key: cord-0041782-uhfabj74
authors: nan
title: Molecular biology of human pathogenic viruses
date: 2004-02-19
journal: J Cell Biochem
DOI: 10.1002/jcb.240470903
sha: 8ee6fe82fe82af78fe1541ef37ac29bf0086dbdc
doc_id: 41782
cord_uid: uhfabj74

nan

The HSV genome consists of > 150 Kbp arranged in two stretches of quasi unique sequences each flanked inverted repeats. After infection, viral DNA circularizes. The HSV genome contains at least 74 open reading frames encoding 71 polypeptides including 8 surface glycoproteins, and numerous enzymes and other proteins involved in nucleic acid metabolism and post translational processing of proteins. Viral gene expression is tightly regulated in a cascade fashion by several viral trans-activating factors at least two of which act by interacting with cis-acting sites in viral DNA. Viral DNA is replicated as a rolling circle by proteins specified by the virus. The capsids containing DNA are enveloped at the nuclear membrane and are transported to extracellular space. The functional complexity of HSV genomes is staggering. By means of genetic engineering of deletion mutants, it has been established that the HSV genes form (a) a minimal essential set required for DNA replication, packaging and virion maturation in restricted repertory of cells, and (b) a supplementary essential set which provides precursors and substrates that may be missing in some tissue cells or factors which make the process of replication more efficient. A fringe benefit of the analyses of HSV gene functions by genetic engineering of novel genomes is the development of attenuated, immunogenic viruses for potential use in prevention of diseases caused by HSV.

In humans and in experimental animal models, HSV multiplies at the portal of entry and then infects and establishes a latent infection in sensory neurons. The latent virus becomes periodically activated and is transported to a site at or near the portal of entry where it can cause lesions. The manifestation of human infections are first infections of the mouth and genitals, recurrent lesions at the portal of entry, in the cornea and CNS, adult herpes encephalitis and devastating infections of the newborn acquired during birth. Studies on deletion mutants have led to the identification of several genes whose products are required for viral replication in the central nervous system. Several other genes which are required for the initiation of replication of latent virus have also been identified. As yet no viral gene or sequence has been demonstrated to be essential for the establishment of the latent state. The specific question as to why viral gene functions related to replication on not expressed routinely in sensory neurons remains unanswered. The efficient immortalization of human B-lymphocytes by Epstein-Ban virus (EBV) is presumably accomplished by several viral genes. One of them, the E.B. nuclear antigen-2 (EBNA-2) has been shown genetically to be required for immortalization. Two additional viral genes likely to be required include that encoding the latent membrane protein (LMP) and EBNA-1. We have found that LMP shares many biochemical properties with receptors for growth-factors. It homes to the plasma membrane, turns over rapidly, its turnover is inhibited by treatment of cells with cycloheximide and is probably preceded by its internalization from the cell surface. Functional mutants of LMP have these properties; non-functional ones do not. We suggest that these properties support a model in which LMP is a receptor-like protein or affects the function of a cellular receptor to alter the growth characteristics of cells in which it is expressed.

It is known that EBNA-1 both activates transcription and mediates plasmid replication by binding specifically to elements within oriP, the origin of plasmid replication of EBV. In an attempt to dissect EBNA-1's contributions to these two functions we constructed a chimeric protein that consists of the estrogen receptor with the DNA binding domain of EBNA-1 in place of its own. This protein activates the enhancer within oriP in the presence of its ligand but does not support the replication of oriP. We suggest that EBNA-1 conmbutes an activity to DNA replication distinct from those required for transcriptional activation. The mechanisms responsible for viral clearance and hepatocellular injury in hepatitis B virus (HBV) infection are not clearly understood, although it is generally believed that a cellular immune response to one or more HBV encoded antigens plays an important role in these processes. Using transgenic mice which express some or all of the gene products of the HBV genome we have identified multiple, direct and indirect, pathways for clearance of the HBV positive hepatocyte. First, we have observed liver cell injury after transfer of HBsAg primed spleen cells and cloned, MHC class I restricted, CD8+, cytotoxic T lymphocytes into syngeneic mice that express the HBV envelope antigens. These results suggest that HBsAg is a potential target antigen on the surface of HBV positive hepatocytes and provide the first definitive evidence that a CTL response to a defined subregion of this antigen can cause liver cell injury in vivo. Second, we have shown that HBsAg positive hepatocytes are exquisitely sensitive to the hepatocytotoxic effects of bacterial lipopolysaccharide (LPS), that LPS-induced liver cell injury can be prevented by the prior administration of antibodies to tumor necrosis factor alpha (TNF-alpha) and to gamma interferon (IFN-gamma), and that the effect can be reproduced by the administration of recombinant TNF-alpha and EN-gamma, but not by recombinant interleukin 1 or interleukin 6. We have also shown that the same cytokines modulate the steady state levels of all of the HBV-encoded RNA transcripts in this model. These results suggest that the release of these cytokines by HBV antigen specific and nonspecific infilnating inflammatory cells may influence not only the expression of the HBV genome by the hepatocyte, but also may lead to its elimination in an antigen nonspecific manner. Finally, dysregulated overexpression of the HBV large envelope polypeptide by the hepatocyte leads to the production of long, filamentous, nonsecretable HBsAg particles which accumulate in the endoplasmic reticulum, causing it to become hyperplastic, and eventually kill the cell. The prolonged liver cell injury characteristic of this model eventually leads to the development of hepatocellular carcinoma. Preliminary evidence suggests that reactive oxygen intermediates may play a central role in the pathogenesis of hepatocellular injury in this lesion. In aggregate, these results suggest that several mechanisms may be involved in clearance of the HBV-infected hepatocyte and in the pathogenesis of HBV-induced liver disease, they indicate directions for further study of the molecular and cellular basis of the chronic carrier state, and they lay the groundwork for the development of therapeutic strategies to terminate chronic HBV infection and prevent its progression to hepatocellular carcinoma. We have been studying the mechanism by which human interferon-a (IFNa) induces expression of a discrete set of genes. The signal transduction pathway used by lFNa has been traced by first identifying primary response genes, studying the physiology of their induction, identification of their regulatory elements, and analysis of the proteins mediating this regulation. We are now studying the biochemical mechanisms responsible for delivering to the nucleus the signal produced by lFNa binding to its specific cell surface receptor.

lFNa stimulated genes are directly induced at the transcriptional level in response to lFNa treatment but not by any other cytokine tested. This process is mediated by pre-existing proteins with no requirement for ongoing protein synthesis. Transcriptional induction is mediated by a cisacting DNA sequence found in the promoters of all IFNa-stimulated genes. This DNA sequence serves as the binding site for a complex of proteins, termed ISGF3, which is activated by lFNa treatment and required for transcriptional induction. These proteins are sequestered in the cytoplasm of untreated cells where they are acted upon following lFNa binding to its cell surface receptor. This activation event stimulates their translocation to the nucleus and assembly of a transcriptionally active complex on DNA. The ultimate transcription factor complex is composed of two distinct components: a 48 kDa DNA binding protein and a regulatory complex composed of three polypeptides of 84, 91, and 113 kDa. These regulatory proteins are the target for activation in response to the liganded receptor, being stimulated to translocate to the nucleus, where they modulate the DNA binding affinity of the 48 kDa DNA-binding protein. One or more of the regulatory polypeptides of lSGF3 is phosphorylated by a kinase activity which can be inhibited by staurosporine or K-252a. However, it is unlikely that a global activation of protein kinase C or calcium-or CAMP-dependent kinases is involved in lFNa signalling. Direct transfer of receptoractivated transcription factor subunits from cytoplasm to nucleus maintains the specificity and rapidity of the lFNa response in the absence of a traditional second messenger signalling system. The ability of cis-regulatory DNA sequences to alter or control the organ specificity and persistence of polyoma in mice is examined. We have previously shown that either deletions or substitutions of cis regulatory can either restrict or expand the acute organ specificity of virus replication and that this effect is generally cisrestricted for DNA replication. We further examine the genetics of acute infections and establish the importance of the Py A enhancer in this specificity. We have also examined the relationship between cell division and Acute Py infections with the aim to understand why Py infections are normally restricted to newborn animals. Host cell division in target organs is not detectably induced by Py virus replication but may be necessary. In addition, Py variants with in vivo pancreas organ specificity are not also specific for replication in transformed cells from the pancreas.

Thus cellular DNA replication control appears to be important for in vivo organ specificity of Py replication. We also examine the cis-regulatory genetics of Py persistent infections. There is a significant variability in the persistent levels of wild type Py DNA, which appears to be due to variable reactivation rates. We propose that there may be an underlying but low and stable level of persistent viral DNA.

We also show the elements within the B enhancer have significant affects on levels and cell types of persistent infections. Several different kinds of RNA-protein interactions can be studied in detail during poliovirus infection; some of these may be shared with its mammalian host and some of them are undoubtedly unique to the specialized lifestyle of a positive-strand RNA virus. Here we will discuss progress in understanding genetic recombination on the RNA level among polivirus genomes, dsRNA editing in human cells using poliovirus as an assay, and the packaging of poliovirus RNA during viral assembly.

Genetic evidence has suggested that RNA recombination of poliovirus genomes occurs during negative strand RNA synthesis, by a template-switching mechanism. We have developed a quantitative PCR method to detect RNA recombinants, thus obviating the need for genetic markers and viable progeny. The timing and strand-specificity of RNA recombination during the poliovirus infectious cycle and the development of totally in vitro RNA recombination assays will be presented.

A dsRNA-specific RNA modification activity has been identified in a variety of mammalian cells. The purpose of this RNA modification activity, which specifically converts adenosine to inosine residues in dsRNA, is unknown, but its proposed functions have ranged from natural antisense regulation to antiviral functions. We have found that this activity is inhibited in cells during the antiviral state that ensues following dsRNA treatment, suggesting that it is unlikely that the dsRNA modification activity is itself antiviral. We have investigated the frequency of A-to-I modification in poliovirus infections initiated with both single-and double-stranded RNA, using the phenotypic reversion of a temperature-sensitive poliovirus mutant as a sensitive biological assay for this conversion.

The final biological activity of poliovirus RNA in infected cells is to be specifically packaged into viral particles. However, which of the various candidate morphogenic intermediates in poliovirus assembly is reponsible for packaging poliovirus RNA has remained mysterious. We have quantified the binding of various subviral particles to biotinylated poliovirus RNA; the binding characteristics of particles from both wild-type poliovirus and a poliovirus mutant defective in RNA packaging will be discussed.

TRANSGENIC MICE EXPRESSING HUMAN POLIOVIRUS RECEPTORS: A NEW MODEL FOR POLIOhCYELITIS, Vincent R. Racaniello', Ruibao Ren', Frank Costantinit, Edward J. Gorgacz+, James. J. Lee+, 'Dept. of Microbiology and tDept. of Genetics and Development, Columbia University College of Physicians and Surgeons, New York, N.Y. 10032, and +Lederle-Praxis Biologicals, Pearl River, N.Y. 10965 Poliovirus infection is limited to primates, and within the infected animal, the virus replicates at very few sites, including lymphoid tissues of the gut and pharynx, and motor neurons within the central nervous system. Examination of the expression of the cell receptor for poliovirus (PVR) revealed the presence of PVR RNA and protein in a wide variety of human tissues, including those that are not susceptible to poliovirus infection. To further study the relationship between PVR expression and poliovirus tissue tropism, a human PVR gene was isolated and used to generate transgenic mice. Several PVR transgenic mouse lines were established that express PVR transcripts and poliovirus binding sites in a wide variety of tissues, including brain, spinal cord, kidney, liver, lung and intestine. Intracerebral inoculation of PVK transgenic mice with poliovirus type 1 Mahoney resulted in viral replication in the brain and spinal cord and development of paralytic poliomyelitis. In contrast, nontransgenic mice did not develop disease after inoculation with this strain. PVR transgenic mice also developed poliomyelitis after intravenous, intraperitoneal, and intramuscular inoculation with type 1 poliovirus. After intravenous inoculation with poliovirus, viral replication was confined to the brain and spinal cord of transgenic mice. When PVR expression was examined by in situ hybridization, RNA was detected in many transgenic mouse organs, but largely in specific cell types. Poliovirus infection is therefore restricted in the PVR transgenic mice, despite widespread expression of virus binding sites. These results suggest that poliovirus tissue tropism may be controlled by other factors required for cell entry or replication. However, the PVR is clearly the major determinant of poliovirus host range. PVR transgenic mice should be useful for studying poliovirus tissue tropism, neurovirulence, and attenuation, and for the development and perhaps testing of poliovirus vaccine strains. We have used molecular genetic approaches to define the protein and RNA determinants of picornavirus replication. The first approach involves a study of the polypeptide requirements for poliovirus RNA synthesis. Our work has focused on the role of polypeptide 3AB (a membrane-associated precursor of the genome-linked protein, VPg) in the initiation of poliovirus RNA synthesis. We employed sitedirected mutagenesis to generate a number of amino acid replacement and insertion mutants in the hydrophobic domain contained within polypeptide 3AB. One 3AE3 mutation produced a mutant virus (Sel-3AB-310/4) that was temperature sensitive for viral RNA synthesis. Biochemical analysis of in vivo and in vitro RNA synthesis of this mutant indicated that it is defective in initiation of ( + ) strand RNA synthesis and in synthesis of VPg-pU(pU), a nucleotidyl protein that has been proposed as a primer for the viral polymerase. Our results provide evidence in support of a model in which polypeptide 3AE3 serves as carrier for VP to initiate (+) strand RNA synthesis. They also suggest that a separate mechanism could be usefin the initiation of (-) strand RNA synthesis. We have extended our study of picornavirus RNA replication in an attempt to define the cis versus trum nature of replication functions supplied by specific viral polypeptides. We have engineered several deletion and insertion mutations into a bacteriophage T7 in vifro transcription vector containing poliovirus cDNA sequences. We have employed complementation analysis using either cotransfected pairs of mutant RNAs or rescue of a single transfected mutant RNA by a superinfecting virus that contains coxsackievirus B3 sequences in the 5' noncoding region of its genome. Slot blot anal sis using template-specific oligonucleotide probes allows one replicating template to be distinguished $om the other. One of the mutant viruses we have recovered, Sel-3D-18, contains an amino acid insertion in the viral RNA polymerase and is ts for RNA synthesis. Our data demonstrate that the RNA synthesis defect of 3D-18 at 39OC can be rescued in trans. A second ap roach that we have initiated to determine the viral se uences required for successful completion o ! a picornavirus life cycle involves the genetic alteration 0 1 sequences in the 5' noncoding region (5' NCR) of poliovirus RNA. We have produced deletion and substitution mutations as well as a limited set of linker scanning mutations within the poliovirus 5' NCR. Genetic and biochemical analysis suggests that lesions which disrupt predicted secondary structures within this 742 nucleotide region of viral RNA produce deleterious effects on virus growth. Some of these effects result from defective translation initiation and may be attributable to altered RNA-protein interactions within the 5' NCR of the poliovirus genome. Understanding the nature of such RNA-protein complexes will provide insights into the mechanisms employed by picornaviruses to insure efficient replication of their genetic information. Transcription factor AP-1 plays a pivotal role in control of cell growth and proliferation, and its activpty is tightly regulated in cells. Stimulation of quiescent cells by exposure to mitogens, growth factors, or othx extracellular stimuli causes a rapid increase in AP-1 activity. Transfection of a number of oncogenes including ras, raf, src, and mos also enhances AP-1 activity, suggesting that at least part of the effect of these oncogenes on gene expression is mediated by changes in AP-1 activity. The two major components of AP-1 purified from HeLa cells are the products of the two proto-oncogenes c-jun and c-fos. Both c-Jim and c-Fos proteins have a bipartite DNA binding domain, which consisits of a region rich in basic amino acids that is involved in sequence-specific DNA binding and an adjacently located leucine repeat that mediates dimer formation. c-Jun forms homodimers that binds DNA in a sequence specific manner and activate transcription from templates containing AP-1 sites. c-Jun can also form a heterodimer with c-Fos that binds DNA more tightly than the c-jun homdmer and is more potent at activating transcription.

To analyze the regulation of c-Jun transcriptional activity we have constructed chimeric proteins in which the DNA binding and dimerization domains of c-Jun are replaced by DNA binding domains of the yeast GAL4 transcriptional activator, or the bovine papilloma virus transactivator E2. The transcriptional properties of these chimeric proteins have been determined by co-transfection assays in several cell-lines. Our analysis reveals that c-Jun contains two transcriptional activation domains (A1 and A2) and a negative regulatory region (6). The A2 domain located near the DNA binding domain is constitutive. A1 located within the amino terminal half is a regulated activation domain which is active in some cell-types but not others. In vivo competition experiments suggest that A1 is regulated by interaction with an inhibitory activity that is cell-type specific and therefore A1 is a cell-type specific activation domain. The 6 region located adjacent to the A1 domain regulates it by facilitating or stabilizing the inhibitory activity. The cellspecific repression of the 6 and A1 domain is recapitulated with in vitro transcription experiments. Our findings may help explain how the retroviral homolog, v-jun, is transforming. Transforming activity of the v-jun coding sequences has been traced to the lack of a 6 region in v-jun. The 6 region is a negative regulatory region and therefore v-Jun transcriptional activity is not as effectively inhibited as that of c-Jun. Consequently, v-jun is a transforming protein probably because it is a more potent transcriptional activator than c-jun. The results of these and other experiments examining the role of the A1 domain in modulating c-jun transcriptional activity in response to physiolgical cues will be discussed.

Papillomaviruses encode a DNA binding protein -E2-that binds to the palindromic sequence ACCGN4CGGT present in multiple copies in the viral genome, predominantly in the long control region (LCR). A strong synergy in transcriptional activation is observed between two such palindromic sites when cloned upstream of an exogenous promoter suggesting that the active form may be a tetramer. The same potential tetrameric structure is required in human epithelial cells for more than additive cooperation with other transcription factors that interact with the viral LCR.

In contrast to the behaviour of other promoter or enhancer binding factors, activation by E2 is optimal when one to several binding sites for transcription binding factors are present between tne TATA box and the E2 binding sites.

The full length BPVl E2 protein, overproduced in yeast, activates in vitro transcription of the HSV TK promoter containing one to several E2 binding sites at position -109. In contrast to the observations made in vivo, no synergy is observed between several E2 binding sites. Finally, when E2 site(s) are present just upstream of the TATA box, as is the case for human genital papillomaviruses, the viral E2 protein strongly represses transcription by preventing the formation of the initiation complex. This inhibition may occur by direct interference with TFllD binding to the TATA box. The great majority of the p53 cDNAs obtained from colon carcinomas contain missense mutations that map at one of four hot spots, amino acid residues 175, 2 4 8 , 273 and 281 out of a total of 393 amino acids. oncogene to transform primary rat embryo fibroblasts. When the mutant human p53 is regulated by an inducible expression vector and is turned off in these cells, the cells change morphology and are no longer transformed, demonstrating the need for a continued presence of mutant p53 protein to maintain the transformed phenotype. A temperature sensitive p53 protein that is wild-type at 32'C and mutant at 39.5'C regulates these transformed cells so as to stop growth, at the G1-S border, at 32'C. At 37'C, half of the p53 is wild-type and half is a mutant form and the cells are transformed. In G1, the wild-type p53 is kept in a complex with hsc70 and mutant p53 protein in the cell cytoplasm.

the entry of the cell into S-phase. In this way, mutant p53 protein might act as a dominant loss of function mutation. The RB-1 product is a 928 aa nuclear protein, which has growth and tumor suppression activity. The mechanisms by which it serves these functions are not known. What is clear is that RB exists in a variety of forms, some overtly phosphorylated and most likely not identical in the state of their phosphorylation. At least one additional RB species is un-or underphosphorylated by comparison with the aforementioned forms. RB phosphorylation occurs in a cell cycledependent manner, with no apparent action in G1 and the initiation of phosphorylation occurring in a clock-work manner at the GllS boundary. The protein is also enzymatically dephosphorylated in M. RB is a known specific binding target of at least three, different transforming gene products of DNA tumor virusesadenovirus EIA, papovaviral large T Ag (T), and the €7 product of transforming strains of HPV.

RB contains a discrete domain, extending over -400 residues of its sequence, which can, as an independent unit, bind to T and E I A with the same specificity as the intact protein. Naturally occurring loss of function RB-l mutations in human tumors which give rise to inactive, but stable, products frequently map to this sequence. These two findings suggest that this domain normally contributes to RB function and that it does so, at least in part, by binding lo one or more cellular proteins.

With this as background, we have searched for RB-binding cellular proteins by fusing this domain to a 'biological hook' and then using the chimeric protein bound to a specific insoluble support as an affinity chromatographic reagent. The results indicate that a set of -7 different proteins can be identified in extracts of a variety of human cells, and their binding, as measured in vitro, fully mimics the specificity noted for the interaction of the domain with E1A and T. These data suggest that one aspect of RB action is to interact with a set of cellular proteins as part of the performance of its growth suppression function. Since T competed with these proteins for binding to the isolated domain, at least part of the mechanism used by T to disrupt RB function may be to block RB access to one or more of its normal cellular targets. Kawahara-Machi, Shogoin, Sakyo-ku, Kyoto, Japan 606. The chicken cellular Jun protein has a low capacity for inducing transformation in chicken embryo fibroblast cultures, while the viral Jun is an effective oncogenic transformer. Two changes are needed to turn the poorly transforming celluar jun gene into an active oncogene: a deletion in the amino proximal region, presumably interfering with the binding of a cellular inhibitor, and removal of 3' untranslated sequences that cause mRNA instability. Besides activating transcription Jun can also stimulate DNA synthesis. Mutants and recombinants of viral and various cellular Jun proteins show a rough correlation between their ability to transform chicken embryo fibroblasts and stimulation of DNA synthesis as well as transactivation. The viral Jun also effectively interferes with differentiation of avian myoblasts in vitro. Myogenic chicken and quail cultures expressing viral Jun show little fusion into multinucleated postmitotic myotubes and fail to express muscle specific proteins. The few myotubes that do form in these cultures fail to overexpress Jun. Initiation of the myogenic program and overexpression of Jun are mutually exclusive. V-jun transgenic mice do not show an increased incidence of spontaneous tumors. However, upon full depth wounding these animals develop persistent hypertrophic lesions at the site of the wound, some of these lesions become sarcomatous . Cell cultures derived from sub sarcomas are uniquely responsive to select growth factors which may play a role in secondary genetic or epigenetic events that are required for tumor development Genetic experiments have shown that seven HSV genes are both necessary and sufficient to support the replication of viral DNA in infected cells. To facilitate the biochemical analysis of these proteins we have expressed each polypeptide using the baculovirus expression system. The activities of the overexpressed proteins togehter comprise the core reactions thought to be necessary for the semi-conservative synthesis of duplex DNA.

