The 'Human Immunodeficiency Virus Type 1 (HIV-1), Human Interaction Database', available through the National Library of Medicine at http://www.ncbi.nlm.nih.gov/genome/viruses/retroviruses/hiv-1/interactions, serves the scientific community exploring the discovery of novel HIV vaccine candidates and therapeutic targets. Each HIV-1 human protein interaction can be retrieved without restriction by web-based downloads and ftp protocols and includes: Reference Sequence (RefSeq) protein accession numbers, National Center for Biotechnology Information Gene identification numbers, brief descriptions of the interactions, searchable keywords for interactions and PubMed identification numbers (PMIDs) of journal articles describing the interactions. In addition to specific HIV-1 protein-human protein interactions, included are interaction effects upon HIV-1 replication resulting when individual human gene expression is blocked using siRNA. A total of 3142 human genes are described participating in 12,786 protein-protein interactions, along with 1316 replication interactions described for each of 1250 human genes identified using small interfering RNA (siRNA). Together the data identifies 4006 human genes involved in 14,102 interactions. With the inclusion of siRNA interactions we introduce a redesigned web interface to enhance viewing, filtering and downloading of the combined data set.

How retroviruses regulate the amount of RNA genome packaged into each virion has remained a long-standing question. Our previous study showed that most HIV-1 particles contain two copies of viral RNA, indicating that the number of genomes packaged is tightly regulated. In this report, we examine the mechanism that controls the number of RNA genomes encapsidated into HIV-1 particles. We hypothesize that HIV-1 regulates genome packaging by either the mass or copy number of the viral RNA. These two distinct mechanisms predict different outcomes when the genome size deviates significantly from that of wild type. Regulation by RNA mass would result in multiple copies of a small genome or one copy of a large genome being packaged, whereas regulation by copy number would result in two copies of a genome being packaged independent of size. To distinguish between these two hypotheses, we examined the packaging of viral RNA that was larger (≈17 kb) or smaller (≈3 kb) than that of wild-type HIV-1 (≈9 kb) and found that most particles packaged two copies of the viral genome regardless of whether they were 17 kb or 3 kb. Therefore, HIV-1 regulates RNA genome encapsidation not by the mass of RNA but by packaging two copies of RNA. To further explore the mechanism that governs this regulation, we examined the packaging of viral RNAs containing two packaging signals that can form intermolecular dimers or intramolecular dimers (self-dimers) and found that one self-dimer is packaged. Therefore, HIV-1 recognizes one dimeric RNA instead of two copies of RNA. Our findings reveal that dimeric RNA recognition is the key mechanism that regulates HIV-1 genome encapsidation and provide insights into a critical step in the generation of infectious viruses.

Human immunodeficiency virus type 2 (HIV-2) has been reported to have a distinct RNA packaging mechanism, referred to as cis packaging, in which Gag proteins package the RNA from which they were translated. We examined the progeny generated from dually infected cell lines that contain two HIV-2 proviruses, one with a wild-type gag/gag-pol and the other with a mutant gag that cannot express functional Gag/Gag-Pol. Viral titers and RNA analyses revealed that mutant viral RNAs can be packaged at efficiencies comparable to that of viral RNA from which wild-type Gag/Gag-Pol is translated. These results do not support the cis-packaging hypothesis but instead indicate that trans packaging is the major mechanism of HIV-2 RNA packaging. To further characterize the mechanisms of HIV-2 RNA packaging, we visualized HIV-2 RNA in individual particles by using fluorescent protein-tagged RNA-binding proteins that specifically recognize stem-loop motifs in the viral genomes, an assay termed single virion analysis. These studies revealed that >90% of the HIV-2 particles contained viral RNAs and that RNAs derived from different viruses were copackaged frequently. Furthermore, the frequencies of heterozygous particles in the viral population could be altered by changing a 6-nucleotide palindromic sequence at the 5'-untranslated region of the HIV-2 genome. This finding indicates that selection of copackaging RNA partners occurs prior to encapsidation and that HIV-2 Gag proteins primarily package one dimeric RNA rather than two monomeric RNAs. Additionally, single virion analyses demonstrated a similar RNA distribution in viral particles regardless of whether both viruses had a functional gag or one of the viruses had a nonfunctional gag, providing further support for the trans-packaging hypothesis. Together, these results revealed mechanisms of HIV-2 RNA packaging that are, contrary to previous studies, in many respects surprisingly similar to those of HIV-1.

