The sequence of the human genome

A 2.91-billion base pair (bp) consensus sequence of the euchromatic portion of the human genome was generated by the whole-genome shotgun sequencing method. The 14.8-billion bp DNA sequence was generated over 9 months from 27,271,853 high-quality sequence reads (5.11-fold coverage of the genome) from both ends of plasmid clones made from the DNA of five individuals. Two assembly strategies-a whole-genome assembly and a regional chromosome assembly-were used, each combining sequence data from Celera and the publicly funded genome effort. The public data were shredded into 550-bp segments to create a 2.9-fold coverage of those genome regions that had been sequenced, without including biases inherent in the cloning and assembly procedure used by the publicly funded group. This brought the effective coverage in the assemblies to eightfold, reducing the number and size of gaps in the final assembly over what would be obtained with 5.11-fold coverage. The two assembly strategies yielded very similar results that largely agree with independent mapping data. The assemblies effectively cover the euchromatic regions of the human chromosomes. More than 90% of the genome is in scaffold assemblies of 100,000 bp or more, and 25% of the genome is in scaffolds of 10 million bp or larger. Analysis of the genome sequence revealed 26,588 protein-encoding transcripts for which there was strong corroborating evidence and an additional approximately 12,000 computationally derived genes with mouse matches or other weak supporting evidence. Although gene-dense clusters are obvious, almost half the genes are dispersed in low G+C sequence separated by large tracts of apparently noncoding sequence. Only 1.1% of the genome is spanned by exons, whereas 24% is in introns, with 75% of the genome being intergenic DNA. Duplications of segmental blocks, ranging in size up to chromosomal lengths, are abundant throughout the genome and reveal a complex evolutionary history. Comparative genomic analysis indicates vertebrate expansions of genes associated with neuronal function, with tissue-specific developmental regulation, and with the hemostasis and immune systems. DNA sequence comparisons between the consensus sequence and publicly funded genome data provided locations of 2.1 million single-nucleotide polymorphisms (SNPs). A random pair of human haploid genomes differed at a rate of 1 bp per 1250 on average, but there was marked heterogeneity in the level of polymorphism across the genome. Less than 1% of all SNPs resulted in variation in proteins, but the task of determining which SNPs have functional consequences remains an open challenge.

The genome sequence of Drosophila melanogaster

The fly Drosophila melanogaster is one of the most intensively studied organisms in biology and serves as a model system for the investigation of many developmental and cellular processes common to higher eukaryotes, including humans. We have determined the nucleotide sequence of nearly all of the approximately 120-megabase euchromatic portion of the Drosophila genome using a whole-genome shotgun sequencing strategy supported by extensive clone-based sequence and a high-quality bacterial artificial chromosome physical map. Efforts are under way to close the remaining gaps; however, the sequence is of sufficient accuracy and contiguity to be declared substantially complete and to support an initial analysis of genome structure and preliminary gene annotation and interpretation. The genome encodes approximately 13,600 genes, somewhat fewer than the smaller Caenorhabditis elegans genome, but with comparable functional diversity.

Lytic growth of Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) in culture

Kaposi's sarcoma (KS) is the leading neoplasm of AIDS patients, and HIV infection is known to be a major risk factor for its development. However, KS can occur in the absence of HIV infection and the risk of KS development varies widely even among HIV-infected patients, with homosexual men with AIDS being 20 times more likely to develop KS than AIDS-afflicted children or hemophiliacs. These and other data strongly suggest that a sexually transmitted agent or co-factor may be involved in KS pathogenesis. Recently, DNA sequences corresponding to the genome of a novel member of the herpesvirus family have been identified within AIDS-KS biopsies, and several reports indicate that these sequences are also present in all forms of HIV-negative KS. These and other findings suggest this new agent, referred to as KS-associated herpesvirus (KSHV) or human herpesvirus 8 (HHV8), as a candidate for the putative etiologic cofactor. However, the role of this agent in KS remains hotly debated. Further progress in understanding its biology has been severely hampered by the lack of a cell culture system for virus growth. Here we report the development of a system for the lytic growth of this virus in a latently infected B cell line and present the first ultrastructural visualization of the virus. This system will facilitate the detailed study of the molecular biology of viral replication, the testing of antiviral drugs and the development of diagnostic tests for viral infection.