The UL9 protein binds specifically to the viral origins of DNA replication. Hydrodynamic measurements have shown that UL9 exists predominantly as a homodimeric structure. The interaction of purified UL9 with wild-type and mutated versions of the origin has been studied in some detail. Our results suggest that: 1) the binding of UL9 to each of the two binding sites correlates well with the ability of the origin to function in vivo, and 2) there is a cooperative interaction between UL9 bound at both arms of the origin palindrome. In addition to origin binding, the purified UL9 protein possesses an intrinsic 3' to 5' helicase activity that is capable of unwinding duplex DNA segments in an ATP-dependent reaction.

The HSV DNA polymerase consists of a stable complex of UL30 (the catalytic subunit) and UL42. This complex can be reconstituted either by mixing of the two purified subunits, or by co-infection of insect cells with recombinant baculoviruses that express each subunit. Experiments using defined primer-templates have shown that the UL42 subunit increases the processivity of the DNA polymerase.. Gel mobility shift and DNase footprint experiments suggest that the UL42 increases the affinity of the enzyme for primer termini by anchoring the polymerase to the double-stranded DNA behind a growing strand, thereby decreasing the probability that the polymerase will dissociate from the template after each cycle of catalysis.

The UL5, UL8, and UL52 genes encode the components of a helicase-primase complex. When expressed individually, the three polypeptides are either not soluble or have no apparent enzymatic acitivty.. Co-expression of all three polypepetides, or of the UL5 and UL52 polypeptides, results in the formation of a soluble complex having both helicase and primase activities. The role of the UL8 polypeptide is therefore currently unknown. We have purified both the UL5/UL8/UL52 complex and he UL5/UL52 complex to homogeneity and are beginning investigate the activity of these proteins on templates that more closely approximate the structure of authentic HSV DNA replication intermediates.

TUMOR SUPPRESSOR PROTEIN C. Prives, P. Friedman, H. Lorimer, K. Moses, I. Reynisdottir, and E. Wang, Department of Biological Sciences, Columbia University, N.Y., N.Y. 10027 We have characterized the interactions of various forms of p53 with SV40 large T antigen (T Ag). It was determined that both murine and human wild type p53 proteins block the DNA replication initiation functions of SV40 T Ag in vitro. Two mutant p53 phenotypes were observed: mutant p53 that fails to inhibit DNA synthesis at all concentrations tested, and mutant p53 that inhibits this process only at low but not at higher concentrations. This latter phenotype suggests that such p53 may preferentially self-associate at higher concentrations. Interestingly wild type human p53 synthesized in bacteria, and thus presumably unmodified, behaves like the second type of mutant suggesting that phosphorylation of p53 may regulate p53:T vs p53:p53 interactions. This hypothesis is currently being tested. Understanding the processes by which wild type and mutant p53 proteins interact with T antigen and with eachother may provide insight into the roles of p53 within cells.

Our studies with polyoma T antigen (Py T Ag) have revealed similarities and differences when compared to its SV40 counterpart. We have established that ATP induces the assembly of Py T Ag into hexamers and alters its interactions with the viral replication origin. Py T Ag, like SV40 T Ag, is a DNA helicase that can unwind duplex DNA. However, unlike SV40 T Ag, Py T Ag displays little sequence specificity for DNA unwinding under normal replication conditions suggesting that additional activities of Py T Ag are involved in its replication initiation functions. Related to this Py T Ag displays a specific interaction with murine but not human DNA polymerase. Another marked difference between the two T antigens is related to the effects of phosphorylation upon their replication function. Treatment of SV40 T Ag with alkaline phosphatase (CIAP) activates its replication function by increasing its binding to the SV40 origin. By contrast, removal of phosphates from Py T Ag by CIAP blocks Py ori-DNA replication in a manner that is apparently unrelated to its binding to the Py origin. The replication of plasmid DNAs containin the Simian Virus 40 origin of DNA replication has been reconstituted in uitro with highly puriaed proteins. In addition to the virus encoded large tumor antigen (TAP), which functions as an origin recognition protein and DNA helicase, seven cellular proteins have been identified as essential components. These include replication factor A (RF-A), a multi-subunit single strand DNA binding protein (SSB) that is required for initiation and elongation; RF-C, a multi-subunit DNA-dependent ATPase and primer binding protein; PCNA, a co-factor for RF-C and DNA polymerase 6; polymerase a/ primase complex; DNA polymerase 6 and the to oisomerases I and 11. The mechanism of DNA replication using these proteins will be discussel To begin to understand the regulation of DNA replication in euka otic cells, we have focused our attention on the initiation and ori unwinding events, both of w h i z require RF-A. RF-A consists of three protein subunits with molecular weights of 70K, 34K, and 11K in both human and yeast cells. The 70K subunit is the SSB, but phosphorylation of the 34K subunit enhances RF-A DNA binding and function. The RF-A kinase has been purified to homogeneity and was identified as the cell-cycle regulated cdc2-cyclin kinase complex.

The 34K subunit is phosphorylated in a cell-cycle dependent manner, being hypo-phosphorylated in the G1 hase and hyper-phosphorylated in the S and G2 phases. These observations provide a direct lint between regulation of DNA replication and cell-cycle control.

Medicine, New Haven, CT 06510 The primary bovine papillomavirus (BPV) gene responsible for morphologic transformation of established lines of rodent fibroblasts is the E5 gene which encodes a 44-amino acid, very hydrophobic protein that is among the smallest transforming proteins known. Efficient transformation by the E5 protein requires a very short "active site" that may consist of as few as seven amino acids. The overall hydrophobicity of the molecule is also important although many random hydrophobic sequences can functionally substitute for the wild type one. We have found that the receptor for platelet-derived growth factor is specifically and constitutively activated in fibroblasts transformed by the BPV and deer papillomavirus (DPV) E5 genes. When the E5 protein is delivered acutely into cells, receptor activation precedes the proliferative response to this protein. December, 1990). The E6 promoted degradation of p53 is ATP dependent and involves the ubiquitindependent protease system. The selective degradation of cellular proteins, such as p53 with negative regulatory functions, provides a novel mechanism of action for dominant acting oncoproteins. Thus the papillomaviruses appear to exert some of their proliferative and oncogenic effects through the interactions of their encoded proteins with key cellular proteins. The long terminal repeat (LTR) of HIV-1 contains multiple cis-acting regulatory elements that modulate the degree of HIV RNA synthesis and viral replication. These include the target for tat transactivation, the TAR region, as well as two binding sites for the cellular transcriptional factor NF-KB and three binding sites for Spl. Previous studies have demonstrated the role of the NF-KB sites in mediating LTR activation by phorbol ester and cytokines in both transient transfection assays and in chronically-infected T cell and promonocytic cell lines. The NF-KB sites are not required for HIV infection of T lymphocytes however, as deletion of these sites from an infectious molecular clone of HIV does not block viral replication. Deletion of the three LTR Spl binding sites results in a virus that is capable of infecting some, but not all, T cells. The Spl deleted virus replicated efficiently in human T cells containing NF-KB binding activity such as periperal blood lymphocytes and MT4 cells, however replicated poorly, or not at all in cells lacking NF-KB activity such as A3.01 cells. Activators of NF-KB could induce replication of the Sp-deleted virus in A3.01 cells, suggesting that NF-KB can functionally substitute for Spl in activating HIV replication. Deletion of both NF-KB sites and all three Spl sites resulted in loss of viral infectivity as did a mutation in the TAR region that eliminated tat transactivation. In recent studies, the roles of the individual NF-KB and Spl binding sites have been evaluated by reconstructing LTRs containing these sites individually and in combination. Viruses containing juxtaposed NF-KB and Spl binding sites are more potent activators of HIV replication than those containing two copies of either site alone; two copies of either site are more efficient in supporting viral replication than are single sites. These studies suggest that the HIV LTR is similar to the promoter/enhancer elements of other viruses in that it is composed of multiple functional elements that contribute differentially to viral replication in different cell types. Human T cell leukemia virus type 1 (HT.LV-1) is a n etiologic agent of adult T cell leukemia (ATL) and also associated with tropical spastic paraparesia (TSP). Viral replication and gene expression are regulated by at least two trans-acting regulators, tax and rex, that are encoded by its own genome. Tax protein is a transcriptional activator that activates the viral enhancer consisting of 21 bp. Cellular proteins (TREB) that bind to the 21 bp viral enhancer and their cDNA were characterized. TREB proteins contain leucine zipper and basic amino acid domain structure and are member of a CREB family. TREB proteins were found to activate or suppress the gene expression in cell line-specific fashion. Interaction of tax protein with these TREB proteins and also with the DNA are currently investigated. Tax also activates cellular genes such as IL-2, IL-2 receptor, GM-CSF, FOS. PTHrP and MHC class I. A cellular DNA binding NF-rB was proposed to be involved in the activation of IL-2 receptor a gene. These activations of cellular gene may explain some phenotypes of ATL cells and ATL patients, however, may not be sufficient to explain malignancy of ATL cells..The other trans-regulator rex modulates RNA processing and stimulates the expression of unspliced mRNA that codes gag, pol and env. On the other hand, rex repress the expression of fully spliced tax/rex mRNA. For these regulation. rex require the specific sequence in the 3LTR that forms a very stable secondary structure. and rex recognizes the secondary structure.

Combination of transcriptional activation with tax and its repression by rex modulating RNA processing makes the viral gene expression transient and keeps it lower level, thus allow HTLV-1 infected cells to escape from the host immune response. 

We have previously documented the emergence in vivo of HIV-1 variants with reduced sensitivity to neutralization by autologous sera. The V3 domain, which has been proposed as the principal target for neutralization, was amplified from sequential virus isolates from four patients by a polymerase chain reaction with nested primers and directly sequenced by a novel semi-automated DNA sequencing method. Results: The deduced amino acid sequence of the V3 domain remained unchanged over time in three patients, despite the emergence of neutralization-resistant virus variants in two patients. In one patient only the emergence of neutralization-resistant virus variants coincided with mutations in the V3 domain. Conclusions: We show that resistance of HIV-1 to neutralization by autologous sera may be caused by mutations both inside and outside the V3 domain. The complete envelope gene will be sequenced to establish the mechanism for virus escape from neutralization in all four patients. The presence of budding C-type and intracytoplasmic A-type particles in Chinese hamster ovary (CHO) cells is well documented. However, extensive screening has failed to detect any evidence of infectivity. Continuous-flow ultracentrifugation has been used to concentrate extracellular particles from culture fluid of a recombinant CHO cell subclone for molecular characterization. Particles exhibiting reverse transcriptase activity and associated with mammalian C-type retrovirus structural antigens banded in sucrose gradients at a density characteristic of retroviruses. Double-gradient purified particles contained RNA which hybridized to probes for murine leukemia virus, and endogenous Chinese hamster intracisternal A-particle (IAP) elements. DNA sequence analysis of a cDNA clone isolated from purified particles revealed multiple interruptions of the endonuclease reading frame, providing one possible explanation for the non-infectious nature of the observed particles. Sequences present as RNA in purified particles were also present as conserved, repetitive, provirus sequences in genomic DNA of all CHO cell lines examined and in Chinese hamster liver DNA. The observed particles are therefore likely to be the products of endogenous retroviruslike elements present in the germline of Chinese hamsters. We analyzed immunological signals and molecular events leading to transactivation of the HIV-1-LTR in human IL-2 dependent, cloned T lymphocytes transiently transfected with a luciferase gene under the control of various regions of the HIV-1-LTR. Specific antigen presentation (tetanus toxoid) by autologous macrophages to a specific T cell clone led to clear transactivation of the HIV-LTR. PMA similarly transactivated the whole HIV-LTR or the HIV-enhancer and translocated NFkB-like proteins to the nucleus. Anti-CD3 antibody induced TNF and lymphotoxin production in the cells and promoted massive nuclear translocation of NFkB-like proteins, however, this did not lead to HIV-LTR transactivaiton.

Neither did IL-2 transactivate LTR activity despite stimulation of T cell proliferation.

These findings suggest that events linked to T cell activation, triggered by potent inducers (antigen recognition, PMA), but not by TNF or anti-CD3 antibodies, are necessary to induce functional NFkB-HIV-enhancer complexes. Efficient transactivation in human T cells may rely on post-translational modifications of one or more proteins of this complex. Infection with HIV-1 may lead either to destruction of the host cell or to a persistant, noncytopathic infection. The outcome has been correlated with the amount of unintegrated viral DNA present in the host cell. While accumulation of large quantities of unintegrated viral DNA is a hallmark of an acute, cytopathic infection, significant amounts of unintegrated viral DNA are not generally observed in persistent HIV-1 infections. This has been ascribed to immunity from reinfection due to the loss or down-modulation of the CD4 receptor, a feature of most continuously infected cell lines. Using PCR of crude cell lysates, we have detected unintegrated HIV-1 DNA in chronically infected T lymphocytic (ACH-2, J1) and promyelocytic (OM-10.1) cell lines. Treatment with AZT or soluble CD4 inhibited accumulation of unintegrated viral DNA at least tenfold within 72 hours; removal of AZT permitted recovery to pre-treatment levels within 72 hours. Our results indicate that unintegrated HIV-1 DNA is unstable in these cell lines and originates from a continuous process of reinfection. Interestingly, the accumulation of unintegrated viral DNA was not correlated with the level of CD4 surface expression. By flow cytometry, OM-10.1 cells had relatively high levels of both surface CD4 expression and unintegrated viral DNA. ACH-2 and J1 cells were CD4-negative by flow cytometry, yet ACH-2 cells had a very low level of unintegrated viral DNA and J1 cells had a level of unintegrated viral DNA similar to OM-10.1 cells. This implies that the number of CD4 receptors is not rate-limiting for reinfection. We have been able to detect unintegrated HIV-1 DNA in PBLs from asymptomatic individuals as well as AIDS patients. Clinical studies are underway in our laboratory to establish a correlation between unintegrated DNA and disease progression. An AIDS patient presented with atypical syncytial variants of high grade B-cell lymphomas of the Burkitt's type.

Electron microscopic examination of two established lymphoma cell lines from the patient indicated that large quantitics of a type D retrovirus were being produced.

Reverse transcriptase template-primer preferences, Southern hybridizations, immunoblot analysis and PCR amplification confirmed that the isolate was not HIV-1, HIV-2, HTLV-I, or HLTV-11. Using degenerate primers to a conserved region of p o l in all known retroviruses, we cloned a 135 bp fragment o f the isolate and DNA sequence analysis of this fragment indicated that the isolate was nearly identical to the Mason-Pfizer monkey virus (MPMV), a type D simian retrovirus that causes an AIDS-like syndrome in rhesus monkeys. Amino acid analysis of the p27 and p14 proteins of the isolate, along with Southern hybridization and immunohlot analysis, confirmed the relatedness of the the isolate to MPMV. The patient's serum recognized the viral isolate by immunohlot analysis, ELISA techniques and by immunoprecipitation of the p27 major capsid protein o f MPMV.

Additionally.

PCR analysis of the patient's original diagnostic hone marrow biopsy with MPMVspecific primers demonstrated the patient was infectcd with the isolate. This i s the first confirmed report of a human infected with a simian type D retrovirus.

The serologic prevelance of this isolate and its possible pathologic effects will be discussed. The endogenous reverse transcriptase reaction of equine infectious anemia virus (EIAV) has been characterized. In contrast to results reported for other retroviruses, synthesis of EIAV full length ( -) strand DNA was not impaired by high concentrations of NP-40, the nonionic detergent used to make the virion envelope permeable. titrated for maximum synthesis of complete ( -) strands and a time course was determined. Minor subgenomic bands were observed and both size and proportion varied with reaction conditions. Conditions established for EIAV DNA synthesis produced only modest yields of full genome length human immunodeficiency virus-1 (HIV-1) DNA synthesis. Prominent subgenomic reverse transcripts were observed in the HIV endogenous reaction, suggesting premature termination was a frequent event. Also with HIV-1, in contrast to EIAV, the synthesis of high molecular weight DNA was drastically reduced at NP-40 concentrations above 0.02%. These results indicated that a detergent stable core is not a property shared by all lentiviruses. The EIAV virion synthetic machinery is unusually stable and provides a convenient system for further study of in vitro replication of retroviruses.

Other components of the reaction were HIV-1 infects susceptible T-cells by binding to CD4 molecules at the cell surface and then directly fusing with the cell membrane. The binding function is mediated by specific association of the external subunit of the viral envelope glycoprotein (gpl20) with CD4, whereas the fusion event involves interaction of the transmembrane subunit (gp41) with the target cell membrane. A similar type of interaction is believed to mediate fusion between HIV-infected cells and CD4+ cells. We are investigating the possibility that CD4 binding might induce fusionrelated structural changes in the envelope glycoprotein (gp12O/gp41) complex. Soluble CD4 (sCD4) was found to promote dissociation of gp120 from the HIV-1 envelope glycoprotein complex expressed on the surface of cells infected with a vaccinia virus vector. sCD4-induced release of gp120 was dependent on time, temperature, and sCD4 concentration, and was associated with specific binding between these molecules. Release of gp120 was also induced by synthetic peptide derivatives corresponding to residues 81-92 of human CD4 (the CDR3like region), which have previously been shown to inhibit HIV-mediated cell-cell fusion. A peptide containing the substitution of gly for glu at the postion correponding to CD4 residue 87 had greatly reduced gp120 release activity, consistent with the effect of this substitution on the cell fusion inhibitory activity of the peptide, as well as on the ability of membrane-associated CD4 to mediate cell fusion. We are currently investigating the region(s) of the CD4 and gp120 molecules involved in CD4-induced dissociation of the gp12O/gp41 complex, as well as developing methods to test whether the fusogenic N-terminus of gp41 becomes exposed upon CD4 binding. Diseases, Centers lor Disease Control, Atlanta, GA 30333. Using the CD4+. latently-HIV-1-infected promyelocylic clone, OM-10.1, we have investigated the mechanisms whereby these cells become CD4-and HIV-l+ (by FACS and IFA respectively) within 24 hours following TNF-a treatment. When TNF-a is removed from these cultures, surface CD4 antigen returns and HIV-1 expression subsides by 3 days.

In TNF-a pulse experiments, where OM-1 0.1 cells were treated for brief periods with TNF-a and then placed back into medium for 24 hours, only a subpopulation of cells were found to undergo HIV-1 activation. The percentage of activated OM-10.1 cells, based on CD4 down-modulation, continued to increase during the 24 hour incubation and was variable depending upon the length of the initial TNF-a treatment (25% CD4-following a 15 minute treatment vs 64% CD4-following a 4 hour treatment). To better understand this control of latency, OM-10.1 cells were again treated for brief periods wilh TNF-a and then cultured in the presence of either anti-TNF-a antibodies or the PK inhibitor H-7 for 24 hours. Following TNF-a treatment, culture of OM-10.1 cells in the presence of anti-TNF-a antibodies was found to prevent the increasing HIV-1 activation. Yet, once the OM-10.1 cells had committed to activation during the treatment p e r i i , anti-TNF-a antibodies could not return these cells to a state of latency. However, H-7 treatment of OM-10.1 cells aHer TNF-a induction prevented further activation and quickly returned the committed cells to a latent HIV-1 state (within 24 hours) as compared to medium or HA1004 control cultures (72 hours).

In conclusion, anti-TNF-a antibody treatment was found to prevent the progressive increase towards HIV-1 activatiin while H-7 treatment caused a reversion of the committed cells and a return to HIV-1 latency in TNF-a-pulsed OM-10. An M-rich synthetic gene encoding HIV I protease was synthesized and cloned into an expression vector controlled by the lambda pL promoter and regulated by the c1857 thermolabile repressor. This plasmid was transformed into E. coli RV308 cells.

SDS-PAGE and Western blot analysis of the heat induced plasmid revealed a greatly elevated level of expression as compared to the natural HIV I protease sequence in the same expression vector. The expressed protease was primarily in the form of inclusion bodies ( granules ). The protease has been purified and refolded to an active form from these granules. This active form of the protease cleaved recombinant HIV I GAG polyprotein properly. Amino acid sequence analysis of the purified protein showed that the expressed protein was indeed HIV I protease and that greater than 95% of the initiator methionine was cleaved off. Possible factors that may lead to the increased synthesis of protease will be discussed. The human immunodeficiency virus (HIV) is the primary etiological agent of acquired immunodeficiency syndrome (AIDS). Current anti-viral therapies against HIV have shown only limited success. In this study, we have examined the possibility of using recombinant antibody molecules to inhibit the functions of the essential viral regulatory proteins, Tat and Rev. We have isolated monoclonal antibodies which interfere with the ability of Tat and Rev to bind to their respective RNA targets, TAR and RRE. Using a PCR approach, we have cloned cDNAs encoding the heavy and light chain variable domains from these "neutralizing" antibodies and have constructed murine-human chimeric antibody molecules and single chain antibodies which retain the ability to bind Tat and Rev. A single chain antibody against the Rev protein has been synthesized in E . coli and is capable of interfering with Rev-RRE interaction in vitro. Modified recombinant antibody molecules capable of localizing to the nucleus will be tested for their ability to interfere with tat and rev function and will be discussed. In addition, recombinant antibodies against gp120 and CD4 have been constructed and their ability to prevent viral replication will be discussed.