Moloney leukemia virus 10 (MOV10) protein is a superfamily-1 RNA helicase, and it is also a component of the RNA-induced silencing complex. Recent studies have shown that MOV10 plays an active role in the RNA interference pathway. Here, we report that MOV10 inhibits retrovirus replication. When it was overexpressed in viral producer cells, MOV10 was able to reduce the infectivity of human immunodeficiency virus type 1 (HIV-1), simian immunodeficiency virus, and murine leukemia virus. Conversely, when MOV10 expression was reduced by small interfering RNAs, HIV-1 infectivity was increased. Consistently, silencing of MOV10 expression in a human T cell line enhanced HIV-1 replication. Furthermore, we found that MOV10 interacts with HIV-1 nucleocapsid protein in an RNA-dependent manner and is packaged into virions. It blocks HIV-1 replication at a postentry step. In addition, we also found that HIV-1 could suppress MOV10 protein expression to counteract this cellular resistance. All of these results indicate that MOV10 has a broad antiretroviral activity that can target a wide range of retroviruses, and it could be actively involved in host defense against retroviral infection.

Since the 1980s, the rapid progression of the HIV/AIDS pandemic has prompted a major international research effort. As a result, the current knowledge on HIV biology, its evolution, and origins [1–5] exceeds that of many, if not all, other viruses. One of the most important areas of HIV research is the detailed understanding of HIV replication. As with all viruses, HIV must exploit the host's cellular machinery and metabolism to copy its genetic material, synthesize viral proteins, and assemble new virions. The viral replication cycle is thus dependent on an intricate network of direct and indirect protein interactions: between the viral proteins, between the virus and the host, and ultimately between the various host proteins that constitute the subverted cellular systems. When we also take into account the host's immune response and intrinsic antiviral factors, there are clearly a large number of host–pathogen relationships that are important to our full understanding of HIV biology. However, until recently [6], this valuable information has remained 'locked' in the published literature, making it time-consuming to study by individual researchers and inaccessible to computational analysis, thus hindering the progress of research. Here we highlight new developments in the area of host–pathogen systems biology that in our opinion will provide helpful insights to the HIV-research community. A compendium of HIV-to-human protein interactions In order to collect and organize the existing knowledge of HIV–host interactions, the 'HIV-1-Human Protein Interaction Database' (HHPID) has been constructed from the published literature [6] and was made publically available at NCBI's website, http://www.ncbi.nlm.nih.gov/RefSeq/HIVInteractions/[7]. The data set currently is composed of 1435 human genes, encoding 1448 proteins that interact with HIV-1 and 2589 unique HIV-1 protein-to-host cell protein interactions, including both direct (e.g., protein binding) and indirect (e.g., gene regulation) interactions. These data were curated from about 3200 papers published between 1984 and 2007, with 58% of the interactions being linked to more than one publication. Although this data set includes information derived from a wide variety of experimental systems (mainly in-vitro studies and model systems) and undoubtedly contains some false positive interactions, it nonetheless constitutes an excellent catalogue of the knowledge available in the published scientific literature. In addition, the data set features detailed interaction definitions and descriptions, with links to supporting literature, and is cross-referenced to other resources at the NCBI, such as Entrez Gene. By using gene ontology annotations [8] to organise all human genes according to the biological processes in which they participate, we can visualize the HHPID data as a detailed map of viral perturbation of the host proteome (Fig. 1); see Ptak et al.[6] for a comprehensive analysis of HIV's relationship to the three branches of the gene ontology. Focussing on the 'biological process' aspect of the gene ontology reveals both how the various parts of the viral replication cycle are delegated to different genes (Fig. 1a) and the way in which HIV-1 gene products cooperate to target specific parts of the human cellular system (Fig. 1b). Comparing the targeted functional categories with information about the different types of interaction (Fig. 1c) highlights the coordinated viral manipulation of particular human processes. For example, Tat is associated with a disproportionately large number of interactions (Fig. 1a), many of which are binding interactions (Fig. 1c) with host proteins involved in DNA packaging (Fig. 1b).Fig. 1: HIV–host interactions visualized in the context of human protein function. The human protein function visualization was constructed by calculating the semantic distance [36,37] between each pair of human gene-products, based on the biological process branch of the gene ontology [8], and constructing a minimum spanning tree (MST) from the semantic