The broad-spectrum antiviral ribonucleoside ribavirin is an RNA virus mutagen

The ribonucleoside analog ribavirin (1-beta-D-ribofuranosyl-1,2, 4-triazole-3-carboxamide) shows antiviral activity against a variety of RNA viruses and is used in combination with interferon-alpha to treat hepatitis C virus infection. Here we show in vitro use of ribavirin triphosphate by a model viral RNA polymerase, poliovirus 3Dpol. Ribavirin incorporation is mutagenic, as it templates incorporation of cytidine and uridine with equal efficiency. Ribavirin reduces infectious poliovirus production to as little as 0. 00001% in cell culture. The antiviral activity of ribavirin correlates directly with its mutagenic activity. These data indicate that ribavirin forces the virus into 'error catastrophe'. Thus, mutagenic ribonucleosides may represent an important class of anti-RNA virus agents.

Kaposi's sarcoma-associated herpesvirus gene expression in endothelial (spindle) tumor cells

The recent discovery of DNA sequences of a new human herpesvirus in Kaposi's sarcoma (KS) has fueled speculation that this virus might cause KS. The mere presence, however, of a virus in a complex multicellular tumor like KS could just as well be construed as evidence of a passenger agent. We sought stronger evidence linking the KS-associated herpesvirus (KSHV) to tumor formation by using in situ hybridization to investigate the specificity, constancy, and timing of KSHV gene expression in KS tumor cells. Here we document expression of a 700-nucleotide viral RNA in every KS tumor examined, from the earliest histologically recognizable stage to advanced tumors in which the vast majority of identifiable spindle tumor cells contain this transcript. Two other KSHV RNAs were also detected in a smaller fraction of the tumor cells in all but the earliest lesion. These viral RNAs were expressed to relatively low levels in this subset; because one of these RNAs encodes a major viral capsid protein, these cells may be producing KSHV. We did not find these KSHV genes expressed in a variety of other tumors and proliferative processes, but we did detect viral gene expression in prostatic tissue, supporting a possible mechanism for sexual transmission of KSHV. The close relationship between KS and KSHV gene expression is consistent with the hypothesis that KSHV is directly involved in the etiology and pathogenesis of KS.

From ab initio quantum mechanics to molecular neurobiology: a cation-pi binding site in the nicotinic receptor

The nicotinic acetylcholine receptor is the prototype ligand-gated ion channel. A number of aromatic amino acids have been identified as contributing to the agonist binding site, suggesting that cation-pi interactions may be involved in binding the quaternary ammonium group of the agonist, acetylcholine. Here we show a compelling correlation between: (i) ab initio quantum mechanical predictions of cation-pi binding abilities and (ii) EC50 values for acetylcholine at the receptor for a series of tryptophan derivatives that were incorporated into the receptor by using the in vivo nonsense-suppression method for unnatural amino acid incorporation. Such a correlation is seen at one, and only one, of the aromatic residues-tryptophan-149 of the alpha subunit. This finding indicates that, on binding, the cationic, quaternary ammonium group of acetylcholine makes van der Waals contact with the indole side chain of alpha tryptophan-149, providing the most precise structural information to date on this receptor. Consistent with this model, a tethered quaternary ammonium group emanating from position alpha149 produces a constitutively active receptor.

Asymmetric localization of a mammalian numb homolog during mouse cortical neurogenesis

During Drosophila neurogenesis, differential segregation of Numb is necessary for daughter cells of asymmetric divisions to adopt distinct fates, at least partly by biasing the Notch-mediated cell-cell interaction. We have isolated a highly conserved mammalian homolog of Drosophila numb, m-numb. During mouse cortical neurogenesis, m-Numb is asymmetrically localized to the apical membrane of dividing ventricular neural progenitors. Depending upon the orientation of the cleavage plane, m-Numb may be distributed into one or both of the daughter cells. When expressed in Drosophila embryos, m-Numb is localized asymmetrically in dividing neural precursors and rescues the numb mutant phenotype. Furthermore, m-Numb can physically interact with mouse Notch1. We propose that some shared molecular mechanisms, both cell-intrinsic and cell-extrinsic, generate asymmetric cell divisions during neurogenesis of vertebrates and invertebrates.