STRONG The HIV-1 tat protein functions as a potent transactivator of viral gene expression, requiring for this activity a 59 base RNA stem-loop structure, TAR, located at the 5' terminus of the HIV-1 transcript. To determine whether tat can bind TAR RNA in vitro, we have conducted gel retardation assays using 32P-labelled RNA and highly purified tat in which cysteine residues were blocked by sulfitolysis. Under nondenaturing binding and gel electrophoresis conditions tat was found to bind,nonspecifically to non-TAR as well as TAR RNA species. In contrast, specific TAR-tat complexes were observed on 7 M urea acrylamide gels following bindinq in the presence of 0.1% to 1% SDS, 4 M urea or 1 M sodium iodide. iMutations in the lower stem, upper stem and bulqe of the TAR-stem-loop resulted in reduced binding under these conditions, but binding was unaffected by mutations in the loop. These data suggest that strong denaturing conditions can reveal protein and/or RNA regions required for formation of specific complexes, and that under these conditions tat cysteine residues are not directly required for RNA binding. The denaturation conditions used may induce protein or RNA conformational changes that mimic changes induced in vivo by cellular factors. Early, regulatory gene products are encoded by multiply spliced mRNA species and are expressed constitutively. In contrast, the late, structural proteins of HIV-1 are encoded by an unspliced or singly spliced class of viral transcripts whose cytoplasmic expression is dependent upon the viral Rev trans-activator. We will present data demonstrating that the viral Vif and Vpr gene products are encoded by singly-spliced viral mRNAs whose expression is activated by Rev. This activation is shown to result from the reduced utilization of splice sites adjacent to or within the vif and coding sequences. The observation that vif and belong to the late class of HIV-1 gene products is consistent with the hypothesis that Rev activates the expression of incompletely spliced HIV-1 transcripts by circumventing a cellular barrier to nuclear RNA export that is imposed by the presence of intact splice sites.

Chin-Yih Ou, and Gerald Schochetman. Centers for Disease Control, Atlanta, GA 3 0 3 3 3 .

Host cell factors are believed to be involved in the interaction of tat with TAR. We investigated whether human chromosome 12 in human-hamster hybrid cells, which supports Highly cytopathic Zaiiian virus HlVl NDK was compared with HlVl B R U prototype to perform genetic analysis of phenotypic differences in growth properties, cell killing ability and fusogenic potential. Recombinant provirus molecules derived from DNA cloned HIVl B R U and HIVl NDK were constructed using conservative BssHII, SpeI, EcoRI and XhoI sites. Recombinant viruses obtained after transfection were standardized according to RT activity and end point dilution. Their replication and killing characteristics were measured by RT and M'IT tests, respectively, in MT4 and CEM cells ; fusogenicity was determined in MT4 cells. Formation of large syncytia, rapid cell killing and early onset of replication were characteristics of HlVl NDK.

HIVl NDK derived SpeI/EcuRI fragment covering part of the gag and pol genes conferred the early onset of replication to recombinant viruses. All those recombinants were also rapidly killing the target cells. However not all of them were highly fusogenic. Combination of HIVl M)K derived fragments Bss/HII/SpeI and EcuRI/XhoI was necessary for the full expression of HIVl NDK fusonegic phenotype ; the HlVl NDK env and gag genes alone were not sufficient to obtain high titer of large syncytia. Different regions of HIVl genome were responsible for the direct-killing and fusogenic effect. LTR and regulatory genes vpr, tat and rev were not responsible for the variation of HIVl replication kinetic. HIV-1 reverse transcriptase forms a heterodimer composed of two subunits (p51 and p66), which have identical N-termini. Compared to p51, the p66 subunit has an additional C-terminal domain, p15. The amino acid sequence of the p15 domain shows homology with RNase H of other retroviruses as well as with E. coli RNase H. T o examine the functional interaction between the domains of HIV-1 RT, we expressed them separately and purified both to homogeneity. The N-terminal domain (p5 1 ) behaves as a monomeric protein exhibiting salt sensitive DNA polymerase activity. The isolated C-terminal domain (p15) has no detectable RNase H activity. However, purified p51 and p15 combined in vitro reconstitutes RNase H activity on a defined substrate. These results demonstrate that domains of HIV-1 RT, although structurally distinct, are functionally interdependent. The p15 domain has been crystallized as large trigonal prisms which diffract X-rays to a resolution of at least 2.5 A. A full three dimensional structural elucidation of the p15 domain is now underway. Why one group of HTLV-I seropositive individuals develops a neoplastic condition, another develops neurologic symptoms, and a third group remains as asymptomatic carriers of the virus is not known. The various mainfestations of an HTLV-I infection may be related to differences in genetic backgrounds of individuals, infection with variant strains of HTLV-I or differences in target tissue tropism of the virus. Another possibility is that differences in host immune responses to HTLV-I contribute to these varied HTLV-I associated conditions. While the humoral response to HTLV-I is well characterized, little is known about the human cellular imune response, such as cytotoxic T cells (CTL) particularly in patients with neurological diseases associated with HTLV-I.

Here we report the presence of high levels of circulating HTLV-I specific CTL in patients with HTLV-I associated neurologic disease but not in HTLV-I seropositive individuals without neurologic involvement. These CTL are CD8t, HLA class I restricted and predominantly recognize the HTLV-I gene products encoded in the HTLV-I regulatory region, pX. These findings suggest that HTLV-I specific CTL may contribute to the pathogenesis of HTLV-I associated neurological disorders. To address this question we have constructed a recombinant baculovirus that expresses a p46 gag representing the HIV gag open readin frame truncated at the frameshift site and lacking the carboxy terminal p6 domain. We show t%at this recombinant produces high levels of p46 gag antigen which is secreted into the culture media as an HIV core like particle. We conclude that the carboxy terminal p6 domain of gag plays no role in particle formation suggesting that failure of p160 gag-pol t o assemble is a consequence of the presence of the p o l domain. In addition, these results suggest that replacement of the p6 domain with foreign antigens may be possible in order t o use truncated gag antigens as particulate carriers. One possibility to account for HIV infection of macrophages is that HIV initially infects cells of early myeloid lineage, which later differentiate into infected macrophages. We have investigated this possibility using the multipotent cell line HL-60, which can be induced to differentiate into various cell types under different conditions. First, we have tested the effects of cell differentiation on gene expression driven by the long terminal repeat (LTR) of HIV. When cells were treated with PMA, which causes differentiation into macrophage-like cells, CAT activity increased by 10 to 20 fold. This induction appears to be mediated through as yet unidentified inducible factor(s), in addition to NF-kB. Also, we have subcloned HL-60 cells infected with HIV. Our preliminary data suggest that viral production increases upon treatment of these cells with PMA, but only transiently. We are currently analyzing the kinetics of RNA expression upon PMA induction. Data will also be presented concerning the effect of HIV infection on cell differentiation, assayed by the expression of several cell surface proteins. vifand vpr genes. The products of these genes are not required for regulation of HIV gene expression nor are they found in mature vim particles. It seems likely that these proteins play a role in regulating virus assembly either by regulating virion protein biosynthesis or, alternatively, by acting as "molecular chaperones" that help direct virion proteins towards the budding particle on the cellular membrane. Viruses carrying mutations in the vpu gene show delayed release of progeny virus, and an accumulation of characteristic virion particles with an "immature" morphology and an unusual pattern of budding. Vpu also shows a homology to a number of membrane-bound proteins including two other viral proteins known to influence virion assembly, the foot-and-mouth disease virus p3A protein. and the influenza virus M2 protein. We have found that vpu plays a role in envelope glycoprotein levels and processing.

Efficient production of gp160. processing of gp160 to gp120 and gp41, and secretion of gp120, is only observed when the vpu and env genes are co-expressed in COS-1 cells. The major effect of vpu is to increase the inuacellular levels of both gp160 and gpl20.

Fourteen times more gp160 is produced from HIV plasmids carrying intact env and vpu genes than from vpu minus plasmids. Vpu is also active in trans, and increases gp160 expression 3to 4-fold. Cleavage of gpl60 to gp12O/gp41 takes place prior to the acquisition of endo-H resistance by the addition of complex sugars in the trans Golgi, and is independent of vpu. Treatment of cells expressing vpu. with NH&I increases the levels of gp160 and gp120, presumably by inhibiting lysozomal degradation of the envelope glycoproteins. NWCI also partially inhibits cleavage of gp160 to gp120/gp41. However, in the absence of vpu NWCI has little effect on either the level of gp160 expression or the rate of gp160 cleavage. lmmunofluaescence microscopy has demonstrated that both vpu and gp160 accumulate in wheat germ agglutinin-stained vesicles of the Golgi, but neither protein is transponed to the cell surface. In contrast. gp120 and gag proteins are found associated with cytoplasmic membranes which are not stained by wheat gem agglutinin, and a small amount of gp120 can also be detected on the cell surface. It is proposed that vpu acts as an ion channel which regulates the pH of the Golgi and thereby indirectly influences envelope protein processsing and accumulation. The latent T cell line, ACH-2, was used to investigate the accumulation of the 2-LTR circular form of (unintegrated) HIV-1 DNA after TNF-alpha (TNF-a) induction. Using PCR to amplify across the 2-LTR junction, we have determined the cellular location and relative abundance of circular HIV-1 DNA following TNF-a stimulation. A ten-fold increase in HIV-1 p24 antigen was detected in the 48-hour supernatant of TNF-a induced ACH-2 cultures (500 Ng/ml) over unstimulated controls (<50Ng/ml). Concommitantly a 10-fold increase in circular HIV DNA was detected in the treated cells by PCR analysis following limiting dilution. To identify the cellular compartmental location of the circular DNA accumulation, the cells were fractionated into cytoplasmic and nuclear components and analyzed by PCR. Southern analysis of 2-LTR amplified product from the treated cells indicated that the nucleus was the source of the 2-LTR HIV-I DNA. To support the PCR analysis, fluorescent _in situ hybridization was performed on ACH-2 cells following TNF-a induction. Using a biotinylated HIV-1 DNA probe, increased nuclear fluorescence was observed in the TNF-a induced cells over controls. These findings indicate that 2-LTR HIV DNA accumulates in the nucleus of TNF-a induced latent cells and offers insight into the role of unintegrated viral DNA in pathogenesis. To characterize a unique Guaymi Indian HTLV-I1 isolate (PNAS, Nov. 15, 1990) a peripheral blood lymphocyte (PBL) cell line (G 12.1) was established by co-culture methods. A continuous IL-2 dependent cell line has been maintained independent of normal feeder PBLs for 10 months of culture. The cell line has sustained production of soluble and cell surface HTLV-I1 p24 capsid proteins (assessed by antigen capture assay and avidin-biotin-complex assays), contains approximately 10 copies of proviral HTLV-I1 DNA per cell as determined Southern blot analysis, budding HTLV particles has been observed by electron microscopy, and lethally irradiated G12.1 cells transmitted the HTLV-I1 infection to rabbits. Western blot analysis using type specific monoclonal antibodies and PCR analysis using strain specific oligonucleotide primers and probes have confirmed the virus type to be HTLV-11. Serum derived from the Guaymi Indian donor was positive for antibody reactivity in HTLV-I1 specific u peptide-based assays and failed to react with HTLV-1 u and gag peptides. Phenotype analysis by FACS indicated the cell line to be CD4+ (Leu 3a) and CDw29+ (4B4) with high expression of CD25 (IL-2R, L243) and CD71 (transfer receptor); the cell line was negative for CD8, CD14, CD16, and CD19 antigens. In addition, the cell line produced the cytokines gamma interferon and GM-CSF, but is negative for IL-4 and TNF-alpha. A G12.1 genomic DNA library has been constructed and 4 molecular clones which hybridize to both HTLV-I1 and gjg specific DNA probes have been isolated. These clones are currently being exploited for nucleotide sequencing and other studies. The G12.1 cell line will allow continued characterization of this unique HTLV-I1 isolate.

Michael', R. Lal', B. Roberts', K. All replication-competent proviral clones of HIV-1 available to date have been obtained from cell cultures of amplified virus. In this project, we sought to molecularly clone and genetically and biologically characterize fulllength HIV-1 proviruses directly from uncultured human brain tissue of a patient with AIDS Dementia Complex (ADC). Specific objectives were (i) to generate replication competent proviral clones from uncultured human tissue thereby allowing an analysis of genome organization and gene structure-function relationships of virus not subjected to in vitro selection pressures, (ii) to generate transfection-derived HIV-1 strains without interim cell culture for analysis of replicative DNA intermediate forms, integration status, and the possible existence of defective andfor helper forms, (iii) to compare and characterize genotypic variation of proviral clones obtained by direct lambda phage cloning to clones obained by PCR amplification. High molecular weight DNA from a brain specimen obtained at necropsy was subjected to lambda phage cloning and 10 HIV-1 proviral clones were obtained out of 8 x lo6 recombinants. These proviral clones contained integrated and unintegrated forms, forms with one or two LTRs, and genomes with large deletions and self-integrated reversed LTR sequences. Eight HIV-1 proviral clones (in lambda) and 11 PCR derived clones from the same brain were sequenced in a 525 bp hypervariable envelope region. All 19 clones were highly related yet distinct with nucleotide variation between 0.1 and 0.3%. Four proviral clones in lambda were full-length by restriction mapping, and one of these was subsequently shown to be transfection-competent in Cos-1 cells and replication competent in human T cells and monocytes after cell free passage. The genomic organization, structure-function characteristics of specific gene products, and the biological properties of these HIV-1 proviral clones derived from uncultured human brain are under study. cycloheximide. W e designated the protein HIVEN86A. Other proteins shared some of the properties of HIVENBBA. In addition, w e identified proteins that were a parently celltype specific. Recently, w e have demonstrated (with W. Greene and colyeagues, D u k e University) that HIVEN86A is the product of the human re1 proto-oncogene. W e w i l l discuss immunologic analysis of Re1 and other proteins that associate with the id3 site that are related to Rel. W e w i l l also discuss the fact that at least one protein is immunologically related to Re1 but demonstrates different binding site requirements than Rel. Functional characterization of the ICB sites within the HN1-LTR by in uitro transcription analysis of extracts derived from cells at different moments within the first two hours of stimulation by mitogenic lectin or tumor promoter reveals that eventhough they contribute to activation of transcription they are not sufficient. The role of Re1 and other proteins that interact with the id3 site in activation are being characterized biochemically and results of these studies will be presented.

Scott D. Gitlin, Cindy A. Bohan, John N. Brady, Laboratory of Molecular Virology, National Cancer Institute, NIH, Bethesda, Maryland 20892.

We have demonstrated that purified HTLV-I Tax, protein can be taken up by 702/3 lymphoid cells and localized to both the nuclear and cytoplasmic compartments. the Tax, protein into the growth medium of 702/3 cells resulted in the rapid and transient induction of NF-kB binding activity in the nuclear fraction. of NF-kB was not sensitive to either staurosporin or prolonged stimulation with the phorbol ester 12-0-tetradecanoylphorbol-13-acetate, suggesting that Tax,-dependent NF-kB activation did not require the protein kinase C pathway. Purified Tax, did not directly increase NF-kB binding activity in 702/3 cytoplasmic extracts, suggesting that NF-kB induction may require cellular factors. Western blot and competitive radioimmunoassays demonstrated that Tax, protein was present in the tissue culture media of HTLV-Itransformed cell 1 ines. gene expression in noninfected cells. We are currently investigating the possibilities that Tax may induce NF-kB through a receptor mediated signal transduction event or via direct effects of internalized Tax, on the stability of the NF-kB-IkB cytoplasmic complex. The human T-cell leukemia/lymphoma virus type I encodes a regulatory protein Tax, which activates expression of both viral and cellular genes and has been implicated in the transforming potential of the virus. Since antibodies to Tax are often found in the serum of patients infected with the virus, it is interesting 10 consider that Tax, may be present as an extracellular protein and may play a role in progression of HTLV-Iassociated disease. We have recently demonstrated that extracellular Tax, is taken up by a variety of cell types including human peripheral blood lymphocytes (PBLs) and transported to the nucleus as an intact 40 kD protein. To investigate the potential role of extracellular Tax, in cell growth regulation, human peripheral blood lymphocytes (PBLs) were cultured from healthy HIV and HTLV-I negative donors. Cultured PBLs were treated with phytohemagglutinin (PHA) followed by interleukin-2 (IL-2), purified Tax, protein, or a mock bacterial extract containing no Tax,. observed in both the Tax or IL-2 treated, but not in the control treated cells. PBLs exposed to Tax, exhibited prolonged sensitivity to interleukin-2-induced proliferation. It has previously been suggested that stimulation of cell growth could result from the transcriptional activation of IL-2 and IL-2 receptor alpha chain by Tax,. Our studies suggest a new biological role of Tax, which may contribute to the leukemogenesis and neurologic complications observed in HTLV-I infected individuals. Site-directed mutagenesis has been used to assess the importance of lysine263 i n substrate binding of human immunodeficiency virus-l reverse transcriptase.

Previous studies have indicated that lysine263 functions in the binding of 2'-deoxynucleoside-5'triphosphate (dNTP) substrates (Basu er al, JBC, 1988) . We have chosen to study this interaction directly by using site-specific mutagenesis to change lysine263 to a serine and an isoleucine. Characterization of highly purified mutant enzymes has shown that K263S and K263I bind natural dNTP substrates and primed polynucleic acid substrates with equal affinity when compared to the wild-type reverse transcriptase. No difference was observed in the binding of AZlTP to either mutant on the basis of K, and Ki determinations. Neither substitution had any effect on RNase H activity.

results indicate that lysine263 is not essential in the binding of substrates to HIV-1 reverse transcriptase. The three subfamilies of retroviruses are ancoviruses. lentiviruses. and spumaviruses.

The latter. also designated foamy or syncytium-forming viruses. are found in a variety of species. including humans. and are highly cytopsthic in tissue culture systems. In the host, spumaviruses appear to establish latency; B clear connection with disease has not yet been established. We have molecularly cloned and sequenced the genome of simian foamy virus type-1 (SFV-1). an isolate obtained from rhesus monkeys. Sequence analysis reveals that SFV-1 is 60% to 80% related to the human foamy virus (HFV) in the gag. pol. and env genes. Both SFV-I and HFV contain open reading frames (0RFs) which extend from the end of the env gene into the 3'LTR.

Analysis of the long terminal repeats (LTR) of both viruses reveals that the R-US and the U3 regions are about 85% and 30% related, respectively. Using transient expression assays in tissue culture systems. wve have demonstrated that SFV-1 encodes a transactivator that acts through sequences in the LTR.

This transactivator functions at the transcriptional level to augment initiation of viral transcripts. A cis-acting target element for the SFV-1 transactivator has been localized to the U3 region of the LTR upstream from the cap site that marks the initiation site for viral transcripts.

RNA transcripts from the O W region have been detected in infected cells and cDNA libraries have been analyzed; multiple splicing events give rise to these viral transcripts. Transient expression assays identified the first ORF (ORF1) as the transcriptional transacivator of foamy virus, designated lof. Current efforu are directed at (i) precisely defining the the cia-acting element. and (ii) elucidating the mechanism of transactivation.

Thus. members of each of the three retrovirus subfamilies (e.g..

the oncovirus human T-lymphotropic virus type-1, the lentivirus human immunodeficiency virus type 1. and the spumavirus SFV-1) encode transactivator genes that act through elements in the LTR to regulate viral gene expression. These investigations provide a basis for comparing pathogenic and nonpathogenic primate retroviruses and far determining the molecular mechanisms that may account for retroviral latency and disease.

The ORFs of SFV-1 and HFV show only 39% similarity. serology were studied. A panel of enu and gag synthetic peptides (SP) and of nef, rev and tat S P were used, respectively, for reg-PE and struct-PE. PCR analysis was performed using three sets of primers specific for select HIU-1 uiral DNA sequences. 20 age-matched HIU-1 positiue children and 15 non at-risk HIU-1 negatiue children were tested as positiue and negatiue controls.

RESULTS: 1 /36 at-risk seronegatiue child (3%) was positiue to all used techniques, 1 /36 was seroreacting to both reg-PE and struct-PE but negatiue to PCR and 1 /36 was only positiue to reg-PE and PCR . As to the positiue controls, 20/20 (1 00%) seroreacted to struct-PE, 12/20 (85%) to reg-PE and 19/20 (95%) were positiue to PCR. CONCLUSIONS: This study suggests the need of further HIU-1 diagnostic eualuation in at-risk seronegatiue children but also indicates that the accuracy of each single technique has still to be ascertained. activated, although, activation of viral sequences is normal. Also, NFkB is not activated by mitogens or cytokines known to activate NFkB binding activity suggesting that the pathways leading to NFkB activation are specifically suppressed following the long term expression of Tax.

Consistent with these findings is the fact that induction offos and jun proteins is normal in these cells, suggesting that activation of protein kinase C is not blocked. Thus, depending on the cellular environment, Tax can have both short and long-term effects on specific transduction systems.

Deaconess Hospital, and Harvard Medical School, Boston, MA 02215.

The long terminal repeat (LTR) of HIV interacts with many cellular factors. Among these, the transcription factor NF-kB has received much attention because it is a potent inducible activator of the HIV LTR. Based on this observation, it has often been proposed that NF-kB might be involved in activation of the dormant HIV provirus, resulting in the switch from the latent state to the productive stage of viral infection. However, it has recently been reported that NF-kB sequences are not essential for viral growth. Therefore, we have looked for the presence of inducible transcription factors other than NF-LB that could act on the LTR. We have consistently found that in certain cell types, there is a significant level of CAT activity even when NF-kB binding sequences are mutated or deleted. Moreover, this expression increases by 10 to 20 fold upon PMA induction.