distances and visualizing it as a network layout using LGL [38] (gray edges). Gene-products close to each other in the MST are close in semantic space and are therefore likely to be implicated in a common biological process. Note, only those human genes assigned an annotation within the relevant branch of the gene ontology are shown. This representation of host functional space is shown with: interactions according to HIV-1 protein. Human gene-products listed as interacting with HIV-1 are shown as square nodes, coloured according to interacting HIV-1 protein. Clusters of nodes of the same colour indicate coordinated manipulation of specific cellular systems by the same HIV-1 protein. Note that interactions with Tat and Env together make up almost 50% of this data set. Functions associated with HIV interactions. Genes annotated with selected high-level gene ontology terms that are significantly enriched in the HIV-interacting set of human gene-products (false discovery rate <1×10−10) are shown by coloured circles. The highlighted gene ontology categories indicate the main cellular processes that are subverted by the virus in the course of its life cycle. And, the type of HIV-1 interaction. The nature of the interactions between HIV-1 and human proteins is shown by the coloured squares. There is a clear tendency for interactions of the same type to cluster together in functional space (e.g., inhibition of host proteins and binding interactions).Figure 1 demonstrates how compilations of detailed interaction data such as HHPID can help us to gain insight into host–pathogen systems and how normal cellular processes are exploited by a pathogen. A good example of this type of analysis is given by Dyer et al.[9], who recently identified human protein functions commonly involved in infections by different pathogens. Although the data analyzed were heavily biased towards viruses (98% of interactions) and HIV–host interactions (78%), such groundbreaking work highlights the potential of this research area to elucidate the intimate relationship between host and pathogen. Learning HIV biology by data integration The availability of a library of known HIV–human interactions allows for a direct comparison of important new experimental results to existing knowledge about the virus [10,11]. For example, the HIV-dependency factors (HDFs) identified by Brass et al.[10] are a set of 273 proteins that HIV needs to survive in vitro in TZM-bl cells (a HeLa cell line engineered to express human CD4, CCR5, and CXCR4). Surprisingly, there is little overlap between the HDFs and the curated HIV interactions, with only 36 proteins belonging to both sets. On the basis of their associated gene ontology terms, we find the HDFs to be enriched for cellular functions related to mRNA transport, protein transport and lipoprotein biosynthesis, but many of the other functions that are over-represented for the HIV-interacting proteins in HHPID [6] are not significantly enriched in this data set. Clearly, this type of experimental data contains extremely useful information about the human proteins that are essential to a particular part of the viral life cycle. However, it is important to consider the possible limitations of HIV interactions determined in vitro, especially where some types of interactions required for viral propagation in vivo (e.g., with the immune system or cell surface) may be missed. Reasons for disagreement between different sets of experimentally determined interactions include the particular experimental protocol [11], cell line, and HIV strain used. In the instance of Brass et al.[10], the RNAi screen for HDFs was applied to cells that were already infected, resulting in relatively few Env-interacting proteins being reported as HDFs (18% of the proteins that belong to both sets were Env-interacting, as compared with 44% of all HIV-interacting proteins in HHPID). In addition, the HIV-1 strain used in this experiment was deficient in the viral proteins Nef and Vpu and had a truncated Vpr protein. This work is a good illustration of the caveats that must be attached to any compilation of interactions, such as HHPID. Careful comparison of the experimental details given in each individual publication will be necessary to reconcile conflicting results wherever they occur. Beyond this basic catalogue of observed interactions, the key to deriving new understanding from the HHPID lies in its integration with other data sources. For example, Bandyopadhyay et al.