than did the GAL4-Tat-1 protein.

Lentivirus Tat proteins specifically associate with a cellular protein kinase, TAK, that hyperphosphorylates the carboxyl-terminal domain of the large subunit of RNA polymerase II: candidate for a Tat cofactor.

Efficient replication of human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) requires the virus transactivator proteins known as Tat. In order to understand the molecular mechanisms involved in Tat transactivation, it is essential to identify the cellular target(s) of the Tat activation domain. Using an in vitro kinase assay, we previously identified a cellular protein kinase activity, Tat-associated kinase (TAK), that specifically binds to the activation domains of Tat proteins. Here it is demonstrated that TAK fulfills the genetic criteria established for a Tat cofactor. TAK binds in vitro to the activation domains of the Tat proteins of HIV-1 and HIV-2 and the distantly related lentivirus equine infectious anemia virus but not to mutant Tat proteins that contain nonfunctional activation domains. In addition, it is shown that TAK is sensitive to dichloro-1-beta-D-ribofuranosylbenzimidazole, a nucleoside analog that inhibits a limited number of kinases and is known to inhibit Tat transactivation in vivo and in vitro. We have further identified an in vitro substrate of TAK, the carboxyl-terminal domain of the large subunit of RNA polymerase II. Phosphorylation of the carboxyl-terminal domain has been proposed to trigger the transition from initiation to active elongation and also to influence later stages during elongation. Taken together, these results imply that TAK is a very promising candidate for a cellular factor that mediates Tat transactivation.

HIV-1 Tat protein is able to efficiently transactivate the HIV-2 LTR through a TAR RNA element lacking both dinucleotide bulge binding sites.

Each of the two stem-loop structures in the HIV-2 TAR (TAR-2) RNA element contains a dinucleotide bulge that specifies a binding site in vitro for the HIV-2 Tat transactivator protein. A TAR-2 RNA with both bulges deleted is very weakly transactivated in vivo by the HIV-2 Tat protein. To gain insight into general features of Tat protein:TAR RNA interactions, we have analyzed the significance of the dinucleotide bulges in TAR-2 RNA for in vitro binding and in vivo transactivation by the related HIV-1 Tat protein. The HIV-1 Tat protein has been shown previously to bind efficiently to wild-type TAR-2 RNA and fully transactivates the HIV-2 LTR. We found that the 5' proximal bulge and the 3' distal bulge appear to specify a high and low affinity binding site in vitro, respectively, for the HIV-1 Tat protein. Wild-type TAR-2 RNA was found to be able to bind HIV-1 Tat proteins simultaneously at each bulge binding site in vitro. A TAR-2 RNA with both bulges deleted was greatly defective for in vitro binding by the HIV-1 Tat protein. Surprisingly, the TAR-2 RNA with both bulges deleted was efficiently transactivated in vivo by the HIV-1 Tat protein, indicating that the HIV-1 Tat protein (but not HIV-2 Tat protein) is able to strongly activate transcription of a TAR RNA with no apparent bulge binding site.

Functional significance of the dinucleotide bulge in stem-loop1 and stem-loop2 of HIV-2 TAR RNA.

Human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) encode related proteins called Tat-1 and Tat-2, respectively, that bind directly to the TAR RNA element contained at the 5' ends of viral transcripts and thereby stimulate transcription through an as yet unidentified mechanism. The determinants in the HIV-1 TAR element (TAR-1) that specify binding by the Tat-1 protein have been extensively characterized, while little is known about determinants in the HIV-2 TAR element (TAR-2) that specify binding by the Tat-2 protein. The HIV-2 TAR RNA element (TAR-2) is known to be composed of two stem-loop structures. A dinucleotide bulge is found in each stem of TAR-2 RNA, analogous to the crucial trinucleotide bulge in the single stem-loop of HIV-1 TAR RNA that is the primary binding determinant for binding by the HIV-1 Tat protein. Our results of a nuclease digestion analysis demonstrated that the 5' proximal bulge in TAR-2 is significantly less sensitive to digestion by single-strand specific nucleases than the 3' distal bulge, suggesting that the 5' bulge may be involved in tertiary interaction with other regions of TAR RNA. Deletion of both bulges reduced binding in vitro by the Tat-2 protein and largely abolished transactivation in vivo by Tat-2. Deletion of either bulge alone simplified the pattern of protein/RNA complexes in a gel shift assay, but did not reduce the overall binding affinity of Tat-2. Deletion of the 5' bulge reduced Tat-2 transactivation in vivo to a level approximately 30% that of wild-type TAR-2, while deletion of the 3' bulge had no measurable effect in vivo. Our results suggest that each dinucleotide bulge specifies a Tat-2 binding site, but in the wild-type TAR-2 element the 3' bulge binding site does not appear to be utilized in vivo.

Exon2 of HIV-2 Tat contributes to transactivation of the HIV-2 LTR by increasing binding affinity to HIV-2 TAR RNA.

Human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) express related Tat proteins that are encoded in two exons. Tat proteins bind directly to the TAR RNA element contained in the 5' ends of viral transcripts and thereby stimulate transcription through an as yet unidentified mechanism. We have investigated the functional significance of exon2 of the HIV-2 Tat protein by examining properties of proteins consisting of exon1 alone or exon1 + 2. In transactivation assays in vivo, exon2 modestly increased HIV-2 Tat stimulation of transcription from the HIV-2 long terminal repeat (LTR) but had no effect on transcription from the HIV-1 LTR. In HeLa cells, exon2 increased transactivation of the HIV-2 LTR by approximately three-fold, while in COS and Jurkat cells this value was less than two-fold. In binding assays in vitro, exon2 increased the binding affinity of the HIV-2 Tat protein to HIV-2 TAR RNA. Results with GAL4 fusion proteins and a synthetic promoter containing GAL4 DNA binding sites indicated that exon2 does not contribute to the HIV-2 Tat activation domain. These observations suggest that exon2 of HIV-2 Tat contributes to transactivation of the HIV-2 LTR by increasing the binding affinity to HIV-2 TAR RNA.