Our preliminary data suggest that the sequence responsible for this induction is located downstream from the NF-kB binding sites, but upstream from the TAT-responsive element. We are determinins the exact location of The HIV-1 promoter can be transactivated by regulatoty proteins encoded by several pathogenic viruses which may act as critical co-factors in the pathogenesis of AIDS. Recent evidence suggests that the immediate early (IE) proteins of the Herpes Simplex (HSV) and Cytomegaloviruses (CMV) may function indirectly by modifying host cell transcription factors, one of which may be LBPl. LBPl binds cooperatively to a large region downstream from the RNA start site, a region that contributes to the basal activity of the promoter and, more recently, has been shown to be critical for responsiveness of the HIV-1 promoter to transactivation by CMV-IE proteins. We purified LBPl to near homogeneity from HeLa cells by sequential chromatography of nuclear extracts on heparin-agarose, DEAE-sepharose, and DNAaffinity resins containing multiple copies of the HIV-1 LBPl binding site. Purified LBPl migrates as a 63kDa peptide in SDS-gels and was used in DNA-footprinting and gel-shift assays to identify recognition sites in well characterized cellular and viral genes and possibly elucidate a functional role for this factor. LBPl was found to bind to transcriptionally important sites in the murine Thy-1 intronic enhancer and the human c-myc P2 and CMV-IE promoters; all regions downstream of the RNA start site. Feline Leukemia Virus ( FeLV 1 was shown to be associated with immunodeficiency syndrome in cats. As most of the retroviruses, Gag gene is translated into the polyprotein precursor Pr65, which is further processed and the Pol gene encodes a protein with several enzymatic domains.The proteinase cleaves itself out of the Pol precursor and is responsable of the processing of Pr65. In an attempt to develop a system where we could evaluate the activity of protease inhibitors against retroviral proteinases, a protease substrate plasmid was constructed to contain, under beta actin promoter control and in frame, part of FeLV Gag including a proteinase cleavage site and the Hepatitis B Virus surl'ace antigen ( HBVsAg ) coding sequences. Upon transfaion of cat fibroblasts, the levels of HBVsAg in the tissue culture was not detectable, due to the N-terminal fusion of the polyprotein produced which blocks the secretion. However if this plasmid was cotransfected with a plasmid carring the protease coding region, HBVsAg was observed. When the substrate plasmid was introduced in FeLV infected cat cells, HBVsAg was also detectable. Immunoblotting, using antibodies against HBVsAg, confirmed the proteolytic cleavage.To test if the murine retrovirus can recognize feline Gag as substrate, murine fibroblasts infected with the virus LP-BMS were transfected with pCag-IIIIV and the results clearly showed that the viral protease is able to process FeLV Gag. During analysis of the cfl'ects of inhibitors of HIV-1 protease,an intracellular accumulation of Gag-HBV was observed. These studies show this syslem to be of potential use for testing compounds that may inhibit retroviral proteinases. Attempts to apply this work to the evaluation of proteinase inhibitors i n vivo are under way.

We have been studying the molecular mechanisms by which cytokines affect For detection of pathogens, in vitro amplification techniques have many advantages compared to conventional methods due to the generality and sensitivity. We have developed an automated solid phase PCR technique which allowes for colorimemc detection and subsequently sequencing of positive samples. This approach makes it possible to perform epidemiological and drug response studies without the need of time consuming cloning procedures. The biotin sueptavidin system was used to capture the amplified fragment on magnetic solid support. The amplified material was subsequently detected using a fusion protein 1ucIlucZ (Fig) . The entire procedure can be performed within a working day. By using a direct genomic sequencing method (no cloning or subcloning) the problems with Taq misincorporation is o---avoided. Results will be presented on solid phase sequence of the envelope (the V3 domain) and R T gene of HIV-I of virus isolates from several patients. The human immunodeficiency virus type 2 (HIV-2) genome encodes a regulatory gene (tat) which is required for viral gene expression and replication. Strain-specific changes in the tat gene product may contribute to differing levels of LTR-driven gene expression. To study the tat activity of different HIV-2 strains, RNA was isolated from HIV-2 infected tissue culture cells and the tat cDNA amplified using the polymerase chain reaction procedure. Two species of tat cDNA were isolated from HIV-2Rod RNA (Genebank accession M15390); one message with the expected coding sequence and one species with an altered splice junction between exons 2 and 3 . This alteration resulted in an insertion of four amino acids. This clone also has a single base deletion which alters amino acids 91-99. These tat cDNA's were cloned into the eukaryotic expression vector pSV2-neo, and cotransfected into RD Mechanisms of action f o r this role have not been elucidated. We have been investigating a recently described human ERS (HRES) with significant homology to both HTLV-1 and Hrab4, a ras-related gene, and have assessed the representation of this sequence in the genome, its similarities to other sequences, and its transcriptional state. Southern blots of numerous cell lines were probed with restriction fragments and PCR products to define regions of HRES which were represented more than once in the genome. While there is a great degree of structural overlap between HRES and Hrab4, they are separate genes, implying a potential gene duplication. Possible significance of this duplication will be discussed. To determine whether HRES was transcribed, Northern blots were probed with an RNA sequence homologous to the coding strand of HRES. Transcripts of different molecular weights were detected. To insure that these transcripts were derived from the same gene, the RNA sequence was used to probe the Southern blots. The transcripts from HRES and Hrab4 were distinct. Investigation is ongoing to clarify a role for HRES in immunologic disease. Bovine lentivirus, known as bovine immunodeficiency-like virus (BIV), is genetically, structurally, and antigenically related to human immunodeficiency virus (HIV). This causes concern because humans consume beef and dairy products and fetal bovine serum is a component of most vaccines prepared in cell cultures. proteins cause human sera to either become HIV seropositive or display indeterminate antibody (Aby) reaction on HIV tests. We developed and used a BIV Western blot (WB) to examine sera from people "at risk" for BIV exposure and found no evidence of seroconversion to BIVspecific proteins. positive (4), Aby negative ( 4 ) , or indeterminate (16). None of these sera were positive by WB to BIV-specific proteins. Many of these sera, however, displayed a strong reactivity to both mock-infected and BIV-infected bovine cell culture antigens. samples of fetal bovine serum, each sample representing serum pooled from 1 -3 fetuses. All samples were negative on BIV WB. This agrees with our observation that calves born to BIVinfected, Aby-positive cows are negative for detectable BIV Aby at birth. To date, we have not detected any human Aby response to BIV-specific proteins. Aby response to BIV proteins is not a source of positive or indeterminate results on HIV WB. T h e use of PCR techniques t o detect HIV infectivity in cell culture and assess the effect of nucleoside analogs on HIV replication was compared with i n vitro p24 antigen production. PHA-stimulated normal PBMC were infected i n vitro with 0.1 t o 1000 tissue-culture infectious doses (TCID) of HIV IIIb in the presence of 0 to 10 p M AZT. Cellular DNA and RNA were prepared from cells collected at various time points and amplified using HIV g g primers.

Results of these experiments showed t h a t HIV RNA and DNA were detected after 24 hours in cells infected with > 1000 TCID while p24 antigen could not be measured until at least 48 hours. A significant antiviral effect could be seen after 48 hours by PCR, and in [3] [4] days by p24 antigen production. In cells infected with < 1000 TCID, HIV g g DNA or RNA could be detected in 48-72 hours while p24 antigen could not be detected for at least 4 days. Significant antiviral activity could be measured in these cultures in 72 hours by PCR, b u t required 5-7 days when measuring p24 antigen production. These results suggest t h a t PCR techniques may be useful to rapidly monitor patient viremia, strain-specific responses t o antiviral therapy, and t h e development of drug resistance. Oligonucleotide directed mutagensis was used to disrupt the "Shine-Delgarno" like sequence in order to increase expression of the 32Kda IN protein.

Expression, purification and DNA binding studies of the IN protein produced by this vector will be discussed. Previous studies from this laboratory have shown that mutants of T antigen which are unable to bind to DNA maintain the ability to transcriptionally activate promoters. These and other data suggest that T antigen mediates transcriptional activation indirectly through the utilization of cellular transcription factors. We have found that the promoter structure necessary for T antigen mediated transcriptional activation is very simple. A TATA or initiator element is required in addition to an upstream element which can be quite variable. We have found that SPI, ATF, API and SPH (TEF- The h u m a n tumor suppressor gene, p53, has been shown t o be involved in several types of carcinomas. We have expressed the wild-type h u m a n gene in S. pombe. In S. pombe h u m a n p53 seems to behave in a similar manner as it does in mammalian cells. The protein is phosphorylated, it is localized to the nucleus, a n d its expression inhibits growth. So-called oncogenic mutants of p53 were also expressed i n S. pornbe. The mutants also inhibit growth, b u t less severely t h a n the wild-type gene. One oncogenic mutant, a change of an arginine to a histidine at residue 175, was found to be dominant to the wildtype allele, thus providing direct evidence for dominant negative alleles of p53. The SV40 large T (Tumor) antigen is a multifunctional protein essential for the expression and replication of the viral genome. To initiate DNA replication, T must be phosphorylated at threonine-124, but the protein has to be dephosphorylated at serine sites. Phosphatase 2A (PP2A) dephosphorylates T at these serine sites and enhances the ability of T to support DNA replication. SV40 small t-antigen intexacts with Pp2A by binding to the regulatory subunit. This observation suggested to us that the interaction of small t with PP2A might modulate T-mediated DNA replication. To test this hypothesis, we used an in vitro DNA replication system which included an Hela cell extract, and as substrate the plasmid pZ189 (which contains the SV40 replication origin). Plasmid replication in this in vitro system is completely dependent on SV40 T-antigen.

When purified small t (400 ng) was added to this system, replication was inhibited -70%. The inhibitory effect was observed under various experimental conditions and using different T and t antigen preparations. We demonstrated by immunoblotting that t-antigen co-precipitated with P E A in this in vitro system. Our results indicate that small t inhibits DNA replication in vitro, and that this inhibitory effect may be mediated by the interaction of small t with PP2A. We have been studying the ability of a serics of SV40 T antigen mutants to immortalize primary mouse cmbryo fibroblasts, to transform established rodent cell lines, to transactivate gene expression from various promoters in monkey kidney cells, and to bind the tumor supressor gene products, p53 and Rb. We found that transactivation activity was reduced by mutations within the first 85 amino acids and that 40% of wildtype activity was retained by an N-terminal fragment of 138 amino acids. Some mutants with defects in origin-specific binding were completely defective for trans-activation. The data suggest that trans-activation is a function of the first 85-100 amino acids of large T. that the structure and orientation of this domain are essential for it to activate gene expression. Mutations at distal sites that reduce trans-activation could do so by altering the conformation of T antigen. We found no correlation between transactivation and immortalization, transformation, R b binding or p53 binding. Most of T antigen was required for transformation of REF-52 cells and immortalization of primary MEFs. Small insertions or deletions at many sites had no effect on either of these activities, but two mutants, with insertions at amino acid 409 and 424, were dcfective for both activities, even though both mutants retained the ability to bind Rb and p53. was defective for both immortalization of primary MEFs and transformation of REF-52 cells. Overall, these studies suggest that R b and p53 binding are important for immortalization and transformation but that other T antigen: host protein interactions likely also play a key role in these processes. The MHC class I was not inhibited in transformed cells by a plasmid carrying the Ad12 12s cDNA. The class I genes were only inhibited in cells expressing the Ad12 13s mRNA product. Using chimeric E1A region between Ad2 and Ad12, we have analyzed the specific domain responsible for inhibition of MHC class I expression.

The target(s) on the H-2Kb promoter was(were) also determined in different transformed cell lines. In transient expression, in primary BRK cells, the transcription from the H-2Kb promoter was not down regulated by the Ad12 13s mRNA product. These results suggest that the MHC class I expression was not directly repressed by ElA gene but was an indirect effect of cellular transformation. We continue to actively investigate this apparently controlled replication state of a "runaway replicon"

and will present analysis of individual cell replication states using in situ hybridization and other assays. The low level inapparent infection state strongly resembles that seen in undifferentiated EC cells but enhancer variants resistant to E1A have still not been found despite many attempts. We will also present results of our protein mapping analysis of this new E1A activity relative to specific conserved regions and to the presence or absence of binding sites for individual E1A associated cellular proteins such as p105 REI and p300. Human Papillomavirus 16 (HPV16) is a DNA tumor virus that contributes to the development and maintenance of ano enital carcinomas. It encodes two transforming proteins, E6 and E7. We previously observed that the H#V 16 E6 protein trans-activated the HSV thymidine kinase promoter. In this study, we sought to delimit the E6 responsive elements by usin mutated and deleted forms of the T K promoter.

Since deletions leaving only 2.5 bp upstream of the TAtA-box did not abolish the E6 responsiveness of the promoter, we investigated whether the TATA element was res onsible for the E6 trans-activating activity, and if functional distinctions could be found between different fATA sequences. Six minimal viral promoters with a slightly different TATA-box were inserted upstream of the chloramphenicol acetyltransferase (CAT) gene. These included the HSV TK promoter spanning from -50 to +51 relative to the transcri tion initiation site, the HPV 16 P97 from -72 to + 1.5, the Ad ML promoter from -44 to + 11, the HIV LTR : om -57 to + 85, the SV 40 early promoter from -140 to + 85, and the Ad E2P from -285 to +40.

Reporter lasmids were co-transfected in triplicate into NIH3T3 cells with a vector expressing E6 from the SV 40 ear& promoter or, as a negative control, the same plasmid containing the E6 ORF in an anti-sense orientation.

In the presence of E6, CAT activity was increased about 6 fold with all the viral promoters tested except with the Ad E2P whose activity was induced 13 fold. The same level of induction was observed at the RNA level.

These results show that E6 activates transcription from several minimal viral promoters which share only a TATA element in common, and that apparently different TATA types are equally effective for the E6 mediated trans-activation. Human papillomavirus type 16 DNA was found in three separate lesions within a female patient. The physical form of the viral DNA in each lesion was determined by two dimensional agarose gel electrophoresis. The primary cervical tumor (papillary squamous carcinoma) contained large amounts of episomal viral DNA as well as integrated HPV DNA. The episomal DNA was resolved into at least five forms equivalent to monomers and larger multimers of viral genomes. Metastatic tumor tissue found in the vagina had reduced levels of episomal DNA and a pattern of integrated DNA distinct from that of the primary tumor. The episomal DNA in the metastatic tissue w a s larger than dimers of the viral genome. The vulvar carcinoma in situ (intraepithelial neoplasia VIN 111) contained both free and integrated forms of viral DNA. The results suggest that both free and integrated forms of viral DNA can be found in tumor tissue and that metastatic cells contain further rearrangements of integrated DNA. Our work is consistent with the idea that mechanisms responsible for HPV 16 DNA integration may also be involved in maintaining episomal viral DNA within the tumor cell.

Galloway, Christine L. Halbert, G. William Demers, Fred Hutchinson Cancer Research Center, Seattle, WA 98104

To investigate the contribution of the E6 and E7 genes of human papillomavirus (HPV) types 6 and 16 in the immortalization of primary human epithelial foreskin cells (HFE), amphotropic retroviruses were constructed containing the genes singly or the contiguous E6/E7 region, and expression was driven by the retroviral LTR. Immortalization was achieved readily in cells containing the HPV 16 E6/E7 sequences and more rarely in cells containing only HPV 16 E7. Those cells transformed by 16 E7 alone contained multiple copies of proviral DNA and expressed high levels of E7 protein. Early after infection with the 16 E7 recombinant retrovirus, the cells demonstrated an altered phenotype in organotypic culture. Experiments are in progress to examine the state of the p53 gene product in immortalized cells which lack HPV 16 E6.

To determine whether the E6 gene of HPV 6 could augment the efficiency of immortalization of the HPV 16 E7 gene, co-infections were performed using viruses containing HPV 6 E6 + HPV 16 E7, HPV 16 E6 + HPV 6 E7, HPV 6 E6 + HPV 6 E7, and HPV 16 E6 + HPV 16 E7. These results will be presented. We have identified apparently specific isoforms of the product of the human cdc2 gene in adenovirus early region 1A (ElA) immune complexes. The activity is present in E1A immune complexes isolated from adenovirus infected cells and from cells that constitutively express the E1A gene. These isoforms of p34Cdc2 are also present in p60/Cyclin A immune complexes derived from cells that express E1A protein or do not. All papillomavirus (PV) eenomes contain a non coding region responsible for viral DNA replication and gene e ression. This region termed Upstream Regulatory Re ion (URR) harbors constitutive as well as inducibxpe enhancer elements, some of which are dependent on bV encoded E2 or, as it has been shown for HPV 18, E6 products. In this study, we sought to determine if the HPV 16 URR could respond to the HPV 16 E6 or the Adenovirus E l a proteins.The whole 800 bp URR was fractionated into 2 sub-fragments. CAT 1 and CAT 97 were generated by inserting respectively a 300 bp distal and a 500 b proximal URR fragments upstream of the TK promoter into the enhancer tester lasmid pBLCAT2. NIHgT3 cells were cotransfected with these reporter vectors and a plasmid e ressing g 6 from the SV 40 early romoter, or as a negative control, the same plasmid containing the E6 8 R F in an anti-sense orientation. $la 13s and 12s were e ressed from their own promoters. E6 activated pBLCAT2 10 fold. No further induction was observeTwith CAT 1 or CAT 97. E l a 12s failed to increase the basal activity of the 3 re orter plasmids. Ela 13s could induce 2 fold pBLCAT 2 and CAT 1, whereas it activated 15 fold the CAT 97Vector. To test if similar results could be observed on the HPV 16 homologous romoter, the whole URR comprising the p97 promoter (CAT 9), and a 5' 300 bp deleted URR (CA! 8) were inserted each in a promoter confi ration into pBLCAT 3. Co-transfections were performed into a human keratinocyte cell line ( H A g T ) .

No induction was observed on both reporter lasmids in the presence of E6, whereas CAT 8 could be activated several fold by Ela 13s. These resuis show that the HPV 16 URR contains an Ela inducible element located within a 500 bp proximal fiagment, but no E6 specific responsive element. Although a single E2P site was activated in response to E2, txe presence of two or more E2Ps resulted in synergistic (rather than additive) cooperative activation. We have therefore designed experiments to determine the contribution of E2 domains to cooperative activation, and to test whether the E2 protein cooperates with cellular factors. The 410 AA E2 rotein has a tripartite structure: a C-terminal DNA bindin /dimerization domain (DBD), an $terminal transcription activation domain (TAD), and a centraf "hinge" domain of -100 AA. Cotransfection experiments using E2 mutants indicated that the central "hinge" facilitates synergistic cooperation between E2P sites located at a distance from each other. To determine whether synergistic cooperation was limited to, and therefore a special function of the E2 transcription activation domain, we have tested the E2 protein for coo eration with cellular factors. The E2 activation domain was found to coo erate synergistically with t i e AP-1 complex of Jun/Fos and with a chimeric cJun TAD/GHF-1 D B g construction. Synergistic cooperation is thus not limited to multiple E2 dimers: the TAD of E2 can cooperate with those of cellular factors. Coexpression of the N-terminal acidic helix domains of E2 or cJun interfered with transactivation of either AP-1 or E2P sites by their cognate factors. Since this "s uelching" is thought to occur upon the de letion of a critical factor in the cell, we conclule that the transcriptional activation by the g 2 and Jun proteins involves at least one common protein target. romoters with multiple E2P sites. high or low risk based on their association with carcinomas. It has been shown previously that the E6 proteins of high risk HPV types (HPV-16, 18) form a stable complex with p53 in vitro, and that complex formation can lead to targeted degradation of p53 in an ATP and ubquitin-dependent reaction. Recent experiments have demonstrated that an additional cellular factor is required for E6-p53 complex formation. This factor forms a stable complex with HPV-16 or 18 E6 in the absence of p53, but not with p53 in the absence of €6. The factor is being purified from reticulocyte lysate, which contains large amounts of this factor. A bacterially produced HPV-I 6 €6 protein has been utilized for affinity purification of the factor. Additionally, ubiquitin affinity chromatography has been used to fractionate the enzymes involved in the ubiquitin-dependent proteolysis sytem in order to determine if the factor is directly involved in this system. We have molecularly cloned the HPV16 E7 gene into a bacterial expression vector driven by a thermoinducible bacteriophage lambda p~ promoter. Expression of the plasmid, designated p16E7, at permissive temperatures in E. coli resulted in an induced protein of approximately 15 kd. This 15 kd protein reacted with a monoclonal antibody to HPV16 E7 in Western blot analysis. We also constructed several derivatives of p16E7, each of which encoded a six residue peptide at the amino terminus of the E7 protein.

Each of the peptides contained amino acid residues capable of chelating metal ions, for use in subsequent purification of the protein by chelating peptide-immobilized metal ion affinity chromatography (CP-IMAC). The chelating peptides (CPs) varied in the number of histidine and tryptophan residues. The CP-E7 proteins were expressed in E.

colt expression levels varied among the constructs. This series of HPV16 E7 proteins containing various N-terminal CP extensions was purified using IMAC. The efficiency of purifying E7 with this series of chelating peptides was compared. One construct was used lo prepare E7 for further biochemical and functional analyses. The results of these analyses will be presented. We have been investigating biochemical properties of the bovine papillomavirus E l protein by site-directed mutagenesis. Mutations were made in sites conserved between E l and the SV40 large T antigen. Mutations between amino acids 100 and 113 have defined the boundaries of the nuclear localization signal for the E l protein as assayed by expressing mutant proteins from a heterologous promoter. We are presently testing the ability of the mutations to interfere with viral DNA replication when built back into the BPV genome. This region also contains a threonine analogous to Thr 124 in SV40 T Ag whose phosphorylation regulates the replication functions of T Ag. We are analyzing mutants at the coresponding Thr in E l to determine if it is phosphorylated and plays a similar regulatory role for BPV replication. Mutations in other regions of the protein are being used to analyze enzymatic activities of E l such as ATP hydrolysis. Interferons ( IFNs) exert potent effects on cellular physiology, including an inhibition of cellular proliferation, by aanscriptional activation of a set of genes in target cells. The molecular mechanisms for gene activation by IFNa have been well studied. IFNa binds to its cell surface receptor and activates one subunit (ISGF3a) of a mutimeric transcriptional activator (ISGF3). Following activation, ISGF3a translocates to the nucleus and forms a complex with the DNA-binding subunit of ISGF3 (designated ISGF3y). ISGF3 binds a specific promoter element (ISRE) in the regulatory region of the IFNa-stimulated genes and increases the transcription of these genes from undetectable to very abundant. This transcriptional activation is rapid and occurs independent of protein synthesis. Previous studies have demonstrated that adenovirus Ela protein can block the transcriptional stimulation of genes by IFNa. It is specifically the 12s form of Ela which causes this repression. We are investigating the effects of Ela on the localization, abundance, and activation of the subunits of ISGF3. In addition, the domains of the 12s Ela protein necessary for disruption of the IFN signal pathway are being defied by viral mutants of Ela. The E2 open reading frame (ORF) of BPV-1 is capable of encoding at least three polypeptides with transcriptional regulatory properties. The entire E2 ORF encodes a 410 amino acid protein that functions as a transcriptional transactivator and the 3' half of the ORF has been shown to encode two smaller polypeptides that repress E2-mediated transactivation. All three E2 proteins share a common carboxyterminal domain that encodes a specific DNA binding function and which contains sequences which promote E2 dimer formation. An amino terminal domain, which is unique to the full-length transactivator polypeptide, encodes the transcriptional activation function.