[12] analyzed human time-course gene expression data in the context of an integrated HIV–host protein interaction network to identify 'activated modules' of connected proteins that were significantly underexpressed or overexpressed at the different stages of the viral life cycle. This work permitted the identification of several human proteins whose expression levels were associated with either the early induction or the latent phase of viral replication. Further development of algorithms for the analysis of experimental data that take the underlying biological network into account is an area of much current research interest in the systems biology community [13–16]. Looking ahead The underlying aim of all HIV research is the development of effective clinical interventions to prevent infection or to stop the virus from replicating, hence preventing progression towards AIDS. HIV–host interaction data, as a representation of existing knowledge about HIV infection on a molecular level, will thus be invaluable to future research in this area. Combined with the potential of a systems biology approach to drug development [15,17,18], there are many reasons to believe that this information will soon be of immense importance for understanding infectious diseases in general, and in particular for the development of drugs and vaccines to counter viral replication or the pathogenic effects of infection. The high mutation and recombination rates of HIV result in rapid adaptations that can quickly suppress the effectiveness of drugs targeting viral proteins. This observation has led to research on disrupting HIV–host interactions with drugs that target host proteins directly [19], with the entry inhibitor maraviroc [20] being the first drug to receive marketing approval in Europe and the United States. Maraviroc binds to CCR5, the chemokine receptor used by HIV in conjunction with the CD4 receptor to enter host cells, thereby making CCR5 unavailable for HIV's envelope protein gp120. This is distinctly different to drugs binding to an HIV envelope protein, such as enfuvirtide [21], a fusion inhibitor that also prevents direct viral–host interaction. Combined with other antiviral drugs, the concurrent use of drugs targeting both viral and host proteins should permit highly effective treatment regimes restricting the evolution of drug resistance. The availability of a large set of HIV-interacting human proteins raises the possibility of identifying putative novel drug targets by in-silico methods [22]. For example, do any of the known HIV-interacting host proteins already represent drug targets? Figure 2 illustrates the complete set of human proteins that are reported to be targeted by FDA-approved drugs [23]. This visualization also demonstrates how large areas of the 'functional space' of human proteins remain unexploited by medical interventions. In contrast, certain classes of these drugs have targets that are over-represented in the set of HIV-interacting proteins. It is possible that some of these targets, either singly or in synergistic combinations [24], may prove to be effective antiviral agents. For example, the HMG-CoA reductase inhibitors (known as 'statins') have been shown to lower viral load in HIV-1-infected individuals by negatively regulating host Rho GTPase activity [25]. In addition, there are host proteins currently bound or down-regulated by HIV-1 that have potent antiviral properties, such as APOBEC3F/G and a specific tetherin (CD317, BST2), which are counteracted by the HIV-1 Vif and Vpu proteins, respectively [26–28]. Indeed, the Vif-ABOBEG3G interaction has recently been described as the target for a small molecule, RN-18 [29]. The future incorporation of additional data, such as known structure-activity relationships [30], predicted 'druggability' [31] or disease association [32] will add greatly to the predictive power of network analysis.Fig. 2: Visualization of the current 'druggable genome'. This is represented by the set of human gene-products that are known to be targeted by existing FDA-approved drugs [23] (large squares). Human gene-products listed as interacting with an HIV-1 protein in the HHPID are shown by small black squares. Coloured squares show the targets of selected categories of drugs that are significantly enriched for HIV-interacting proteins (false discovery rate <1 × 10−10). See legend to Fig. 1 for further details.In conclusion, the compilation of a curated set of experimentally verified human–HIV interactions will enable new possibilities for using computational methods in the study of basic HIV biology and in drug and vaccine development. It permits the construction and visualization of the essential map of HIV's subversion of our biology (pro-pathogen interactions), that is, those promoting HIV's replication cycle and our reaction to HIV infection (prohost interactions). Detailed characterization of the relationships between these interactions and such factors as gene expression patterns, RNA interference, cell type, disease stage, propensity to progress to AIDS, etc., will lead to improved understanding of the conflict between HIV and the human host. Combined with knowledge gained from genomic data taken from cohorts of HIV patients, this type of information is starting to reveal how the genetic differences between HIV-infected individuals can affect their progression to AIDS [33,34], knowledge that in turn can reveal new strategies for inhibiting viral replication [35]. Thus, a systems biology perspective focusing on data integration and a detailed understanding of the complex interplay between virus and host could be key to developing new therapeutics that obviate an infection's pathogenic outcome. Acknowledgement J.E.D is supported by a Wellcome Trust studentship and J.W.P by a BBSRC project grant (BB/C515412/1). R.G.P, W.F. and BES-B were funded by Federal funds from the NIH, NIAID under Contracts N01-AI-05415 and N01-AI-70042 and wish to thank Dr Roger Miller, Project Officer, DAIDS.