Restricted expression of Kaposi sarcoma-associated herpesvirus (human herpesvirus 8) genes in Kaposi sarcoma

Kaposi sarcoma (KS) is the leading neoplasm of HIV-infected patients and is also found in several HIV-negative populations. Recently, DNA sequences from a novel herpesvirus, termed KS-associated herpesvirus (KSHV), or human herpesvirus 8 (HHV-8) have been identified within KS tissue from both HIV-positive and HIV-negative cases; infection with this agent has been proposed as a possible factor in the etiology or pathogenesis of the tumor. Here we have examined the pattern of KSHV/HHV-8 gene expression in KS and find it to be highly restricted. We identify and characterize two small transcripts that represent the bulk of the virus-specific RNA transcribed from over 120 kb of the KSHV genome in infected cells. One transcript is predicted to encode a small membrane protein; the other is an unusual polyadenylylated RNA that accumulates in the nucleus to high copy number. This pattern of viral gene expression suggests that most infected cells in KS are latently infected, with lytic viral replication likely restricted to a much smaller subpopulation of cells. These findings have implications for the therapeutic utility of currently available antiviral drugs targeted against the lytic replication cycle.

Activated antigen-specific CD8+ T cells persist in the lungs following recovery from respiratory virus infections

The poor correlation between cellular immunity to respiratory virus infections and the numbers of memory CD8(+) T cells in the secondary lymphoid organs suggests that there may be additional reservoirs of T cell memory to this class of infection. Here we identify a substantial population of Ag-specific T cells in the lung that persist for several months after recovery from an influenza or Sendai virus infection. These cells are present in high numbers in both the airways and lung parenchyma and can be distinguished from memory cell populations in the spleen and peripheral lymph nodes in terms of the relative frequencies among CD8(+) T cells, activation status, and kinetics of persistence. In addition, these cells are functional in terms of their ability to proliferate, express cytolytic activity, and secrete cytokines, although they do not express constitutive cytolytic activity. Adoptive transfer experiments demonstrated that the long-term establishment of activated T cells in the lung did not require infection in the lung by a pathogen carrying the inducing Ag. The kinetics of persistence of Ag-specific CD8(+) T cells in the lung suggests that they play a key role in protective cellular immunity to respiratory virus infections.

Description and clinical treatment of an early outbreak of severe acute respiratory syndrome (SARS) in Guangzhou, PR China

Severe acute respiratory syndrome (SARS), now known to be caused by a coronavirus, probably originated in Guangdong province in southern China in late 2002. The first major outbreak occurred in Guangzhou, the capital of Guangdong, between January and March 2003. This study reviews the clinical presentation, laboratory findings and response to four different treatment protocols. Case notes and laboratory findings were analysed and outcome measures were collected prospectively. The SARS outbreak in Guangdong province and the outbreak in Guangzhou associated with hospitals in the city are described, documenting clinical and laboratory features in a cohort of 190 patients randomly allocated to four treatment regimens. Patients were infected by close contact in either family or health-care settings, particularly following procedures likely to generate aerosols of respiratory secretions (e.g. administration of nebulized drugs and bronchoscopy). The earliest symptom was a high fever followed, in most patients, by dyspnoea, cough and myalgia, with 24 % of patients complaining of diarrhoea. The most frequent chest X-ray changes were patchy consolidation with progression to bilateral bronchopneumonia over 5-10 days. Thirty-six cases developed adult respiratory distress syndrome (ARDS), of whom 11 died. There was no response to antibiotics. The best response (no deaths) was seen in the group of 60 patients receiving early high-dose steroids and nasal CPAP (continuous airway positive pressure) ventilation; the other three treatment groups had significant mortality. Cross-infection to medical and nursing staff was completely prevented in one hospital by rigid adherence to barrier precautions during contact with infected patients. The use of rapid case identification and quarantine has controlled the outbreak in Guangzhou, in which more than 350 patients have been infected. Early administration of high-dose steroids and CPAP ventilation appears to offer the best supportive treatment with a reduced mortality compared with other treatment regimens.

Diagnosis and management of antibody-mediated rejection: current status and novel approaches

Advances in multimodal immunotherapy have significantly reduced acute rejection rates and substantially improved 1-year graft survival following renal transplantation. However, long-term (10-year) survival rates have stagnated over the past decade. Recent studies indicate that antibody-mediated rejection (ABMR) is among the most important barriers to improving long-term outcomes. Improved understanding of the roles of acute and chronic ABMR has evolved in recent years following major progress in the technical ability to detect and quantify recipient anti-HLA antibody production. Additionally, new knowledge of the immunobiology of B cells and plasma cells that pertains to allograft rejection and tolerance has emerged. Still, questions regarding the classification of ABMR, the precision of diagnostic approaches, and the efficacy of various strategies for managing affected patients abound. This review article provides an overview of current thinking and research surrounding the pathophysiology and diagnosis of ABMR, ABMR-related outcomes, ABMR prevention and treatment, as well as possible future directions in treatment.