We have recently demonstrated that the BPV-1 E2 polypeptides are phosphorylated primarily on two serine residues at a site adjacent to the carboxy-terminal DNA binding domain, which is common to all three E2 proteins (McBride et a/., 1989. J.Virol. 63: 5076). These serine residues, at amino acid positions 298 and 301, were substituted with alanine residues and the mutated E2 ORFs substituted in the entire BPV-1 genome. The mutated BPV-I genomes were transfected into rodent cell lines and assayed for focus formation, viral gene expression and extrachromosomal viral DNA replication. The mutated viruses were able to transform C127 cells with wild-type efficiency. However, the viral genome containing the serine to alanine substitution at position 301 of the E2 polypeptide is present at very high copy number within these cells. The E7 proteins encoded by the human papillomaviruses (HPVs) associated with anogenital lesions share significant amino acid sequence homology with the Ad E1A proteins and the large T antigens of polyomaviruses. Each of these viral oncoproteins interacts with the product of the retinoblastoma susceptibility gene, pRB. Mutagenesis studies have shown that these conserved amino acid sequences are involved in binding pRB. The E7 proteins of the high risk HPV types 16 and 18 bind pRB with an affinity 5 to 10 fold higher than that of the E7 proteins of the low risk HPV types (HPV 6 and HPV 11). The biochemical and biological properties of a series of HPV-G/HPVlG E7 chimeric proteins were studied. The amino-terminal portion of the E7 protein determines the affinity to pRB, the ability to cooperate with an activated ras oncogene to transform baby rat kidney cells and the aberrant mobility of HPV-16 E7 on SDS polyacrylamide gels. Studies done in collaboration with J. A. Pietenpol and H. L. Moses (Vanderbilt University, Nashville TN) have revealed that this portion of the E7 sequence also governs the ability to abrogate the TGFp mediated repression of the c-myc promoter. The E7 proteins of the high risk and the low risk HPVs were able to transactivate the Ad E2 promoter with similar efficiencies indicating that this property must be mediated by a different mechanism. Minute virus of mice (MVM) is an autonomous parvovirus, one of a group of small single-stranded DNA viruses that are pathogenic for many vertebrate species, including man, and exhibit teratogenic, immunosuppressive, and tumor suppressive activity in vivo.

A series of mutations that only affect the small nonstructural protein (NS2) were constructed and introduced into the infectious clone of MVM(p). Analysis of these mutants indicates that NS2 participates in MVM DNA replication and is required for efficient viral growth in a cell-type specific manner. The majority of the NS2 mutants were severely defective for replication following transfection of normal host murine A9 fibroblasts; however, all were found to replicate more efficiently and produce infectious virus in certain other cell types such as human NB324K cells. The isolation of viral stocks from NB324K cells permitted a more detailed analysis of the mutant defect on A9 cells. NS2 mutant NS2-2018 was shown to be approximately 10-fold deficient for viral monomer replicative-form DNA production within a single burst cycle in infected A9 cells and produced a reduced amount of progeny single strand. Recent results indicate that NS2 plays a role in the regulation of the accumulation of spliced viral transcripts which encode the two viral nonstructural proteins.

Analyses of NS2 mutant DNA replication, transcript and protein accumulation will he presented. Virus infection induces the production of interferon in host cells probably by generating dsRNA as a byproduct of viral metabolism. Interferon induces the synthesis of DAI and other enzymes which comprise the cell's antiviral response. DAI, the dsRNA-activated inhibitor of protein synthesis, is a kinase which is activated by autophosphorylation and phosphorylates the initiation factor eIF-2 causing the cessation of protein synthesis. To preserve translation in the infected cell, adenovirus has elaborated a defense against this aspect of the antiviral response. The key viral product is a small (160 nt) RNA molecule, VA RNAI, which is produced abundantly i n cells at late times of infection. VA RNA is a highly structured molecule and competes with dsRNA for binding to DAI to counteract its activation by dsRNA.

We are studying the features of VA RNA that are important for its function. Structure analysis of VA RNAI suggested a complex stem-loop model structure which contains a n apical stem loop, a central domain (CD) and a terminal stem. Introduction of deletion and substitution mutations into VA RNA, defined two distinct regions: a duplex structure which is necessary for binding to DAI and the CD essential for blocking DAI activation. To further study the structure and function of the CD and to verify the suggested model we made a variety of mutations in this region. Changes in the CD altered the RNase protection pattern of this region but did not affect the structure of the apical stem-loop or the terminal stem. However, the analysis suggested the existence of tertiary interactions between the CD and a neighboring interior loop. Some mutants containing deletions and substitutions i n the CD are functional in kinase inhibition assays despite having altered CD structure. The data begin to define the structural requirement of the CD and suggest that an alternative conformation may be important for function.

Barnes', P. Raychaudhuri', K. Munger3, P.M. Howley3, and J.R. Nevins2. Div of Virology, Burroughs Wellcome Co., RTP, N.C. 27709', HHMI Duke Univ Med Ctr, Durham, N.C. 27710' and LTVB, NCI, Bethesda, Md. 2089Z1. The E7 protein of the human papillomaviruses is a multifunctional oncoprotein encoding transforming and transcriptional functions and has substantial amino acid sequence similarity to Ad E1A (Phelps et a 1 . u 53:539). The transcriptional activating function of E7 has been demonstrated using the Ad E2 promoter and recent analysis of a series of E2 promoter mutants reveals that the same sequences are required for activation by E7 and E1A. Despite these similarities, other promoters known to be E1A responsive are not activated by E7 including the Ad E3, Ad E4 and the HSP70 promoters: moreover, the amino acid sequences in E7 critical for transactivation of the E2 promoter do not include sequences homologous to the previously defined trans-activating domain 3 of E1A. Recent experiments suggest that the 125 E1A product independent of domain 3 , can activate the E2 promoter: however, neither the E3 nor E4 promoters is activated, similar to results with E7. Finally, biochemical studies have suggested a mechanism for this activation that involves the ability of E1A to dissociate complexes involving the E2F factor (Bagchi et al., Cell 62:659). We now find that the E7 protein can biochemically mimic the activity of the 125 E1A product. The transforming functions of adenovirus-5 E l a have been mapped to two regions which are highly conserved among Ela proteins (CR1 and CR2). CR2 is involved in binding to the product of the "recessive oncogene" Rb, and this feature is shared with transforming proteins of several other DNA viruses. 'The function of CR1 is less well characterized, although it appears to have a major role in trans-repression of transcription and also shows strnctural homology to papovavirus and papilloma virus transforming proteins. Although release of a block to cell replication via Rb binding can account for part of its oncogenic potential, other biological effects of E l a are similar to those produced by "dominant" oncogenes such as niyc. In order to investigate the relationship between E l a and rnyc transforming activities, we constructed Elalmyc chimeras to ask whether essential transforming domains of E l a could complement rnyc transforming domains. We found that the N-terminal domain of E l a could complement the C-terminal domain of vrnyc for immortalization of primary REFs and collaboration with EJ ras. Chimcras constructed using domains from transformation-defective mutants of either E l a (eg. ACRl) or rnyc were inactive, suggesting that both domains contribute to the function of the chimera. The reciprocal recombinant (niyclEla) also imniortalized primary REFs and collaborated with EJ ras. We were unable to transform or immortalize REFs by complementation of Ela and rnyc domains in trans, although this could reflect either a strict requirement for complementation in cis or inefficiency in transfection or trans-complementation. Although work by other groups has shown that any functional similarity between E l a and iiiyc probably is not due to any strnctural "homology", o w experiments with Elalrnyc chimeras sug gect that part of the transformation potential of Ela may arise through mimicry of nyc functions. Human papillomavirus type 6 (HPV 6 ) is the cause of the most prevalent viral sexually Abigail Farr, and He Wang, Department of Microbiology and Immunology and The transmitted disease, condyloma acuminata (genital warts). Such lesions are characterized by a) a change in viral gene expression as the target cell, the keratinocyte, differentiates, and b) increased cellular proliferation. We have been investigating both of these aspects using HPV6-W50, an HPV 6 genome cloned from a condyloma and verified to be an authentic representative of the genome present in the tissue. The entire long control region, LCR, of the genome and subfragments thereof have been cloned into pSVEcat, an enhancerless plasmid that contains the cat gene driven by the SV40 early promoter. Experiments in HeLa cells indicate that both silencer and enhancer elements are located in the LCR. Silencer activity is detected as the ability of a subfragment of the LCR, when inserted into pSVEcat, to decrease expression of CAT activity, and as an increase in CAT expression when this subfragment is removed from the entire LCR. The silencer activity is also detected in undifferentiated keratinocytes. Insertion of the silencer into pSV2cat, a plasmid containing the SV40 enhancer and promoter upstream of the cat gene, decreases CAT activity in both HeLa cells and keratinocytes. We propose that interactions with this silencer may change during cellular differentiation altering viral transcription and/or replication. In a second biological assay, the HPV6-W50 genome alone is sufficient to stimulate rat embryo fibroblast colony formation. Insertion of stop codons into individual viral open reading frames suggests that more than one gene is capable of stimulating cellular proliferation.

Institutes of Health, Bethesda, Maryland 20892. The HPV-16 9 7 promoter directs the transcription of the viral E6 and E7 transforming genes. In primary human epithelial cells, the BPV-1 E2 gene product can act as a transcriptional repressor of this promoter through interactions with E2 binding sites in the viral regulatory region. The mutation of the E2 binding site proximal to the 9 7 promoter relieves that repression, and the mutation of an additional binding site results in the transcriptional transactivation of the promoter (J. Virol.64: 2849, 1990) . In order to assess the effect of HPV-16 E2 in the context of the full viral genome, mutations were created either in the E2 binding sites upstream of the P97 promoter or in the E2 gene itself. These mutants were then examined in a quantitative primary human epithelial cell immortalization assay. Mutations of E2 or of its cognate sites adjacent to the P97 promoter resulted in an increase in the immortalization capacity of the HPV-16 genome, confirming that E2 repression of the P97 promoter occurs in vivo. We have also determined that the disruption of the El ORF of HPV-16 results in an increased immortalization efficiency. The effect of El and E2 mutations on DNA copy number and viral transforming gene expression are being examined. Overexpression 01 the E5 and E6 ORF alone, but not the E2 ORF. results in translormation in vitro. E2 ORF encodes a set of transcription faclors which interacl with the BPV-1 enhancer to regulate expression of the BPV-1 early genes. Although in vitro experiments failed to prove a direct involvment of E2 in BPV-1 transformation of cultured cells, the E2 proteins, as transcription factors, have a potential to perturb expression of cellular genes and may act synergistically with the E5 and E6 oncogenes to effect cellular transformation and tumrigenesis.

Seven independent lines of transgenic mice harbouring the E2 ORF fused to the p actin promoter (pCBAE2) were generated. About 10% of mice of four of transgenic lines develop focal hyperplastic growth in the upper portions of urethers. Polypoidal projections into the lumen of urethers were caused by abnormal proliferation of both the Stromal (fibroblastic) and epithelial (transitional epithelium of the urether) components of the urether wall. These hyperplastic changes were histologically benign in most of examined cases. pCBAE2 transcripts. correctly initiated at the p-actin cap site are detectable in several tissues with the highest levels of expression in the urinary tract, specifically bladders and urethers prior to development of overt macroor microscopical lesions. Therefore it appears that E2 gene products can induce abnormal proliferation of fibroblasts and/or epithelial cells.

E2 proteins and their DNA binding specificities are conserved among different members of papilloma virus family, which includes several viruses implicated in human skin and genital premalignant and malignant lesions. This conservation 01 E2 may be important not only in the context of viral functions, but also relevant to the oncogenic potential of human papilloma viruses. T h e papillomavirus E2 proteins a r e important regulatory factors which affect viral transcription, episomal replication and transformation. In the bovine papillomavirus type 1 (BPV-1 ), several viral promoters can be transactivated by E2 through E2 dependent enhancer elements located in the viral long control region (LCR), including promoters involved in E2 expression itself. Characterization of the elements involved in the regulation of the major E2 promoter, P,,,,, indicated that a binding site for the transcription factor S p l exists directly upstream of the P,,,, TATA box which is critical for the basal level of transcription from this promoter. Point mutations in this S p l site eliminated P,,,, promoter activity in transient expression assays for E2 and resulted in a loss of transforming activity when introduced into the full viral genome. These mutations also resulted in a reduction in the E2 transactivation potential consistent with a loss in the production of E2. RNA analysis from cell lines stably expressing these mutated viral genomes confirmed that the P,,,, promoter w a s nonfunctional.

Surprisingly, however the viral genome w a s present as a stable plasmid, although at reduced levels We have shown by promoter deletion analysis that sequences within the 98 base pair repeat downregulate the expression of the late promoter in glial cells.

We have localized this "silencer" to a region which contains a poly-dA tract adjacent to a pentanucleotide repeat sequence (PRS) AGGGAAGGGA.

This region, when present within a heterologous promoter, can also function as a negative regulator of gene expression.

Using gel shift and UV-crosslinking assays, we have found that the PRS interacts specifically with a nuclear protein of approximately 56 kDa (p56). We are presently purifying p56 to homogeneity.

Recent evidence suggests that p56 is capable of interacting with several factors present in brain nuclear extract. We will present experiments to determine whether p56, either alone or in concert with accessory factors, regulates the JCV late promoter in vitro. We suggest that unique interactions between p56 and other nearby regulatory factors may determine the cell type-specific expression of JCV in glial cells and function to temporally regulate early and late gene expression during the viral lytic cycle. Homologous, subgenomic fragments of the viral LCR and E6fE7 transforming genes of Human Papillomavirus (HPV) types 18 and 16 were amplified from several primary cervical, penile, and vulvar tumors and cloned into a pUC-18 derived vector.

When assayed by a quantitative transformation assay using primary human keratinocytes (Schlegel R. et al. EMBO J. 7(10):3181, 1988), the subgenomic regions of HPV-16 and HPV-18 exhibited transforming activities similar to that of the full-length, prototype HPV genomes.

More importantly, the HPV-18 LCR-E6-E7 region was approximately 10-50 fold more active than that of HPV-16. These studies demonstrate that the transforming activity differences previously observed between prototype HPV-16 and HPV-18 map to the LCR-E6-E7 region, and that individual and independent isolates of HPV-16 and HPV-18 exhibit consistent differences in transforming potential, even when isolated from different anatomic sites. Using an in vitro immunoprecipitation assay, the recA/El fusion protein has been shown to bind specifically to a single region on BPV DNA (nucleotides 7819-93 on the BPV map). Furthermore, recA/El can be covalently labeled with oxidi~ed-~~P-ATP, suggesting the presence of a specific ATP binding site. Demonstration of these two activities by the bacterially expressed product indicates that recA/El and similar fusions should be useful for characterizing the biochemical activities associated with authentic El protein(s).

The three dimensional structure of canine parvovirus and its functional implications. The three-dimensional atomic structure of the single-slranded DNA canine pawovirus ha8 been determined Phases for the observed smcture amplitudes were initially derived in pan from a spherical shell model of t i x virus and eventually extended to 3.6 A resolution. Icosahedral infectious virions contain 60 proteins that an W 2 or VP3 with same copies of VPl. while empty capsid8 contain primarily VP2 (the precursor of VP3) and VP1 The central structural motif of VP2 has rhe same topology (an eight-stranded mliparallcl &-barrel) as has bee1 found in many other icoaahedral RNA and DNA viruses. There is a 22A long spike on the three-fold axes. a 151 deep depression at the twofold axes OD the viral surface. By analogy with rlrinoviruses, the canyon may be t h d sire of receptor attachment. Residues related to the antigenic properties of the v i m are found on the threefolc spikes. Electron density dong the fivefold axes shows that some of the amino termini run Ira the exterior consistat with the observation that some VP2's in full particles can be cleaved by trypsin. The fivefold axes w surrounded by 811 unusual cylindrical structure formed by Five &ribbons. A substantial volume of elcctro~ density is shown to correspond to at least an DNA nucleotides or about 12% of the total genome. Polyomavirus enhancer, required for viral replication and iranscription, contains a small region,the a-domain. consisting of three cellular protein binding sites. Variants of the a-domain that were active for replication activation were tested for their ability to activate Py early gene transcription. The presence of a single binding site, PEA3, PEAl or PEA2 was sufficient to activate replication of an origin-containing, enhancer-deleted construct in an FM3A-derived. Tantigen expressing cell line(F0P). Combining the PEA3 and the PEA1 binding sites synergistically activated replication to high levels. Combining PEA1 and PEA3 or PEAl and PEA2 in the order opposite that found in Py resulted in repression of replication to levels below that of the monomers. To investigate the effect of these binding sites on transciption, a plasmid containing the luciferase coding sequence, and SV40 polyadenylation and small T-antigen splice sites, driven off the Py early promoter was constructed. Monomers, multimers and combinations of PEA3, PEAl and PEA2 are being tested for transciption activation activity in FOP, 3T6 and aza myoblast cells. Early results suggest that a single combination of PEA3 and PEAl is sufficient to activate transcription in FOP cells. The effect of reversing the order of the PEA3lPEAl and PEAllPEA2 combinations on transcription will be determined. The human papfllomavirus type-16 (HW-16) is one of the papfllomavlrus genotypes which is frequently associated with cervical carcinoma. In cervical carcinoma tissue and in cell lines derived from cervical carcinoma. two translational open reading frames of the HW-16 genome, E6 and E7, are found to be structurally intact and actively transcribed. implicating their gene products as agents in the development of this epithelial cancer. The E6 and E7 gene products have been characterized as immortalizing agents on the basis of Ln uitro assays in tissue culture cells. The E6 and E7 gene products are also known to interact with tumor suppressor gene products. p53 and retinoblastoma. respectively. To study the specific activities of the E6 and E7 gene products in a differentiating epithelial tissue fn uluo, we created transgenic mice canying a chimeric DNA fragment in which the E6 and E7 open reading frames are fused to the transcriptional regulatory signals for the UA crystallin gene. This promoter directs the expression of E6 and E7 specifically to the developing ocular lens, since it is activated before terminal differentiation of the lens epithelial cells.

Transgenic mouse lineages have been established from three founder animals. All transgenic mice developed bilateral microphthalmia which was overtly apparent from late embryonic stages onward. Hlstological examination of lenses from transgenic mice indicated abnormalities in fiber cell differentiation. similar to those which developed when other immortalizing gene products were expressed in the developing lens. Blochemical analyses are in progress to correlate the observed histopathology with onset of transgene expression and to evaluate at the biochemical level the perturbation in lens cell differentiation. Preliminary results indicate that the transgenes are expressed specifically in the ocular lens. Our results suggest that the primary combined effect of the HPV-16 E6 and E7 gene products on differentiating epithelia is to inhibit normal differentiation. Thls property may relate to the etiologic role of these genes in cervical carcinoma. l'hcre is little information concerning thc T cell response to human parainfluenza virus type 1 (hl'IV-1), despite the intention to producc a vaccine for this childhood infection. W e have analysed I'BL from adults and find that memory CD4' and CD8' T cell responses are readily detected in vitro. hPIV-1-specific prolifcratire responses, with stimulation indiccs of 5-10 fold over background, and IL-2 production by CD4' 1' cclls are readily demonstrable following incuhation for 1 week with hI'IV-1. Each of 18 adults and 2 children tested were found to rcspond. T cell lincs cultured for long periods contained class 11 HLArestricted, virus-specific cytotoxic activity, however, this is less evident in short-term cultures and appcars to Iic dependent on the form of virus used for rcstimulation. CTL which are class 1 HLA-restricted wcrc oliscwed in short-term culturcs, following 1-3 stimulations with virus, from each person tested and the frequency of circulating hPIV-1 specilic precursor CTL was similar to that rcported for other viral infections. We have investigated the spccificty for viral proteins of the hPIV-1 specific, €ILA-A2.1-restrictcd T cell response which from our initial stutiics appcars to be a strong responder type and since this allcle is comnion in the population. The small (15.5 kDa) HBV e antigen (HBeAg), encoded by the C gene, is present in the serum of infected patients when viral replication occurs. Two steps of processing am r e q u i d for its secretion. 'Ihe first one eliminates the signal peptide located at the N-terminus of the 25 kDa precursor, giving rise to a 22 kDa protein (P22). The second processing event remove the arginine-rich domain located at the C-terminus of the P22 intermediate. This cleavage is due to a cellular aspanyl protease localized inside secretory vesicles. To study the molecular details of this step, we have introduced non-sense mutations in the 3' exmmity of the HBV C gene. Adenovirus-based vectors containing the C gene or the mutated derivatives were transiently expressed in human cells and the C gene relatcd products analyzed by immunoprecipitation. The results show that the HBeAg secretion is dependent of the presence of a sequence of 15 amino acids located in the arginine-rich domain. These results suggest that this squence could contain a soning signal allowing the entry of HBeAg in secretory vesicles. We have also studied the influence of this sequence on the secretion of normally nonsecreted reporter gene products. . 9 (1990) 379-384) was applied to the functional test of MV matrix (M) genes cloned from brain cell RNA of patients w h o died of subacute sclerosing panencephalitis (SSPE). The functional M gene of the full length MV cDNA was exchanged with the M genes o f four SSPE cases. One of the genes tested gave rise to infectious virus, whereas only the first halves of the other three genes were functional. These results demonstrate that M gene inactivation is not obligatorily correlated with SSPE development. We are currently applying the same experimental approach to the functional test of F protein intracellular domains, which are often altered in SSPE cases.