Specific interaction of the human immunodeficiency virus Tat proteins with a cellular protein kinase.

The human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) Tat proteins are related transcriptional activators whose effects are likely to be mediated by a cellular factor. Using an in vitro kinase assay, we have shown that the Tat protein of HIV-2 and the activation domain of the Tat protein of HIV-1 specifically bind to a cellular protein kinase. Mutations in Tat that abolish transactivation activity in vivo abrogate the ability of the mutants to bind to the kinase in vitro. This is the first demonstration of a cellular factor that binds to Tat that is specific for a functional activation domain of Tat and that displays a biochemical activity. Additionally, we show that the Tat protein of HIV-2 serves as a substrate of the kinase in vitro. Consistent with the in vitro results, the Tat protein of HIV-2 interacts with a cellular kinase in HIV-2 Tat-transfected cells and is phosphorylated in vivo. These results suggest that a cellular serine/threonine kinase may act as a mediator of Tat function.

Trans-activation of the human immunodeficiency virus long terminal repeat sequences, expressed in an adenovirus vector, by the adenovirus E1A 13S protein.

The human immunodeficiency virus 1 (HIV-1) long terminal repeat (LTR) sequences were inserted into adenovirus in place of the E1 region. The HIV-1 LTR contained in this recombinant adenovirus responds to trans-activation by tatIII in a HeLa cell line constitutively expressing that HIV-1 gene product. In addition, the HIV-1 LTR is activated by the adenovirus E1A 13S, but not 12S or 9S, gene product when it is supplied in trans by a coinfecting wild-type adenovirus. The Rous sarcoma virus LTR, in a similar recombinant adenovirus, is insensitive to tatIII but is also trans-activated by the E1A 13S protein. The action of the 13S E1A and tatIII proteins are additive for the HIV-1 LTR in the context of adenovirus and they appear to act at the transcriptional level. As in HeLa cells, the adenovirus-borne HIV-1 LTR is inactive in the absence of a trans-activator in H9 and Jurkat cells, two human leukemic T-cell lines. This suggests that recombinant adenoviruses have diagnostic potential for the detection of trans-activators of the HIV-1 LTR that are present in circulating human lymphocytes.

Transcriptional but not translational regulation of HIV-1 by the tat gene product.

Human immunodeficiency virus-1 (HIV-1), which causes AIDS (acquired immune deficiency syndrome), possesses an essential gene, tat, whose product, acting through the long terminal repeat (LTR) sequences of HIV-1, activates viral genes and replication. The mechanism by which tat trans-activates HIV genes is unclear. Some studies have reported that an increase in messenger RNA accumulation directed by the HIV-1 LTR can explain the action of tat, but others suggest that this increase in mRNA levels can only partially explain trans-activation, and that translational control mechanisms may also be involved. To test those possibilities we have established an efficient adenovirus system for delivering the HIV-1 LTR attached to a reporter gene (chloramphenicol acetyltransferase; CAT) into cells and monitoring its activity. The HIV-1 LTR expressed from this adenovirus responds to trans-activation in a HeLa cell line constitutively expressing the tat protein by increasing the transcription rate of the HIV-1 LTR and the accumulation of mRNA encoding CAT. In this system the translational efficiency of this CAT mRNA in the cell is unaffected by the presence of tat.

Vaccinia virus induces cellular mRNA degradation.

The infection of mouse L cells with vaccinia virus induced a rapid inhibition of cellular polypeptide synthesis and a diversion of protein synthesis to the exclusive production of viral polypeptides. This shutoff of cell-specific protein synthesis was achieved by a novel mechanism by which the virus induced the rapid degradation of cellular mRNAs. Concurrent with the degradation of cellular mRNA, the virus proceeds in the orderly temporal expression of its own genetic information. The effect of vaccinia virus infection upon two abundant L-cell mRNAs was assessed by using the highly conserved cDNA sequences that encode chicken beta-actin and rat alpha-tubulin. Hybridization analyses demonstrated that throughout infection there is a rapid and progressive degradation of both of these mRNAs. In fact, after 3 h of infection they are reduced to less than 50% of their concentration in uninfected L cells, and between 8 to 10 h they are almost entirely degraded. This observation explains in part the mechanism by which vaccinia virus inhibits host cell protein synthesis.

Interferon-mediated, double-stranded RNA-dependent protein kinase is inhibited in extracts from vaccinia virus-infected cells.

The interferon-inducible, double-stranded RNA (dsRNA)-dependent protein kinase which phosphorylates an endogenous HeLa 69 kilodalton polypeptide or exogenous initiation factor eIF2 was inhibited during vaccinia virus infection. High interferon doses (20,000 reference units per ml) did not prevent this inhibition. The inhibition required protein synthesis but not viral DNA synthesis during infection, suggesting that an early vaccinia virus gene function was responsible. An active dsRNA-dependent protein kinase could be recovered from an inactive extract by purification on polyinosinate X polycytidylate-cellulose. An inhibitor of the protein kinase, therefore, must be present in the inactive extract. Similar results have been obtained with mouse L929 cells. At early time points of infection, the protein kinase in cell extracts required exogenous dsRNA for activity. This argues against endogenous viral dsRNA and activation of the kinase in the intact cell. At late time points of infection (when vaccinia virus dsRNA was almost certainly formed), the inhibitor of the kinase is present. Accordingly, it seems unlikely that the kinase played any role in the interferon-mediated inhibition of virus growth observed in these cells under these particular conditions.