DLX-1, DLX-2, and DLX-5 expression define distinct stages of basal forebrain differentiation

The homeobox genes in the Dlx family are required for differentiation of basal forebrain neurons and craniofacial morphogenesis. Herein, we studied the expression of Dlx-1, Dlx-2, and Dlx-5 RNA and protein in the mouse forebrain from embryonic day 10.5 (E10.5) to E12.5. We provide evidence that Dlx-2 is expressed before Dlx-1, which is expressed before Dlx-5. We also demonstrate that these genes are expressed in the same cells, which may explain why single mutants of the Dlx genes have mild phenotypes. The DLX proteins are localized primarily to the nucleus, although DLX-5 also can be found in the cytoplasm. During development, the fraction of Dlx-positive cells increases in the ventricular zone. Analysis of the distribution of DLX-1 and DLX-2 in M-phase cells suggests that these proteins are distributed symmetrically to daughter cells during mitosis. We propose that DLX-negative cells in the ventricular zone are specified progressively to become DLX-2-expressing cells during neurogenesis; as these cells differentiate, they go on to express DLX-1, DLX-5, and DLX-6. This process appears to be largely the same in all regions of the forebrain that express the Dlx genes. In the basal telencephalon, these DLX-positive cells differentiate into projection neurons of the striatum and pallidum as well as interneurons, some of which migrate to the cerebral cortex and the olfactory bulb.

The size and conformation of Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) DNA in infected cells and virions

The genome of a novel human herpesvirus has been detected in specimens of Kaposi's sarcoma (KS) and in several AIDS-related lymphoproliferative disorders. Here we examine the size and genomic conformation of the DNA of this virus (known as KS-associated herpesvirus or human herpesvirus 8) in latently and lytically infected cells and in virions. Pulsed-field gel electrophoresis of viral DNA shows that the viral genome is similar in size to those of other gammaherpesviruses (160 to 170 kb). As with Epstein-Barr virus, KS-associated herpesvirus DNA is stably maintained in latently infected B cells as episomal monomer circles and induction from latency is associated with the selective accumulation of linear genomic forms.

Nicotinic receptor binding site probed with unnatural amino acid incorporation in intact cells

The nonsense codon suppression method for unnatural amino acid incorporation has been applied to intact cells and combined with electrophysiological analysis to probe structure-function relations in the nicotinic acetylcholine receptor. Functional receptors were expressed in Xenopus oocytes when tyrosine and phenylalanine derivatives were incorporated at positions 93, 190, and 198 in the binding site of the alpha subunit. Subtle changes in the structure of an individual side chain produced readily detectable changes in the function of this large channel protein. At each position, distinct features of side chain structure dominated the dose-response relation, probably by governing the agonist-receptor binding.

Cationic liposome-mediated intravenous gene delivery

Systemic gene transfer provides new opportunities for the analysis of gene function and gene regulation in vivo, as well as for human gene therapy. We used the chloramphenicol acetyltransferase reporter gene to examine several parameters important for the development of efficient, cationic liposome-mediated, intravenous (IV) gene transfer in mice. We then demonstrated that this approach can produce high level expression of biologically important genes. Specifically, we assessed the relationship of expression vector design to the level of systemic gene expression produced, and compared transfection levels produced by intravenously injecting DNA alone versus DNA-liposome complexes. We found that both the position of the heterologous intron, and the promoter element used in the expression plasmid, significantly affected the level of systemic gene expression produced. Although intravenous injection of plasmid DNA alone transfected every tissue analyzed, liposome-mediated delivery was much more efficient. We also established that repeated i.v. injection of DNA-liposome complexes produced high level systemic transfection. The second injection of DNA-liposome complexes produced levels of gene expression at least as high as those following a single i.v. injection. Thus, unlike some viral vectors, a neutralizing host-immune response does not limit re-expression, following reinjection of DNA-liposome complexes. Finally, we showed that the expression vectors which produced the highest levels of chloramphenicol acetyltransferase reporter gene expression could also produce high level expression of two colony stimulating factor genes in mice. Specifically, i.v. injection of liposomes complexed to expression vectors into which we had inserted either the murine granulocyte-macrophage-colony stimulating factor cDNA or the human granulocyte-CSF cDNA, produced circulating levels of the corresponding colony stimulating factor gene product comparable to levels which have been shown previously to be both biologically and therapeutically significant.