Three other lines of research are in progress. First, to define the sequence requirements for paramyxoviral RNA editing, full length MV cDNAs mutated in the canonical mRNA editing signal were constructed, and will be tested for production of infectious virus. Second, preliminary results indicate that two full length MV cDNAs containing the framework for an additional (seventh) gene in the form of a duplication of an MV intercistronic sequence and additional unique restriction sites, produce infectious virus. Third, we are trying to substitute the MV persistently infected cells used t o provide helper function with cells transiently expressing only those MV proteins directly involved in replication or transcription. To locate sites modifying virulence in the enterovirus coxsackievirus B3 (CVB3), we generated recombinants of the 5'NTR and capsid encoding regions of cardiovirulent and attenuated CVB3 strains. Inoculation of C3H/HeJ weanling mice has demonstrated that an important site of CVB3induced murine cardiovirulence resides downstream of the 5'NTR. Intertypic recombinants with portions of the poliovirus 1 (PV1) 5'NTR substituted for that of CVB3 has shown that substitution of less than the entire 5'NTR can reduce the ability of the recombinant to replicate but that recombinants with the entire 5'NTR of PV1 are as viable as the parental CVB3 in HeLa cell culture. However, the replication of this PVl:CVB3 recombinant in C3H/HeJ mice is reduced compared the CVB3 progenitor. Substitution of a 3' region of PV1 (with part of the replicase encoding region and the 3'NTR) for the corresponding region of CVB3 severely attenuates replication of the recombinant in HeLa cell culture. Replication of this recombinant increases when 5'NTR sequences are from PV1 as well. The essential determinant of attenuation in the genome of Sabin live oral poliovirus vaccine (OPV) is the presence of uridine at position 472 rather than cytidine characteristic of wild type virus. Using a new highly sensitive method we have found that all lots of type 3 OPV contain some amount of C at position 472. The amount of 472-C was increased in lots which failed intraspinal neurovirulence test in monkeys. Passaging of type 3 OPV in vitro led to rapid accumulation of these mutants at a rate dependent on cell type and infection conditions. There were no changes at the position 2034 previously implicated in increased neurovirulence of some virus isolates or at several other positions which were reportedly changed in vaccine lots that failed monkey neurovirulence test. Our data emphasize the significance of the position 472 for the neurovirulence of type 3 OPV in the monkeys and suggest the use of this approach for evaluation of OPV lots and finding optimal conditions for manufacturing of safer vaccines. Human rhinoviruses (HRVs) are members of the picornavirus family and are the major causative agent of the common cold. Previous studies have demonstrated that 91 of the 102 known serotypes require a single cellular receptor, identified as the adhesion ligand ICAM-1, for attachment and subsequent infection. ICAM-1, which interacts with LFA-1 in the execution of immunological and inflammatory functions, is predicted to have 5 immunoglobulin-like domains. Using an in vitro transcriptionftranslation system, we have previously mapped the binding site of three MAbs that block virus attachment to the first 82 amino acid residues (domain 1) of ICAM-1. However, no virus binding was detected using the in vitro synthesized ICAM-1 fragments. sylation appears to play a role in virus binding since native deglycosylated receptor purified from HeLa cells also failed to bind virus. In order to study virus binding, we expressed modified ICAM-1 molecules in CEF cells using a RSV vector. Removal of each of the 4 glycosylation sites in domain 2 failed to compromise virus binding. To further map the region(s) of ICAM-1 required for virus binding, a synthetic gene was constructed that inserts 29 additional restriction sites into domains 1 & 2. Mutagenesis studies using the synthetic gene are in progress using the RSV expression system. An infectious, full-length, poliovirus cDNA clone was placed under control of the GAL1 0 promoter and transformed into yeast. Expression of the cDNA by growth on galactose resulted in production of a poliovirus-specific mRNA, but did not result in altered growth characteristics of the yeast. In fact, we find that yeast are unable to translate poliovirus RNA and that this effect is a consequence of the poliovirus 5-untranslated region (UTR). In cell-free translation assays, an activity present in yeast extracts inhibits the ability of HeLa extracts to translate RNAs containing the 5-UTR. This activity is concentration dependent and can be fractionated over DEAE Sephacel columns. These results suggest that yeast contain a translational inhibitor which does not allow efficient internal initiation of translation as mediated by the 5'-UTR of poliovirus RNA. Virus strains of the same serotype differ with regard to their ability to induce the lesions even though these tissues characteristically support luxuriant virus replication. Because the development of lesions in these zymogen organs is associated with the release of proteolytic enzyme autodigestion, the pathological process is triggered by an unidentified event. In our studies, CD8 and CD4 lymphocytes were depleted by the intraperitoneal inoculation of Mab before infection of adult male A/J mice with an encephalotropic strain of the encephalomyocarditis virus. Elimination of the CD4 cells prevented the development of lesions in the zymogen organs and reduced the severity of the associated encephalomyelopathy without dramatically affecting virus replication. Since disease in these animals develops 3-5 days after the inoculation of virus, traditional autoimmune processes would not appear to play an important role in the development of the disease. Accordingly, our research has focused on the question -what role do CD4 cells play in the pathogenesis of the infection and the development of lesions in target organs? The possible mechanism involved will be considered in this presentation. Stephan Schaefer, Maria Seifer Departments of Medical Microbiology and of Pharmacology*, University of Gottingen, D 3400 Germany. Transfection of SV40-immortalized non-tumorigenic fetal mouse hepatocytes (FMH202) with cloned dimeric HBV DNA results in highly tumorigenic hepatocyte clones while FMH cells transfected only by the selection marker remain non-tumorigenic (EMBO J 9:1137). Transfection of permanent mouse fibroblasts which do not contain SV40 T antigen (NIH 3T3 and LTK') by dimeric HBV DNA generates also clones with increased growth efficiency in soft agar. Transfection with an HBV DNA fragment containing enhancer I and X gene generates also FMH clones which grow in soft agar and induce nude mouse tumors, if the clones express much 17 kD X protein. Although the X clones express more X protein than the HBV clones, they do not grow as efficiently in soft agar and have a longer latency period before tumor development. X transfected 3T3 clones grow also significantly better than transfected control clones, but again not as good as the HBV transfectants. The data suggest that X protein of HBV is a viral oncogene -at least in vitro-but further factor@) of HBV contribute probably to oncogenicity. The nude mouse tumors of HBV transfected FMH clones showed constitutive expression of c-fos, induction of endogenous retrovid elements IAP and MoMLV and rearrangements of the integrated HBV DNA. X protein was found to transactivate the serum response element of the c-fos promoter in these cells. Such a transcription element is also found in the LTR's of retroviral element IAP and of MoMLV. Possibly, X protein initiates increased transcription of cellular protooncogenes and furthermore rearrangements by endogenous retrotransposons. To further study the structural requirements of the hydrophobic region of 3AB, we made a number of amino acid replacement, deletion, and insertion mutants aimed at decreasing its hydrophobic character. Changes of Thr-67 to Arg or Lys, change of Gly-74 to Glu, and mutants containing both amino acid changes resulted in non-viable clones, primarily due to an RNA synthesis defect. This defect could not be complemented by a polio-coxsackie recombinant virus, suggesting a cis acting function directed by this domain. More dramatic changes including the insertion of 9 hydrophobic amino acids between residues 80 and 81 and the deletion of residues 72 to 80 also resulted in non-viable cDNAs. These mutants have been further analyzed by in vino translation of RNA transcripts of the mutated clones in the presence of microsomal membranes to evaluate the association of the mutated polypeptides with membranes. Our results suggest a correlation between the hydrophobic character of polypeptide 3AB and its role in poliovirus RNA synthesis. Virus infected target cells express newly synthesized viral antigens which can associate with class I MHC molecules and sensitize these cells for lysis by specific cytolytic T lymphocytes (CTL). Antigens processed by this "endogenous" pathway appear to be generated by the cleavage of viral proteins into small peptides which can then associate with class I MHC molecules and be transported to the cell surface. In order to determine whether the position of an epitope within the native viral protein is important for antigen processing by the endogenous pathway, we have generated several hemagglutinin (HA) mutants of the influenza virus A/JAP/305/57 in which one immunodominant HA epitope (amino acid residues 202-221) has been repositioned within the HA protein.

To generate these mutants, the epitope was removed from the HA gene by site-directed mutagenesis and synthetic oligonucleotides encoding the epitope were inserted at six locations throughout the HA protein. The mutant HA gene was then inserted into a vaccinia virus expression vector and recombinant viruses were produced. Recombinant vaccinia viruses expressing the altered HA protein were used to sensitize target cells for lysis by specific CTZ clones in a standard 51Cr release assay. We found that all six classes of mutant viruses were able to sensitize target cells as efficiently as wild type influenza virus for lysis by speciftc CTL clones. These data suggest that residues flanking antigenic site recognized by class I MHC restricted T cells do not influence the processing of the newly synthesized antigen and the presentation of the processed antigen fragment to T cells. We are also examining the role of individual amino acid residues for the presentation of the epitope to T cells by expressing the minigene encoding the antigenic moeity inside the cytoplasm. HCV is a major etiological agent of non-A, non-B hepatitis worldwide. The molecular cloning of the virus and determination of the complete primary structure of the viral polyprotein revealed that HCV is related to human flavi-and animal pestiviruses and more distantly to plant poly-and carmovimses. We have determined the nucleotide sequence at the extreme 5'-and 3'-termini of the HCV genome. Our analyses of these sequences show: 1) The nucleotide sequence in the 5'-untranslated region is highly conserved among HCV isolates collected from five continents. This suggests tat there is a strong evolutionary constrain operating on this sequence, indicating a likely role for this sequence in viral replication. 2) Within this region there are blocks of nt sequence homology with pestiviruses, but not with other viruses. 3) The relative position of short ORFs present the 5'-untranslated region of the HCV genome is similar to that of the pestiviral genome. 4) In vitro translation of viral RNA is inefficient unless these ORFs are removed, indicating that the 5'-unstranslation region is inhibitory for translation. 5) RNAs truncated at both the 5'-and 3'-ends are found suggesting the presence of both subgenomic mRNA and defective viruses. 6 ) Poly A tails are present at multiple sites on 3-subgenomic RNAs. These data differentiate HCV from the flaviviruses and HCV appears to be substantially different from other known pestiviruses. These data are consistent with the assignment of HCV to a separate viral genus. Our data provide new insights into the organization of the HCV genome, that may have important ramifications regarding the replication strategy and evolution of the virus. 

Heidemarie Holzmann, Franz X. Heinz, Christian W. Mandl, Farshad Guirakhoo and Christian Kunz. Institute of Virology, University of Vienna, Vienna, Austria Seven variants of tick-borne encephalitis (TBE) virus were selected by the use of neutralizing monoclonal antibodies (MAbs). These mutants did not bind the selecting MAb and differed from the wild type sequence by only single amino acid substitutions, which were located at different sites in the envelope protein E of TEE virus. The mutants were compared with respect to their virulence characteristics in the mouse model. One of the mutants, which had an amino acid substitution from tyrosine to histidine at residue number 384, revealed a strongly reduced pathogenicity after peripheral inoculation of adult mice, but retained its capacity to replicate in the mice and to induce a high titered antibody response. Infection with the attenuated mutant resulted in resistance to challenge with virulent virus. Comparison of nonconservative amino acid substitutions present in other attenuated flaviviruses provided evidence that a structural element including residue 384 is an important determinant of flavivirus virulence in general. Biology, University of Utah School of Medicine, Salt Lake City, UT 84132 All picomaviruses generate their functional proteins by proteolytic processing of a large polyprotein which represents the single primary translation product of the viral genome. The cleavage pattern of the HAV polyprotein has been predicted by computer-assisted amino acid sequence alignments, but no viral proteins other than capsid proteins have been observed in infected cells, most likely due to slow HAV replication cycles and failure to inhibit host cell protein synthesis. Thus, no viral protease activities or cleavage sites in viral precursor proteins have been identified. We have utilized cell-free translation of 'I7 RNA polymerase transcripts of HAV cDNAs representing various portions of the P3 region to study the protease activity of the 3C region. Comparison of the mobilities of the translation products on SDS-PAGE, pulse-chase analyses of the various products and immunoprecipitation of the translation products with antisera specific for 2C, 3C and 3D proteins demonstrated that a 2C3ABCD precursor is first cleaved at the 2U3A junction, and this is most likely an autocatalytic, cis cleavage. Mutational analyses confmed that 3C sequences were responsible for the observed cleavages, and showed that the C-terminal region of 3C is important for protease activity. .5-kb RNA genomes differ at 12 nucleotide positions: 1 in the 5'-noncoding region (5'-NC), 1 each in nsP2 and nsP3 nonstructural protein genes, 6 (one silent) and 2 (one silent) in the E2 and El envelope glycoprotein genes, respectively, and 1 in the 3'-noncoding region. We have constructed a fullycharacterized, full-length cDNA clone of VEE TC-83 virus in the polylinker site of plasmid pUC18. Subsequently, TRD-specific cDNA regions were spliced into the TC-83 cDNA clone backbone to construct 18 TC-831TRD recombinant clones. Recombinant viruses derived by transfection of BHK-21 cells with genomic RNA transcribed from linearized plasmid were tested for virulence in mice. The genetic virulence factor in the TRDITC-83 virus pair is a constellation of genetic determinants dominated by the 5'-NC and a Thr-Val-Thr amino acid triad (TRD positions 120-192-296) in the E2 protein. E2 position 120 appears to be the most critical structural protein component of virulence. The TRD-specific 5'-NC appears to enhance the virulence potential of downstream TRD amino acid specificities.

TRhNSGENIC been investigated so far are not polyadenylated but capable of forming a stable secondary structure. They further include several highly conserved sequence elements that were proposed to mediate functions of the viral replication cycle.In contrast, the genome analyses of tick-borne flaviviruses revealed the occurance of two different types of 3'-nOn coding regions: The first type was found in some strains of tick-borne encephalitis (TBE) virus and in Powassan virus. It is similar to that of mosquito-borne flaviviruses; it is not polyadenylated and may form a 3'-secondary structure resembling that predicted for other flavivirus genomes. However, there is no significant primary sequence homology with mosquito-borne flaviviruses in this part of the genome. The second type of 3'-nOn coding regions was found in certain strains of TBE virus. It is much shorter, lacks a 3'-secondary structure, but carries a poly-A tail. Currently we aim to identify common tick-borne-specific sequence elements. Furthermore, it is investigated, whether the two types of 3'-termini can be correlated with different biological properties of the respective virus isolates.

Department of Medical Microbiology, University of Nijmegen, POB 9101, 6500 HB Nijmegen, The Netherlands Evidence is accumulating that enteroviruses can persist in chronic diseases as polymyositis, chronic myocarditis and end-stage dilated cardiomyopathy. Persistence may be caused by a defect in replication control. To study the association between enterovirus persistence and dilated cardiomyopathy more directly, we have developed a PCR assay based on homologous sequences in the 5' NCR of the enterovirus group. A specific amplification could be obtained from 60 of 66 different enterovirus types. No amplification was obtained from ECHO 16, 22, 23 and Coxsackie A 11, 17 and 24. The enteroviral classification of ECHO 22 and 23 has recently been doubted. ECHO 16 is now being cloned for further characterization. No information is available for Coxsackie A 11, 17 and 24, but from their growth pattern and from the appearant lack of the conserved regions in the 5 ' NCR, the question raises whether they are properly classified as well.

With this broadly enterovirus specific PCR assay we were able to detect enteroviral sequences in 4 out of 7 cardiac biopsy specimens from patients with dilated cardiomyopathy. By sequencing the amplified products we are currently investigating which enteroviruses are involved. proteins. The precore protein contains the entire sequence of the core protein plus an amino-terminal extension of twenty-nine amino acids. This amino-terminal extension contains a signal sequence for the secretion of the precore protein. Translocation of the precore protein across the endoplasmic reticulum (ER) membrane results in the cleavage of the signal sequence and the production of the precore protein derivative named P22. We now demonstrate that both P22 and the core protein are phosphoproteins. Microsomal fractionation and trypsin digestion experiments demonstrate that, although a fraction of phosphorylated P22 is located in the ER lumen, a substantial portion of it is trypsin-sensitive indicative of a cytosolic localization. Phosphorylation appears to occur in the carboxy terminus of P22, since the P22 derivative, P16, which lacks the carboxy terminus of P22, is not phosphorylated. Linking the carboxy terminus of the precore/core protein to heterologous cytosolic and secretory proteins led to the phosphorylation of the resulting chimeric proteins. This result suggests that phosphorylation of P22 and core protein is mediated by cellular kinases. The F protein is synthesized as an inactive precursor F, which is cleaved by one or more host-cell proteases into the active form of the protein which consists of the disulfide-linked F, and F, subunits. The cleavage site of the SV5 F protein consists of five arginine residues and is cleaved by a ubiquitous protease (or proteases) that is present in all cell types examined so far. We have used a variety of approaches including the use of inhibitors of intracellular transport and cell fractionation, followed by analysis of the extent of post-translational modification of the F,, F, and F, polypeptides, in order to determine the subcellular compartment in which cleavage of the SV5 F protein takes place. The data suggest that the cleavage event occurs in the trans-Golgi network. Many studies have pointed to the possible coupling of the vesicular stomatitis virus (VSV) transcription process to viral N S protein phosphorylation. We previously showed that at least two distinct protein kinase activities which remain bound to virion cores can phosphorylate the viral N S transcription factor in vitro (Beckes and Perrault, to be submitted). We show here that one of these ( V S V K I ) gives rise to the phosphorylated N S 1 species while the second (VSVK2) converts N S 1 to faster migrating species, collectively referred to here as N S 2 . Both kinases are active when virion cores transcribe their genome in vitro. However, N S I to N S 2 conversion is specifically inhibited by pre-treatment of virion cores with 5'-pfluorosulfonylbenzoyl adenosine (FSBA) with no effect on the transcription process. Likewise, addition of cell extracts inhibits the appearance of N S 2 without affecting transcription. The latter phenomenon is probably due to cellular phosphatases since extracts specifically remove phosphates from N S 2 following in vitro phosphorylation. We conclude that N S I to N S 2 conversion is not required concomitantly with the transcription process. The deoxyguanosine analogue, 2'-CDG, inhibits HBV replication by greater than 95% in the HBVproducing cell line, 2.2.15, as monitored by a decrease of secreted HBV DNA, HBV polymerase activity, and intracellular HBV DNA. Transcription of HBV RNA from chromosomally-integrated DNA was unaffected. The maximally effective concentration of 2-CDG was 2 2 0 ng/ml; 5 to 10 ng/ml produced = 50% inhibition; 5 2 ng/ml produced little detectable inhibition. Guanine is incorporated into both DNA and RNA from deoxyguanosine (dGuo) using the "salvage" pathway. Although the incorporation of guanine into DNA from 2'-CDG was several times that from dGuo, there was no detectable incorporation of guanine into RNA from 2'CDG.

There were similar rates of DNA incorporation of guanine from 2'CDG in 2.2.15 cells and in the parental cell line, HepG2, which does not contain HBV sequences. The DNA was analyzed to determine whether 2'-CDG acts as a chain terminator. Virtually all of the analogue was present at internal sites within the DNA, and there was no detectable exchange of radioactivity into other DNA nucleotides. The data indicated that although 2'-CDG could be incorporated into cellular DNA, this entry was not through the "salvage" pathway, but rather by de novo phosphorylation. The incorporation of 2'-CDG into DNA was not dependent on active HBV replication, or on the presence of HBV-specific proteins. The 2'-CDG did not act as a chain terminator of DNA synthesis. We are currently investigating whether 2'-CDG triphosphate can specifically inhibit the HBV polymerase. We acknowledge a generous gift of 2'-COG from J.A.Secrist and Y.F.Shealy of the Southern Research Institute. The work was supported by NIH grant CA34818. Borna disease virus (BDV) causes a rare neurological disease in horses and sheep. The virus has not been classified since neither an infectious particle nor a specific nucleic acid had previously been identified. Recently, infection with a BDV-like agent has been associated with schizophrenia in humans. Antibodies to novel BDV-specific antigens were found in sera and CSF of such patients. To identify the genome of BDV a subtractive cDNA expression library was constructed with poly A selected RNA from a BDV infected MDCK cell line. A clone (Be) was isolated that specifically hybridized to RNA isolated from BDVinfected brain tissue and BDV infected cell lines. This clone hybridized to four BDVspecific positive strand RNAs (10.5, 3 . 6 , 2.1 and 0.85 kb) and one negative strand RNA (10.5 kb) in BDV-infected rat brain. Nucleotide sequence analysis of the clone suggested it represented a full length mRNA which contained several open reading frames. In vitro transcription and translation of the clone resulted in the synthesis of the 14 and 24 kD BDV-specific proteins. The 24 kD protein, when translated in vitro from the clone, was recognized by antibodies in the sera of patients (3/7) with behavioral disorders. This BDV-specific clone will provide the means to isolate the other BDV-specific nucleic acids, and identify the virus responsible for Borna disease. In addition, the significance of BDV or a BDV related virus as a human pathogen can now be more directly examined.