The ‘Human Immunodeficiency Virus Type 1 (HIV-1), Human Protein Interaction Database’, available through the National Library of Medicine at www.ncbi.nlm.nih.gov/RefSeq/HIVInteractions, was created to catalog all interactions between HIV-1 and human proteins published in the peer-reviewed literature. The database serves the scientific community exploring the discovery of novel HIV vaccine candidates and therapeutic targets. To facilitate this discovery approach, the following information for each HIV-1 human protein interaction is provided and can be retrieved without restriction by web-based downloads and ftp protocols: Reference Sequence (RefSeq) protein accession numbers, Entrez Gene identification numbers, brief descriptions of the interactions, searchable keywords for interactions and PubMed identification numbers (PMIDs) of journal articles describing the interactions. Currently, 2589 unique HIV-1 to human protein interactions and 5135 brief descriptions of the interactions, with a total of 14 312 PMID references to the original articles reporting the interactions, are stored in this growing database. In addition, all protein–protein interactions documented in the database are integrated into Entrez Gene records and listed in the ‘HIV-1 protein interactions’ section of Entrez Gene reports. The database is also tightly linked to other databases through Entrez Gene, enabling users to search for an abundance of information related to HIV pathogenesis and replication.

Although many interactions between HIV-1 and human proteins have been reported in the scientific literature, no publicly accessible source for efficiently reviewing this information was available. Therefore, a project was initiated in an attempt to catalogue all published interactions between HIV-1 and human proteins. HIV-related articles in PubMed were used to develop a database containing names, Entrez GeneIDs, and RefSeq protein accession numbers of interacting proteins. Furthermore, brief descriptions of the interactions, PubMed identification numbers of articles describing the interactions, and keywords for searching the interactions were incorporated. Over 100,000 articles were reviewed, resulting in the identification of 1448 human proteins that interact with HIV-1 comprising 2589 unique HIV-1-to-human protein interactions. Preliminary analysis of the extracted data indicates 32% were direct physical interactions (e.g., binding) and 68% were indirect interactions (e.g., upregulation through activation of signaling pathways). Interestingly, 37% of human proteins in the database were found to interact with more than one HIV-1 protein. For example, the signaling protein mitogen-activated protein kinase 1 has a surprising range of interactions with 10 different HIV-1 proteins. Moreover, large numbers of interactions were published for the HIV-1 regulatory protein Tat and envelope proteins: 30% and 33% of total interactions identified, respectively. The database is accessible at http://www.ncbi.nlm.nih.gov/RefSeq/HIVInteractions/ and is cross-linked to other National Center for Biotechnology Information databases and programs via Entrez Gene. This database represents a unique and continuously updated scientific resource for understanding HIV-1 replication and pathogenesis to assist in accelerating the development of effective therapeutic and vaccine interventions.

Lactoferrin (Lf) is a multifunctional glycoprotein that plays an important role in immune regulation and defence mechanisms against bacteria, fungi and viruses. Bovine lactoferrin (bLf) has been recognized as a potent inhibitor of human herpetic viruses, such as cytomegalovirus and herpes simplex virus type 1 and 2. BLf has been found to prevent viral infection by binding to heparan sulphate containing proteoglycans that also act as cell receptors for herpetic viruses.In this study we further investigated the inhibiting activity of bLf against herpes simplex virus type 1 (HSV-1) in Green Monkey Kidney (GMK) cells and found that, in addition to the viral adsorption step, bLf also targets the HSV-1 entry process and cell-to-cell viral spread. Our study showed that the inhibition of HSV-1 infectivity by bLf is dependent on its interaction with specific structural viral proteins. Apart from the prevention of early phases of viral infection, cell-to-cell spread inhibition activity of HSV-1 by bLf confirmed that this protein is an outstanding candidate for the treatment of herpetic infections since it would offer the advantage to prevent also viral infections caused by cell-associated virus.