Differential expression of mammalian Numb, Numblike and Notch1 suggests distinct roles during mouse cortical neurogenesis

During Drosophila neurogenesis, asymmetric cell divisions are achieved by differential segregation of Numb (d-Numb) into one of the daughter cells to cause a bias in the Notch mediated cell-cell interaction. We have isolated a second mammalian gene with significant sequence similarity to d-numb, mouse numblike. When expressed in dividing neural precursors in Drosophila, Numblike is symmetrically distributed in the cytoplasm, unlike endogenous d-Numb or expressed mouse Numb (m-Numb), both of which are asymmetrically localized to one half of the cell membrane. In d-numb loss-of-function mutant embryos, expression of Numblike allows both daughter cells of a neural precursor to adopt the fate of the cell that normally inherits d-Numb. In mice, numblike mRNA is preferentially expressed in adult and embryonic nervous system. In the developing neocortex, Numblike is expressed in postmitotic neurons in the cortical plate, but not in progenitors within the ventricular zone where m-Numb and Notch1 are expressed. We have also found that, in dividing cortical progenitors, Notch1 is distributed around the entire membrane, unlike m-Numb which is asymmetrically localized to the apical membrane. We propose that an interplay between cell-intrinsic mechanisms (executed by m-numb and numblike) and cell-extrinsic mechanisms (mediated by Notch1) may be involved in both progenitor cell proliferation and neuronal differentiation during mammalian cortical neurogenesis.

Proteoglycans mediate cationic liposome-DNA complex-based gene delivery in vitro and in vivo

The factors controlling cationic liposome-DNA complex (CLDC)-based gene transfer in cells and in animals are poorly understood. We found that cell surface heparin/heparan sulfate-bearing proteoglycans mediate CLDC-based gene transfer and expression both in cultured cells and following intravenous gene delivery into animals. CLDC did not transfect Raji cells, which lack proteoglycans, but did efficiently transfect Raji cells stably transfected with the proteoglycan, syndecan-1. Fucoidan, heparin, or dextran sulfate, all of which are highly anionic polysaccharides, each blocked CLDC-mediated transfection both in cultured cells and following intravenous injection into mice, but had no effect on transfection by either recombinant adenovirus infection or electroporation. Intravenous pretreatment of mice with heparinases, which specifically cleave heparan sulfate molecules from cell surface proteoglycans, blocked intravenous, CLDC-mediated transfection in mice, confirming that proteoglycans mediate CLDC gene delivery in vivo. Modulation of proteoglycan expression may prove useful in controlling the efficiency of, as well as targeting the sites of, CLDC-based gene transfer in animals.

Combined chemokine and cytokine gene transfer enhances antitumor immunity

The probability of producing a specific antitumor response should be increased by multiplying the number of T lymphocytes that encounter the malignant cells. We tested this prediction in a murine model, using a recently discovered T-cell chemokine, lymphotactin (Lptn). This chemokine increased tumor cell infiltration with CD4+ lymphocytes but generated little antitumor activity. Coexpression of the T-cell growth factor interleukin-2, however, greatly expanded the T lymphocytes attracted by Lptn, affording protection from the growth of established tumor in a CD4+ and CD8+ T cell-dependent manner. Lesser synergy was seen with GM-CSF. Hence coexpression of a T-cell chemokine and T-cell growth factor potentiates antitumor responses in vivo, suggesting a general strategy to improve cancer immunotherapy.

A novel mechanism to ensure terminal initiation by hepatitis C virus NS5B polymerase

Hepatitis C virus (HCV) nonstructural protein 5B (NS5B) RNA-dependent RNA polymerase (RdRp) has acquired a unique beta-hairpin in the thumb subdomain which protrudes toward the active site. We report here that this beta-hairpin plays an important role in positioning the 3' terminus of the viral RNA genome for correct initiation of replication. The presence of this beta-hairpin interferes with polymerase binding to preannealed double-stranded RNA (dsRNA) molecules and allows only the single-stranded 3' terminus of an RNA template to bind productively to the active site. We propose that this beta-hairpin may serve as a "gate" which prevents the 3' terminus of the template RNA from slipping through the active site and ensures initiation of replication from the terminus of the genome. This hypothesis is supported by the ability of a beta-hairpin deletion mutant that utilizes dsRNA substrates and initiates RNA synthesis internally. The proposed terminal initiation mechanism may represent a novel replication strategy adopted by HCV and related viruses.