Department of Immunology and Microbiology, Wayne State University, Detroit, MI 48201. A cloned line of persistently infected (PI) cells by a strain of the normally lytic, echovirus 6 has been established. defective viral particles. The present study was undertaken to determine whether replication of nonlytic viral RNA occurred after transfection. cells were transfected with RNA recovered from PI cell extracts and nonlytic viral particles produced by PI cells. Cytoplasmic RNA extracts were prepared at various times after transfection and examined for the presence of viral RNA. The viral RNA was detected by hybridization of Northern blots of cellular RNA extracts with cDNA probes of wild type, lytic echovirus 6 . cellular extracts at 48 hours after transfection. Replicate transfected cultures, which underwent 6 division cycles during a 66 day growth period retained viral RNA. RNA extracts from the transfected cells did not produce cytopathology or lytic virus. These In order to characterize the HCV structural proteins, we convtructed a vector for in virro transcription coutaining the iirst 906 codons of the HCV polyprotein linked to a synthetic p-globin 5' UTR. Translation of RNA produced by this vectior in rabbit reticulocyte lysate supplemented with canine microsomal membranes generated polypeptides of 11, IS, 35 and 72 kD. The gene order, proteolytic processing, glycosylation and meinbrane association of these polypeptides nerc anallzed by standard techniques. Based on these studies, the first 906 codons of the HCV genome encode Ihree structural protcins which are integrated into niicrosoinal membranes when expressed in virro: an 18 kD pre-core (prc-C), a 35 k l l envelope-l glycoprotein (El) and a 72 kD envelope-2 glycoprotein (E2). The l l k D polypeptide appear-tu represent a portion of the adjacent non-structural region. Expression of the HCV structural region using recombinant vaccinia viruses produced species of E l and E2 which were essentially identical to those produced by in v i m trailslation. Pre-C was not detected in this system, although its expression was inferred from the cleavage of the plyprotein to generate E l (and E2), and by truncation experiments. Analysis of the the subcellular localization of E l and F2 by indirect immunofluorescent antibody staining and subcellular fractionation showed that both El and E2 ere retained within the endoplasmic reticulum as core-glycosylated integral membrane proteins which lacked terminal sialic acid rcsidues. Experiments with E2 mutants demonstrated that at least a portion of E2 could be processed within the golgi (sialated) and secreted if the C-terminal anchor regiou was removed. These experiments suggest that if the El and E2 species produced in the reticulocpe and vaccinia systems accurately reflect the processing of the IICV envelope proteins in a natural infection, then the hepatotropism of HCV may arise through binding of its envclopc proteins to glycoprotein receptors such as the asialoglycoprotein receptor or mannose receptor. the middle and small surface antigens, while the much weaker preS1 promoter codes for an mRNA that is mainly translated into the large surface antigen. All three forms of the surface antigen, which are co-linear in their carboxy-terminal portions, are found in the virion envelope, but over-expression of the large antigen can prevent secretion of all three forms and cause cellular injury. differentially regulate the two promoters in hepatoma cells. preSl promoter is activated by an up-stream HNF-1 site, but only in cooperation with an adjacent Oct-1 site. promoter, which is activated by sequences within 70 bp up-stream of the major start sites. In addition, a region in the X gene acts as an enhancer that activates the S but not the preSl promoter. The X gene product, a transcriptional trans-activator, has no signficant effect on either promoter. promoter and the enhancer, and characterizing the cellular factors that bind to them.

Our results show that the Neither of these sites affects transcription from the down-stream S

We are further defining the cis-elements in the S

Diseases, The National Institutes of Health, Bethesda, M D 20892

Vaccinia virus late gene transcription is dependent on DNA replication and the expression of three transactivator genes. Nuclease S1 analysis showed that expression of the transactivator genes is controlled in a manner distinct from that of either the early or late class of genes.

responsible for directing transcription of the transactivator genes, promoter fragments were ligated to a reporter gene. CD4(178)-PE40 is a genetically engineered hybrid protein consisting of the HIV gpl20-binding region of human CD4 linked to the translocation and ADP-ribosylation domains of Pseudomonas aeruginosa exotoxin A. In previous experiments with human T-cell lines, we showed that the hybrid toxin kills HIV-I infected cells with high selectivity and potency, and inhibits HIV-1 spread. We now report that CD4(178)-PE40 selectively kills HIV-1 infected cell lines of the monocytehnacrophage lineage, and that it inhibits HIV-1 spread in primary T-cell and macrophage cultures. Furthermore, we observe strong synergistic activity between CD4( 178)-PE40 and reverse transcriptase inhibitors (AZT or ddl) in blocking HIV-1 spread. When both types of drugs are present in cultures of T-cell lines infected with HIV-1, virus spread is completely blocked, as judged by protection of the target cell population from virus-mediated killing and complete inhibition of virus production. When drug administration is terminated after an initial treatment period of several weeks, no infectious virus remains; quantitative PCR analysis indicates complete elimination of HIV DNA. The surviving population is CD4-positive and readily killed upon re-exposure to fresh HIV, indicating that the drug treatment does not merely select for HIV-resistant cells. We conclude that the combination treatment completely eliminates infectious HIV-1 from the culture. Control experiments indicate that these synergistic effects are due to selective killing of the HIV-infected cells by CD4(178)-PE40, rather than to a simple neutraliztion effect by the CD4 moiety of the hybrid toxin. These results highlight the potential value of therapeutic regimens involving combination of a virostatic drug which prevents infection of healthy susceptible cells in the population, plus an agent capable of killing those cells that are already infected. E.Norrby. Twentyfive 13 to 35 amino acids lonn peptides representing regions of human immunodeficiencv virus type 2 (HITT-?.), strain SBL6669, envelope proteins were evaluated for their immunogenic activity in guinea pigs. A number of the HIV-2 peptides were found to be capable of inducing strain SBL6669 neutralizing and antibody-dependent cellular cytotoxicity (ADCC) antibodies. TWO overlapping peptides covering the amino acid sequence 311-337 representing the central and carboxy terminal part of the V3 region showed the most pronounced capacity to induce neutralizing antibodies. Two additional regions in the gp 125 containing linear sites reacting with neutralizing antibodies were identified, aa 119-137 and 472-509. The transmembrane protein, gp 36, pf YIv-2 harbored two regions of importance for induction of neutralizing antibodies, aa 595-614 and 714-729. These findings pare the way for development of synthetic vaccines against HIV-2 and possibly also SIV infections. The capacity of such a product to induce protective immunity can be evaluated in macaque monkeys. Stanford, CA 94305. Very little is understood about the genetic control or mechanisms involved in CMV pathogenesis. The large 230 kilobase pair DNA genome of human and murine CMV encode over 200 gene products that regulate viral gene expression during viral growth as well as gene products presumably involved in virus/host interactions and latency. The murine CMV immediateearly gene 2 @2) encodes a 391 amino acid protein that is dispensable for growth in cell culture. In experimentally inoculated mice, k2deficient recombinant viruses do not exhibit significantly altered growth in target tissues such as the salivary gland, peritoneal cavity, liver, kidney, spleen, and lungs. We have recently demonstrated that the k2 gene product can function in transient gene expression assays as a transcriptional activator of the major immediate-early kl promoter as well as its own promoter. The murine and human CMV enhancer regions share many similar sequence motifs that may be important in the regulation of immediate-early gene expression. Preliminary results demonstrate that 1E2 can mediate transactivation through known human CMV enhancer repeat elements, suggesting that the murine CMV enhancer may be a target for IE2-mediated transactivation. Based on the predicted protein sequence, IE2 contains a putative leucine zipper and adjacent charged domains. A 4-3 repeat of hydrophobic residues occurs in this same region of the k2 sequence that may be important for formation of a coiled-coil structure. Interestingly, IE2 shares an overall amino acid similarity of 44% to Epstein-Barr virus BZLFI, a known transactivator of the EBV immediate-early promoters. IE2 also demonstrates 49% amino acid similarity to the bovine papillomavirus E2 transactivator as well as 47-5076 similarity to a HCMV protein family encoded for in the short (S) component of the viral genome, suggesting that these proteins may be functionally similar to MCMV IE2. Functional characterization of IE2 as a transactivating protein and murine and human CMV enhancer elements as targets for IE2-mediated transactivation should yield insights into murine CMV gene regulation and may be applicable to the overall understanding of human CMV pathogenesis. Wilcox and Johnson (1) have previously reported the infection in vino of primary fetal rat neurons with Herpes simplex virus type 1 (HSV-1) under conditions that lead to viral latency. as defined by the following operational criteria : i) lack of expression of the viral gene products that are detectable during a productive infection. and ii) reactivation of the virus by specific stimuli. Transcriptional studies show that neurons harboring a latent infection in vino express the latency-associated transcript (LAT); however these cells do not produce transcripts from other viral genes until the virus is reactivated by nerve growth factor (NGF) deprivation. Since viral expression restricted t o LAT transcription is widely accepted as a marker for latency, these data suggest that our system not only mimics latency at an operational level. but is relevant for mechanistic studies of latent infection by HSV-1.

The LAT contains two open reading frames. but so far a corresponding protein encoded by LAT has not been detected. Experiments w i l l be. described. in which we demonstrate by immunocylcchemical techniques that an antiserum directed agaimt a bacterially expressed fusion protein containing a pan of an LAT-encoded polypeptide recognizes an antigen specifically expressed in latently infected neurons cultured in virro. This protein was not detected in mock-infected neurons nor in productively infected Vero cells. Western blot analysis revealed a 42 kd protein present in latently infected neurons. m e significance of this novel latency-associated antigen, or LAA. w i l l be discussed. 60 pat'ents who had undergone allogeneic bone marrow transplantation for different haemtoloyical disorders were followed up for the development of HCMV infection using a highly specific and sensitive PCR technique in addition to conventional and rapid virus culture as well as hybridization and immunocytologlcal analysis. 43 of the patients analysed were found to be HCMV positive by the PCR assay, In more than 80% o f these patients the virus could be additionally detected by at least two of the other techniques described. PCR technique allowed detection of the virus more than two weeks prior to all the other techniques applied. A decrease in the leukocyte, lymphocyte count and particularly the absolute number of CD4t-T-cells after HCMV detection was associated with the development o f symptomatic HCMV infection as observed in 26 patients. In contrast in 17 patients HCMVpositive in two assays no symptoms related to HCMV infection were observed. Organ involvement of HCMV as clinically suspected due to an increase in the serum levels of the transaminases or the development of interstitial pneumonia and proven by demonstration o f the virus in the lung and liver by at least two of the techniques described was associated with marked immunohistological alterations on the cells infected by the virus, suggestive of an immunopathogenesis o f HCMV disease after allogeneic bone marrow transplantation. The pattern c.f expression correlates to the cellular phenotype. Four known and at. least one not yet identified promoter is involved in this expression. I.iess_ promoters are partly cell type specific, so that Wp, Zp and the LMP---gulatory sequence (LRS) are utilized in different cell types. Cellular transcription factors are likely to be important for this cell type specificity. We have also found that methylation patterns are speci,'ic in the promoter/enhancer regions of highly methylated NPC-and BL-cells (type I). The LRS in NPC i s an 800 bp unrnethylated island, surrounded by extensive CpC methylation, in tumors where LMP-expression can be detected. It is completely methylated in LMP non-expressors. We are now investigating whether Wp also shows an expression related methylation pattern. The ori P region with the EBER-coding genes i s always unmethylated (see F a l k et al., this abstract volume). We propose that this specific methylation pattern has a role in the viral gene regulation. The three dimensional and antigenic structure of poliovirus are well understood, and have allowed the construction of intertypic poliovirus antigen chimaeras exhibiting composite antigenic and immunogenic characteristics. ' We have exploited this observed flexibility in the sequences that can be accomodated at certain positions of the poliovirus particle to develop the Sabin 1 vaccine strain of poliovirus as an epitope expression vector2.

Known and predicted epitopes, of 12 to 18 residues in length, derived from the transmembrane and surface glycoproteins (gp41 and gp120) of HIV-1 have been engineered into antigenic site 1 of poliovirus, without abrogating virus viabilitya. The chimaeric virus particles have been characterised with regard to their antigenicity and their ability to induce neutralizing antibodies. Results obtained demonstrate that poliovirus chimaeras may have a role in HIV-1 epitope expression for diagnostic4 and vaccine development purposes. The promoter-proximal downstream sequences of genes transcribed by RNA polymerase I 1 have recently been recognized as containing transcriptional regulatory elements. I have identified and characterized a novel regulatory domain downstream from the human cytomegalovirus that is recognized at the DNA level. This regulatory domain was shown to enhance the number of functional initiation complexes without significantly altering the apparent elongation rate by RNA polymerase II transcription. I have constructed and analyzed a series of mutations in the first exon by in vivo and in vifro transcription assays. Run-off in vifro transcription and DNAbinding experiments (DNase I-footprinting and mobility shift assays) identified two downstream elements that specify the interaction of cellular transcription factors. One of these elements contains a reiterated sequence motif, present twice within exon I . The second element is an 18-bp sequence, located at approximately nucleotide position +33, that is conserved between species strains of CMV. The conserved element is recognized by two cellular nuclear proteins designated LTF A and B (for Leader Transcription Factor A and B) of 74 kDa and 50 kDa respectively as determined by photoactivated crosslinking. Screening a battery of different celltype nuclear extracts for LTF A and B activity suggests that these proteins are ubiquitous transcription factors. This study of promoter-proximal downstream transcriptional elements describes a unique DNA target regio n for regulating latency and reactivation of HCMV. Among potential cofactors, herpesviruses have received much attention because they are highly prevalent among HIV infected hosts. It was reported by us and others that HSV can activate HIVl gene expression, as determined by HIV LTR CAT expression and superinfection of acutely and chronically HIVl infected cell lines. The ACHZ cell line, derived from a chronically HIVl infected T lymphocytic line, CEM, produces few progeny virions, but is inducible to increase virion production by treatment with PMA and GM-CSF. Resting ACHZ cells have an aberrant pattern of viral RNA expression which is altered by treatment with phorbol esters to reproduce the RNA patterns of actively replicating virions. We have found that HIV expression dramatically increased following superinfection of ACH2 cells with HSV, as determined by a 10-fold rise in reverse transcriptase activity and p24 antigen expression. The HIV RNAs expressed in resting ACH2 cells consist mainly of subgenomic RNA species, with little or no genomic RNA. We found that within 12 hours of HSV superinfection, there was a dramatic rise in the 2 kilobase mRNA (encoding viral regulatory proteins), followed by a marked increase in the genomic mRNA. Such a shift of transcriptional phases recapitulates the early-to-late transition of the productive infection observed during a single step growth cycle of HIV and supports our previous observation that upregulation of HIVl expression occurs following HSV superinfection. These data also suggest that activation of HIVl by certain heterologous viruses occurs at the level of RNA expression. four months or the same treatment combined with four 2 week courses of interleukin 2 (IFN / 112) starting at week 3 of treatment. The interval between iL 2 courses was 2 weeks. IL 2 was given S.C. at 2.25 x 106 IU/m2 2 x per day for 2 days followed by 1.8 x 100 IU/m2/day for 12 days. The intervai between the end of the 1, treatment and stort of retreatment avaraged 13.4 months (range 6-25 months). Twelve patients with CAH-B each were randomized to each treatment arm HBV-DNA became negotive in 11 patients (7 on IFN, 4 on IFNIIL2). HBeAg became undectable in 4 patients (3 on IFN, 1 on IFN/iL2) at the end of treatment. Ail patients remained HBsAg positive. Side effects were frequent and reversible in both groups, however more severe in the IFN/IL2 group than in the IFN group. 112 had b e discontinued in 2 patients because of side effects. Severe local side effects (inflammatory reaction) occured only after lL2 opplication. The study shows that retreatment of patients with CAH-B previously unresponsive to IFN alpha was of low effectiveness. The addition of lL2 at the doses given and the schedule reported to IFN appeared to be of no benefit. We have developed a rapid, semiautomated amplified probe assay for HSV utilizing LCR technology. Traditional detection of HSV by virus culture and cytologic monitoring, which often take more than two days, has limited sensitivity, and depends on the presence of intact, culturable virus in test specimens. Our HSV LCR assay detects less than 100 HSV DNA molecules, and since the target region is short, the assay accommodates partially degraded specimens. A segment of the HSV genome approximately 50 bases in length is amplified by LCR and the product detected by an IMx@ microparticle immunoassay. Our HSV LCR system eliminates blunt-end ligation background, resulting in an extremely sensitive non-radioactive assay with excellent specificity. hours. Sensitivity compares favorably with that of PCR, and the HSV LCR assay readily discriminates HSVl from HSV2.

Presently, a batch of 24 specimens can be analyzed in less than three Poxviruses are a large group of complex DNA viruses that infect a variety of organisms and cause a wide range of disease pathologies. Poxviruses carry out their complete replicative cycles within the cytoplasm of infected cells and represent unique model systems for the study of numerous biological processes. Shope fibroma virus (SFV) is a tumorigenic leporipoxvirus which causes a localized benign fibroma in immunocompetent rabbits which regresses after two weeks due to a vigorous cell-mediated immune response. Malignant rabbit fibroma virus (MRV) and myxoma virus (MYX) are related tumorigenic leporipoxviruses which induce rapidly lethal systemic infections characterized by severe immunosuppression. Using a monoclonal antibody directed toward class I MHC proteins, we observe a rapid and specific decrease in the cell surface expression of class I MHC antigens in cell culture during infection with MRV or MYX. By 24 hours post infection, class I MHC antigens are nearly undetectable on the surface of infected cells. This downregulation requires late viral gene expression, and does not affect the cell surface levels of other glycoproteins, such as the transferrin receptor. Neither the benign SFV nor the orthopoxvirus vaccinia significantly decrease MHC antigen levels. An attenuated mutant of MYX, created by interuption of an ORF of unknown function designated as M-11, shows minininl downregulation of class I antigen expression. Therefore, decreases in cell surface class I MHC antigen expression during poxvirus infection correlate well with virus virulence. We are currently characterizing the nature of this downregulation at the genetic and molecular levels in order to assess the physiological relevance of these observations. Since many animal virus virulence factors act to attenuate host immune responses, these studies should further our understanding of the mechanisms underlying virus-induced immunopathology.

THE GLYCOPROTEIN GP116 OF HUMAN CYTOMEGALOVIRUS CONTAINS TWO Heidi Meyer and Michael Mach, Institut fur Klinische und Molekulare Virologie, Universitat Erlangen-Nurnberg, LoschgestraRe 7, D-8520 Erlangen, FRG.

The glycoprotein gp116 of human cytomegalovirus (HCMV) is a target for neutralizing antibodies. It is a component of the gCI-complex consisting of gp58 and gp116. Like its homologue glycoprotein gB of HSV I it posseses a highly antigenic region at the very aminoterminal part of the molecule between aa 27-84. In Western blot analyses with a series of procaryotic expression clones and human sera as well as a monoclonal antibody, we have defined two antibody binding sites. Site I (aa68-77) contains a neutralizing epitope and was recognized by human sera and the human monoclonal antibody C23. This region is conserved among all HCMV strains tested so far. Site I1 corresponds to aa 50-54, which is not conserved among the two HCMV laboratory strains of known sequence (AD169 and Towne). Using affinity chromatography we purified antibodies against site I1 from pooled human sera and investigated the biological function in vitro. Antibodies against site I1 did not neutralize HCMV in tissue culture. VIEek', Zbyn6k Kozmrk'and Martin Schwyze?; 'Dept. Human Genetics, Yale Univ. Sch. Med., New Haven, CT 06510; 'Inst. Molec. Genetics, Czechoslovak Acad. Sci., CS-16637 Prague, Czechoslovakia; 31nst. of Virology, Univ. Zurich, CH-8057 Zurich, Switzerland A segment (15 kbp) of the Pseudorabies virus (PRV) genome, comprising and encompassing its immediate-early (IE) gene, was sequenced. Several open reading frames (ORF) were found in it. One of them (ORF3) overlaps with the IE gene in the opposite orientation. It is in the same region where latency-related transcripts in opposite polarity t o the IE gene were reported. There is a potential promoter (P2) upstream of ORF3 containing binding motifs for several transcription factors. Specific binding of nuclear proteins t o four S p l sites was detected. Promoter P1 is upstream of the IE gene and it contains a TATA-box, CCAAT motif, and S p l and NF-WE1 binding sites. Four closely associated NF-WE1 and oct-I motifs were found further upstream from P I . These upstream elements were essential for efficient expression of PRV IE gene in vivo. It also increases efficiency of the HSVl t k promoter. Two potential enhancer elements (E3, E4) were identified in the sequenced region by an enhancer trap experiment. Linked to P I , E3 acted as an enhancer and E 4 as a silencer. The PRV IE gene product repressed transcription from its own promoter and activated the SV40 early promoter. The transactivating virion protein Vmw65 of H S V l had the opposite effect on these promoters. The regulatory elements of PRV will be compared with those of H l V l . Defective EBV (hetDNA), first identified in a subclone of the P3HR-1 cell line, was shown to be capable of disrupting EBV latency i n v i t r o , and to consist of a deleted and rearranged genome. The specific rearrangement which appears to be responsible for activation of replication is the abnormal juxtaposition of BamHI W and Z fragments which results in upregulation of ZEBRA protein (€81, Zta, BZLFl). ZEBRA is an immediate early gene product which binds the AP-1 binding site and transactivates the EBV productive cycle. Detection of the Barn W-Z rearrangement in clinical samples can be done using polymerase chain reaction (PCR) to frame the abnormal W-Z junction. W e have detected this rearrangement in 2 tissues containing replicating EBV; a) oral hairy leukoplakia, a nonmalignant lesion, and b) a thymic carcinoma. Identification of the Barn W-Z rearrangement i n v i v o suggested that hetDNA may play a biologically significant role similar to that shown i n v i t r o of replication activation. Western blot analysis using monospecific anti-ZEBRA serum has been performed on tissues containing EBV. We have found evidence of ZEBRA expression in two tissues thus far. Correlation with presence of hetDNA and with disease process will be discussed.

A SITE-SPECIFIC MUTATION IN THE EXTRACELLULAR DOMAIN OF HERPES SIMPLEX VIRUS 1 (HSV-I) GLYCOPROTEIN B (gB) AFFECTS ENTRY AND CELL-TO-CELL SPREAD. Lenore Pereira, David Navarro, Pedro Paz, and Ishtiaq Qadri. Division of Oral Biology, School of Dentistry, University of California San Francisco, San Francisco, CA 94143. HSV-1 gB is an envelope glycoprotein that is required for infectivity and promotes virion penetration into cells. To map the functional domains of this molecule, we constructed a mutant virus by cotransfecting intact viral DNA with plasmid DNA containing a gB gene with a site-directed mutation at cysteine 382. We selected a recombinant, RVCys-5, that failed to react with several monoclonal antibodies having virus-neutralizing activity. Analysis of this mutant showed that RVCys-5 forms plaques at the permissive temperature (37OC) but not at the nonpemissive temperature (39OC). This defect was overcome by treating the mutant-infected cells with polyethylene glycol. Analysis of the viral proteins made at 39OC in cells infected with a high multiplicity of RVCys-5 showed that synthesis of the major HSV-1 regulatory protein, 0.4, and the viral glycoproteins was comparable to that of the parental virus, with the exception of gB. Although the mutant gB formed stable dimers, they were partially glycosylated and remained sensitive to endoglycosidase H. Our results showed that mutant RVCys-5 specifies an underglycosylated form of gB with conformation-dependent changes and suggest that these altered properties affect enhy and cell-to-cell spread of virus.

Pitha and Waldemar Popik, The Johns Hopkins University School of Medicine Oncology Center, Baltimore, MD 21205 We have been studying the effect of interferon (IFN) on the HSV-1-mediated activation of HIV transcription, using HeLa cell lines containing an integrated tat defective HIV (dt4) and a cell line which contains both tat defective HIV and a plasmid in which the expression of human IFN-a, is directed by HIV-LTR (a,). In dt4 cells, HIV viral transcripts were detected as early as 2 hr post-induction. However, no HIV transcripts could be detected in a, cells infected with HSV-1; infection, however, led to IFN synthesis ( 5 0 to 100 U) in a, cells. Using isolated nuclei from dt4 and a1 cells, we found that the IFN-induced inhibition occured at the transcriptional level. We have previously shown that IE175 and IEllO gene products of HSV-1 were essential for the activation of HIV-LTR. The detailed analysis of the IE175 and IEllO mRNAs and proteins in dt4 and a, cells showed only marginal inhibition of IE175 and IEllO proteins in the a, cells. We found, however, a difference in the binding pattern of nuclear proteins from dt4 and a, cells to the enhancer region of HIV-LTR. While the extracts from HSV-1-infected dt4 cells formed the NF-KB complex, the extracts from a, cells formed several complexes with different mobilities. Thus IFN, possibly through induction of new transcriptional factors, seems to interfere with the formation of a functional NF-KB complex.

DELETION OF THE HSV-1 ICP22 GENE AFFECTS VIRAL HOST RANGE, Kimber L. Poffenberger, Pat Raichlen and Ronald C. Herman, SYNTEX RESEARCH, 3401 Hillview Ave., Palo Alto, CA 94304. The five immediate early (IE) proteins of HSV-1 are responsible for the progression to early and late gene expression during productive infection. Proteins ICP4 and ICP27 are essential; ICPO synergizes with ICP4; ICP47 is nonessential in tissue culture. Little is known about the function of the fifth IE protein, ICP22. Previous studies, including one with a virus having a deletion encompassing part of both ICP22 and the adjacent Us2 ORF, as well as another using an antisense oligonucleotide to ICP22, suggest that viral growth is inhibited in certain cell types when ICP22 expression is affected. We describe here the construction and characterization of an HSV-1 mutant (del22Z) from which only the coding region of the ICP22 gene has been removed and replaced by the bacterial LacZ gene.

This virus synthesized no detectable ICP22 transcript or protein product upon infection of tissue culture cells. The loss of ICP22 caused a decreased yield of progeny virus in all cell types tested, especially those of human origin. The normal cascade of HSV-1 gene expression was altered in del22Z-infected cells. There was prolonged expression of some early gene products and delayed appearance of some late gene products. Although the absence of ICP22 was partially overcome in some host cells, our results demonstrate that ICP22 enhanced virus production both in vitro and in vivo. Sciences, The Johns Hopkins Medical School, Baltimore, MD 21205 Epstein-Barr virus DNA replication during latency requires a cis-acting Ori-P (origin of plasmid replication) and a aans-acting viral nuclear protein EBNA-1 of 641 amino acids encoded by the BamH1-K fragment. A bacterial expressed C-terminal fragment of EBNA-1 protein binds as a d m r to 24 high-affinity DNA binding sites in On-P and to two low-affinity sites in the BamH1-Q fragment in the EBV genome. Ba131 nuclease deletion analysis defined the minimal DNA-binding domain between amino acid 493 to amino acid 584. A potential DNA recognition motif from amino acid 513 to amino acid 533, and a protein dimerization function from amino acid 554 to amino acid 584 have been dissected out using site-directed mutagenesis. EBNA-1 binding to a consensus DNA binding site bends the DNA and induces structural changes in the protein that causes it to become resistent to proteinase-K digestion . These physical interactions of EBNA-1 with the DNA may play a direct role in DNA replication. We have used an in vitro mRNA decay system (Ross and Kobs, 1986, J. Mol. Biol., 188: 579-593) to analyze the mechanism by which HSV-1 destabilizes mRNAs. Polysomes from uninfected human erythroleukemia cells were used as a source of mRNA. These polysomes were mixed 1:l in cell-free reactions with either polysomes or post-polysomal supernatant (S 130) from HSV-1-or mock-infected murine erythroleukemia cells. Normally stable mRNAs such as y-globin were degraded rapidly with extracts from infected cells. S130 from cells exposed for only 15 min to wild-type virus caused a four-fold decrease in y-globin mRNA stability in vitro. In contrast, S 130 from vhsl-, vhs ASma-, or mock-infected cells did not destabilize y-globin mRNA. The stability of normally unstable mRNAs, such as histone and c-myc, was not accelerated by S130 from infected cells. No mRNA destabilizing activity was detected in polysomes from infected cells. Experiments are in progress to characterize how the vhs protein and any other required factors induce mRNA decay and to delineate the pathway of host mRNA decay following infection.

Burkitt's lymphoma (BL) cells carrying EBV only express EBNA-1, while in vitro transformed lymphoblastoid cell lines (LCL) express EBNA 1-6 , as well as LMP 1 and 2. We analyzed two cell lines for methylation/ restriction enzyme sensitivity

an LCL established with EBV from Rael. Cellular and viral DNA's were cleaved with the methylation sensitive restriction enzyme (Hpa 11) and with it's methylation insensitive isoschizomer (MspI) followed by Southern blotting analyses. Rael was found to be extensively methylated and CBM1-Ral-Sto unmethylated

Belgium Dextran sulfates [prepared from dextran fractions with molecular weights (m.w.) ranging from 1,000 to 500,0001 are potent inhibitors of enveloped (i.e. retro-. herpes-, toga-, arena-, and rhabdo-) viruses. They are not active against non-enveloped viruses such as polio, Coxsackie or reovirus. Within the m.w. range of 3,000 to 500.000. no much variation is observed in the antiviral effects of the different dextran sulfate samples. However, the 1,000 m.w. sample behaves quite differently

HUMAN MYOCARDITIC HEART. Steven Tracy, Nora M. Chapman, Mark A. Pallansch, Beth-Ann Coller, Melinda A . Beck, Peter Kolbeck, and Yvonne DeCory-Woronoff, Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, NE 68198 and Centers for Disease Control, Enterovirus Laboratory, Atlanta, GA 30333. Enteroviruses can cause human myocarditis and can be molecularly detected in inflamed heart muscle. An enterovirus was isolated from both pre-mortem throat and post-mortem heart muscle of a 9-day old child with fulminant myocarditis. Anti-coxsackievirus B3 (CVB3) neutralizing antibody neutralized both viruses, but sequence analysis of the 5' non-translated region suggests identity with modern coxsackievirus B2 isolates. The heart isolate induced inflammatory heart disease in the mouse, whereas the throat isolate did not. Isolation of the enterovirus from myocarditic human heart, propagation in cell culture, demonstration of murine cardiovirulency, and virus re-isolation from murine heart is the first rigorous test of Koch's postulates for any enterovirus as an etiologic agent of human myocarditis, and validates the CVB3-induced murine myocarditis model for the human disease. Pending sequence analysis, this may be the first report of a naturally occurring genetic recombinant coxsackievirus. The genome of this virus is being cloned for sequencing and infectious chimeric genome analyses. i v e t . r a n s a c t t v a t i a n a 1 a c t i v i t y un b o t h v i r a l a n d cal:l.ular prumot.rrs.In cirriw t.o charactw-ize t h i s t r a n s a c i i . v a t i n i ) fnnc.t,:ion we t.nqinF?arrd two s e t e of pla::,msdf. preS2/S 3'' dele-Lion iniit:wts r e s p e c t i v e l y , W e fotiiid t.tia? t h e i--terminal wni.naat:id d i d n o t tienerate t h r t . r a n s a r t i v a t i n q d a p p e a r e d when 23 aminoacide more were removed, b a t h iin S pr oteins.. The c,Fiert o.f trurxateil pi,.eS/S protc:ins o c-myc regnl.atory sequences and the 1F'A rt+sponslv e m n t <lW) is concerL.t dependeini arid require?; eel ltilar pr w t p i r i IKjnass C vi.ty. Tmmunoiluoresren iiiimtinopi,ecipita.tion s t u d i e s d o nut s u p p u r t .the h y e s i s *.hat t h e ;~cqu;.sit.ton u t t.ha new p r o p e r t i e s by t h e t r u n c i r t e d preS/S p r o t e i n s :is due t o a di.ffei*ori t, sub c e 1 1 i.ir ar c omn at-t. i men-l R t : ELISA against peptides that correspond to certain regions of the HIV-1 envelope. A correlation between high neutralizing activity arid smng seroreactivity against the peptides wggested that the corresponding regions might be involved in neutralization. Pcptidcq representing these regions were also able to inhibit neuimliwtion mediated by swum from HN-1 antibudy positive individuals. Neutralizing antibodies to the novcl rcgion of gp41 were found to be cross-reactive whereas hey were found to be. sUair-specific for the V3 region of gp120. The rcsults indicate that the conserved B-cell cpitope of the HIV-I gp41 (&a. 647-671) elicits a virus neutralizing antibody response during natural infection in humans and may thereforc bc considered for inclusion in a vaccine against MIV-1.

Per A. Brolidenl,2, K Broliden2, L Akerblom?, J Rosen4, E Norrby2 and B Wahrenl. The importance of the dependencc on single amino acids in the V3 region of Mv-1 gp120 was evaluated for virus neutralization and antibody dependent cellular cytotoxicity (ADCC). Synthetic overlapping 15-mer peptides and a set doinission peptides covering a.a. 301-317 were uscd. Sera from 29 H W l infected individuals at different stages of disease were tested for neuualization, ADCC and specific IgG xactivity. Six HIV-1 neutralizing monoclonal antibodirs (Mabs) acted as controls. All MAbs reacted with aregon (a.a. 304-318) of gp120, previously shown to induce *wutralizing antibodies. The amino acids essential for reactivity were identified to be within the sequence GPGR (a.a. 312-315). The importance of this region lor occurrence of neutralizhg antibodies in infected humans was investigated using the same set of peptides. Out of 29 individuals, 21 were found to have neumlizing antibodizs in titers between 100-1ooO. Among the neuhlimtion-positive sera, 17/21 (81%) reac*:.d with a.a. 304-318, compared with only one of eight sera (13%) negative in neutralization. When any of the foul mino acids G, P, G or R were deleted, the sercmactivity of these neutralization-posiuve sera decreased considerably. The conserved sequence GPGR was therefore considered to be the most important fpr neuaalizarion in this region also in human sen One of the MAbs were able to mediate ADCC, indicating that the V3 region may be one important target for cytotoxicity. However, no correlation was found among the hitman sera ktween ADCC activity and IgG reactivity m a specific sequence within V3. Thus, the conserved sequence GPGR in the V3 region of gp120 is critical for virus ileutralization by human HlV-I specific antibodies. The two EBV-associated human malignancies Burkitt lymphoma (EL) and Nasopharingeal carcinoma (NPC) have a restricted expression of EBNA 2-6 and LMP in comparison with LCLs that express all EBV antigens. Expression of EBV antigens correlates with the expression of Bblast markers. In order to examine the heno pe dependent viral gene expression we have constructed a serie of somatic cell hybrids Arivedxy the fusion of EBV positive BLs and LCLs with non Bcell lines. Anal sis of EBV antigens in these hybrids showed that the expression of EBNA 2 to 6 is down regdated in the hybrids in parallel with the extinction of the B-cell specific markers CD 19, CD20, CD21, CD23 HLA class II antigens and lg, suggesting that the expression of EBNA 2 to 6 is dependent on the B-cell phenotype. EBNA-1 was the only EBV antigen regularly expressed in the hybrids indicating that EBNA-1 can be autonomously expressed. Our results rise question weather B-cell specific transacting factors play a role in the regulation of EBNA 2 to 6 and weather EBNA 1 can be transcribed independent from the big primary transcript that generates all EBNAs after alternative splicing.

TRANSACTIVATION OF THE MAJOR CAPSID PROTEIN GENE OF HERPES SIMPLEX VIRUS TYPE t REQUIRES PROMOTER B I N D I N G OF A NOVEL CELLULAR FACTOR, Robert M i l l e t t e , Shin Chen, Ronald Wobig, Rosemary Lown, and Robyn Hauser, Department of Biology, P o r t l a n d S t a t e Univers i t y and Department of Microbiology and Immunology, Oregon Health Sciences University, Portland, OR 91201The p r o r 1; class of l a t e genes of herpes simplex v i r u s type 1 (HSV-l), as exemplified by t h e malor capsid p r o t e i n (VP5) gene, a r e v i r t u a l l y s i l e n t i n uninfected cells. However, t h e s e genes a r e e f f e c t i v e l y t r a n s a c t i v a t e d by s u p e r i n f e c t i n g HSV-1 or by c o t r a n s f e c t e d immedi a t e e a r l y ( I E ) HSV-1 genes.Using Eight mutants were constructed in a plaque purified isolate of the NYCBH strain of vaccinia virus. Individual genes were altered by either deletion or insertion mutagenesis. Single gene mutations were generated in the hemagglutinin (HA-), thymidine kinase (TK-), small subunit of the ribonucleotide reductase (sRR-), Bam HI site in Hind I11 F (BamF-), host range (HR-) and both of the growth factor (GF-) genes. Two large deletion mutants, one 2 0 Kb to the left and one 1 2 Kb to the right of the host range gene, were generated through condensation reactions between vaccinia promoters ectopically inserted at the 3' end of the host range gene. A TKdeletion mutant in the WR strain was also constructed. All of the mutants were characterized in a mouse model with regard to intracranial and intranasal LD-50, replication in the brain, dissemination after a respiratory infection, growth and pock formation in the skin, immunogenicity and transmission to naive cage mates. The results from these experiments have provided unique insight into vaccinia pathogenesis and attenuation. Intertypic chimaeric polioviruses previously constructed were shown to possess composite antigenicity and dual immunogenicityl. This prompted the investigation of Sabin type 1 as a presentation vector for neutralizing epitopes of other pathogens including HIV-1 and human papilloma virus HPV162.3. We have extended this work f o r the expression of defined epitopes of human CD4, the receptor for HIV. The ability of antisera to these chimaeras to block HIV infection and syncytia formation will be discussed. Those chimaeras containing CD4 residues which are believed to be involved in binding HIV are being assessed for their ability to bind gp120. The authentic RNA sequence of Sabin 3 used in vaccine in the United States has been determined. The sequence of the Sabin 3 vaccine strain, called LED 3: is unique. A novel base discovered in LED 3 is contained in one of the viral capsid protein coding regions and predicts a new Sabin 3-specific amino acid change as compared to published sequences (Stanway et al., 1983; Toyoda et al., 1984) . The reversion is observed to occur after replication of vaccine virus in the gut of primary vaccinees and after replication in nervous tissues of neurovirulence test monkeys. Passage conditions have been identified that lead to the reversion of the LED 3 specific base in the capsid region as well as the reversion of the attenuated base to the wild type base at 4 7 2 . The observation that these two base positions are found to revert to wild type bases during passage suggests that there is a selective advantage for virus containing the wild type bases at these positions. HHV-6 is a recently isolated herpesvirus which is highly prevalent in adult populations. The virus is tropic for CD4 positive lymphocytes and persists in the peripheral blood of a majority of seropositive individuals. We have recently determined the restriction endonuclease map of a Ugandan isolate of HHV-6 (U1102). The genome consists of a largely unique sequence of 141 kb flanked by direct repeat units of 10 kb. This structure differs from that determined for the other human herpesviruses, including human cytomegalovirus (HCMV). Previous sequencing studies of HHV-6 U1102 have, however, demonstrated that both the arrangement of genes and the predicted amino acid sequence of their products closely resembles that of HCMV. We now present evidence, based on sequencing a further 34 kbp of selected regions of the genome, that the organization of coding and non-coding features of HHV-6 U1102 is collinear with that of the unique long component of HCMV. In particular, a single region of HHV-6 contains a local deficiency of CpG dinucleotides which identifies the major transactivating (immediate early, IE) genes in HCMV. The sequence of this region in HHV-6 contains a tandem array of repeats, each containing binding sites for cellular transcription factors NFkB and AP-2, which is likely to constitute the major IE enhancer. This region has been amplified using the polymerase chain reaction and the product shown to transactivate a range of heterologous promoters, including the HIV LTR. Evidence for the importance of this enzyme as a target in infected cells was demonstrated by glycan analysis of virion glycoproteins vhich showed the presence of high glucose glycopeptides.

6-0-butanoyl CAST (BUCAST) completely inhibited HIV qlycoprotain synthesis after acute infection and caused a major restriction in the cleavage of gp120 from gp160 in treated chronically infected cells. This a-glucosidase 1 inhibitor had potent effect on the growth of human cytomegalovirus which was also related to the reduced expression OP a virion associated glycoprotein. The critical dependence of both viruses on a-glucosidase 1 for the correct expression of glycoproteins suggests that this enzyme is a favourable antiviral target. HIV-LTR transactivation occurs following cell activation via PKC but not PKA-dependent pathways and by herpesvirus infection. We tested if CMV-IE could: 1) influence HIV-LTR transactivation by PKA-dependent stimuli, and 2) further amplify transactivation induced by PKC-dependent signals.HIV-LTR linked to a luciferase reporter gene was transiently transfected in a CD4+ T cell lymphoblastoid line (Jurkat clone) with or without CMV-IE 1-2 genes. CMV-IE alone transactivated the HIV-LTR(10-fold). Stimulation of PKA by Forskolin or a CAMP analog (8bromo-CAMP) did not transactivate HIV-LTR, but did so in the presence of CMV-IE (20-fold). Stimulation of PKC-dependent pathways by TNF and PMA increased HIV-LTR activity 20 and 100-fold, respectively in the presence of CMV-IE compared to 3 and 4 fold in its absence. Similar results were observed upon stimulation by cross-linked CD3 and CD28 antibodies. CMV-IE mediated amplification was independent of the NFkb binding motif within the HIV LTR. Furthermore, all the above stimuli were shown to transactivate the CMV promoter and to increase CMV-IE gene transcription.CMV-IE can mediate HIV-LTR transactivation by PKA-dependent stimuli and amplify PKC-dependent stimuli in T-lymphocytes. The oncogenic activity of Human Herpesvirus-6 (HHV-6) DNA (GS strain) in NIH 3T3 cells has been reported previously (Razzaque, Oncogene 5, 1365, 1990). In the current study, we confirm and extend this observation to include another strain, HHV-6 GD, which was isolated by Dr. J. Luka from peripheral blood lymphocytes of a patient with chronic fatigue syndrome. Restriction enzyme analysis and monoclonal antibody (to GS strain) reaction did not show any difference between GS and GD strains of HHV-6. Like GS strain, the genomic DNA from GD strain could also morphologically transform NIH 3T3 cells. Two focus-derived cell lines established from this DNA transfection produced tumors in nude mice when injected subcutaneously. The transforming activity of this strain was identical to that of HHV-6 GS strain. With polymerase chain reaction using primers corresponding to the major capsid protein gene of HHV-6, viral DNA sequences were detected in both the focusderived and tumor-derived cell lines. Further analysis on the DNA sequences in these transformants and their role in transformation will be discussed. The effect of the EBV BNLF-1 gene product (latent membrane protein 1, W 1 ) on homotypic adhesion of a murine B-cell, A/J-95 was compared to that of 12,13-phorbol dibutyrate (PDBu). Three independent clones of A/J-95 cells, expressing high levels of the LMPl protein after retroviral transfer of the BNLFl gene under the HSV-1 TK-gene promoter, displayed a 5-7-fold increase in homotypic adhesion compared to controls. The addition of PDBu to the parental A/J-95 cells resulted in a similar increase in homotypic adhesion. The increased adhesion of the cells could in both cases be inhibited almost completely with a monoclonal antibody to the CDlla subunit of the adhesion molecule CDlla/CDl8 (Leu-CAMa, LFA-1) but not with a monoclonal antibody to CD54 (ICAM-1). The high level constitutive cell surface expression of CDlla, CD18 and CD54 was not significantly altered in cells expressing LMP-1, suggesting that the increased homotypic adhesion observed was caused by functional activation rather than by a quantitative increase in the adhesion molecules. The present experiments validate previous observations of LMPl-mediated aggregation of human B-cells in a murine system, demonstrate that LMP1, like phorbol esters, activate the CDlla/ CD18 dependent homotypic adhesion between B-cells to a comparable degree, and suggest that LMP1, like phorbol esters, participates in intracellular signalling pathways to accomplish this end.

The transmembrane region of retoviruses is immunosuppressive. In order to produce a vaccine against HIV-1, with a superior immune response, the transmembrane region of HIV-I was replaced with that of the VSV G TM and CTregions. This rvv, vvE13 was compared to a rvv expressing gpl60,vvEl. The level of expression of gpl20-VSV G TM CT was lower than that of gp160 by immunoblot analysis and immunoprecipitation with anti gp160 sera for both proteins and anti VSV G CT for gpl20-VSV G TM CT and anti gp41 sera for gp160. vvE13 did not induce syncytia, whereas vvEl did induce syncytia in CD4+ cells. Both rvvs produced a cell surface protein recognized by antibody to gp120.To test the immune response of the two r w s mice were immunized. Antibody response and l y m p y t e proliferation were superior with wE13 as compared to vvEl. These results indicate that replacement of the HIV-1 transmembrane region gp41, by the VSV G TM CT regions can improve the immunogenicity of HN-1 rvvs, which may be of importance in vaccine development.