Suppression of the malignant phenotype of human glioma cells by overexpression of manganese superoxide dismutase

Manganese superoxide dismutase (MnSOD) has been previously shown to suppress the malignant phenotype of human melanoma and breast cancer cells. To test the possible role of MnSOD in glioma malignancy, MnSOD was overexpressed in wild type human glioma U118 cells and subcloned U118-9 cells by transfection of human MnSOD cDNA. The MnSOD-transfected cell lines demonstrated expression of exogenous (plasmid) MnSOD mRNA, increase in MnSOD immunoreactive protein, and a three- to eightfold increase in MnSOD enzymatic activity. The MnSOD overexpressing cell lines became less malignant as demonstrated by requiring a higher serum concentration to grow in vitro and much slower tumor growth in nude mice than the parental and neo control cell lines. These findings further support the hypothesis that MnSOD may be a tumor suppressor gene in a wide variety of human tumors.

Mouse numb is an essential gene involved in cortical neurogenesis

During neurogenesis of the mammalian neocortex, neural progenitor cells divide to generate daughter cells that either become neurons or remain as progenitor cells. The mouse numb (m-numb) gene encodes a membrane-associated protein that is asymmetrically localized to the apical cell membrane of dividing cortical progenitor cells and may be segregated to only the apical daughter cell that has been suggested to remain as a progenitor cell. To examine m-numb function during neural development, we generated a loss-of-function mutant allele of m-numb. Mice homozygous for this mutation exhibit severe defects in cranial neural tube closure and precocious neuron production in the forebrain and die around embryonic day 11.5 (E11. 5). These findings suggest that m-numb is an essential gene that plays a role in promoting progenitor cell fate during cortical neurogenesis.

De novo initiation of RNA synthesis by hepatitis C virus nonstructural protein 5B polymerase

RNA-dependent RNA polymerase (RdRp) encoded by positive-strand RNA viruses is critical to the replication of viral RNA genome. Like other positive-strand RNA viruses, replication of hepatitis C virus (HCV) RNA is mediated through a negative-strand intermediate, which is generated through copying the positive-strand genomic RNA. Although it has been demonstrated that HCV NS5B alone can direct RNA replication through a copy-back primer at the 3' end, de novo initiation of RNA synthesis is likely to be the mode of RNA replication in infected cells. In this study, we demonstrate that a recombinant HCV NS5B protein has the ability to initiate de novo RNA synthesis in vitro. The NS5B used HCV 3' X-tail RNA (98 nucleotides) as the template to synthesize an RNA product of monomer size, which can be labeled by ¿gamma-(32)Pnucleoside triphosphate. The de novo initiation activity was further confirmed by using small synthetic RNAs ending with dideoxynucleotides at the 3' termini. In addition, HCV NS5B preferred GTP as the initiation nucleotide. The optimal conditions for the de novo initiation activity have been determined. Identification and characterization of the de novo priming or initiation activity by HCV NS5B provides an opportunity to screen for inhibitors that specifically target the initiation step.

A complex translational program generates multiple novel proteins from the latently expressed kaposin (K12) locus of Kaposi's sarcoma-associated herpesvirus

The most abundantly expressed latent transcripts encoded by the Kaposi's sarcoma (KS)-associated herpesvirus derive from the genomic region surrounding open reading frame (ORF) K12 (kaposin A). Here we show that these transcripts, initially described as limited to ORF K12 itself, more frequently encompass upstream sequences spanning two sets of 23-nucleotide GC-rich direct repeats (DRs) (DR1 and DR2). Although the DRs lack AUG codons and were previously presumed to be noncoding, a monoclonal antibody raised to infected cells detected multiple polypeptides encoded by this region. These proteins are expressed during latency and upon induction of lytic viral replication in both primary effusion lymphoma (PEL) cell lines and KS tumors. Biochemical and genetic analyses reveal that these proteins are derived from variant translational initiation at CUG codons. The predominant translation product in the PEL cell line BCBL-1 derives from the 5'-most CUG codon in the transcript, resulting in a protein (termed kaposin B) which is encoded largely by the repeats themselves and which does not include K12 sequences. Other non-AUG codons in alternate reading frames are also used at lower efficiency, including one that initiates translation of a DR-K12 fusion protein (kaposin C) that is predicted to sort to a different subcellular locale than kaposin B. Thus, the products of the K12 region, which is the most abundantly transcribed region in latency, are surprisingly complex and may encompass multiple biological functions.