Protease activity within nascently released human immunodeficiency virus type 1 (HIV-1) particles is responsible for the cleavage of the viral polyproteins Gag and Gag-Pol into their constituent parts, which results in the subsequent condensation of the mature conical core surrounding the viral genomic RNA. Concomitant with viral maturation is a conformational change in the packaged viral RNA from a loosely associated dimer into a more thermodynamically stable form. In this study we used suboptimal concentrations of two protease inhibitors, lopinavir and atazanavir, to study their effects on Gag polyprotein processing and on the properties of the RNA in treated virions. Analysis of the treated virions demonstrated that even with high levels of inhibition of viral infectivity (IC(90)), most of the Gag and Gag-Pol polyproteins were processed, although slight but significant increases in processing intermediates of Gag were detected. Drug treatments also caused a significant increase in the proportion of viruses displaying either immature or aberrant mature morphologies. The aberrant mature particles were characterized by an electron-dense region at the viral periphery and an electron-lucent core structure in the viral center, possibly indicating exclusion of the genomic RNA from these viral cores. Intriguingly, drug treatments caused only a slight decrease in overall thermodynamic stability of the viral RNA dimer, suggesting that the dimeric viral RNA was able to mature in the absence of correct core condensation.

ABSTRACT Frequent human immunodeficiency virus type 1 (HIV-1) recombination occurs during DNA synthesis when portions of the two copackaged RNAs are used as templates to generate a hybrid DNA copy. Therefore, the frequency of copackaging of genomic RNAs from two different viruses (heterozygous virion formation) affects the generation of genotypically different recombinants. We hypothesized that the selection of copackaged RNA partners is largely determined by Watson-Crick pairing at the dimer initiation signal (DIS), a 6-nucleotide palindromic sequence at the terminal loop of stem-loop 1 (SL1). To test our hypothesis, we examined whether heterozygous virion formation could be encouraged by manipulation of the DIS. Three pairs of viruses were generated with compensatory DIS mutations, designed so that perfect DIS base pairing could only occur between RNAs derived from different viruses, not between RNAs from the same virus. We observed that vector pairs with compensatory DIS mutations had an almost twofold increase in recombination rates compared with wild-type viruses. These data suggest that heterozygous virion formation was enhanced in viruses with compensatory DIS mutations (from 50% to more than 90% in some viral pairings). The role of the SL1 stem in heterozygous virion formation was also tested; our results indicated that the intermolecular base pairing of the stem sequences does not affect RNA partner selection. In summary, our results demonstrate that the Watson-Crick pairing of the DIS is a major determinant in the selection of the copackaged RNA partner, and altering the base pairing of the DIS can change the proportion of heterozygous viruses in a viral population. These results also strongly support the hypothesis that HIV-1 RNA dimers are formed prior to encapsidation.

Primate lentiviruses are composed of several distinct lineages, including human immunodeficiency virus type 1 (HIV-1), HIV-2, and simian immunodeficiency virus SIVagm. HIV-1 and HIV-2 have significant differences in the mechanisms of viral RNA encapsidation. Therefore, the RNA packaging mechanisms of SIVagm cannot be predicted from the studies of HIV-1 and HIV-2. We examined the roles of the nucleocapsid (NC) zinc finger motifs on RNA packaging by mutating the conserved zinc finger (CCHC) motifs, and whether SIVagm has a preference to package RNA in cis by comparing the RNA packaging efficiencies of gag mutants in the presence of a wild-type vector. Our results indicate that the SIVagm NC domain plays an important role in Gag-RNA recognition; furthermore SIVagm is distinct from the other currently known primate lentiviruses as destroying either zinc finger motif in the NC causes very drastic RNA packaging defects. Additionally, trans-packaging is a major mechanism for SIVagm RNA encapsidation.

Approximately one million people in the world are dually infected with both HIV-1 and HIV-2. To identify potential interactions between these two human pathogens, we examined whether HIV-1 and HIV-2 Gag proteins can coassemble and functionally complement each other. We generated HIV-1- and HIV-2-based vectors with mutations in Gag; compared with wild-type vectors, these mutants had drastically decreased viral titers. Coexpression of the mutant HIV-1 and HIV-2 Gag could generate infectious viruses; furthermore, heterologous complementation in certain combinations showed efficiency similar to homologous complementation. Additionally, we used bimolecular fluorescence complementation analysis to directly demonstrate that HIV-1 and HIV-2 Gag can interact and coassemble. Taken together, our results indicate that HIV-1 and HIV-2 Gag polyproteins can coassemble and functionally complement each other during virus replication; to our knowledge, this is the first demonstration of its kind. These studies have important implications for AIDS treatment and the evolution of primate lentiviruses.

After their release from host cells, most retroviral particles undergo a maturation process, which includes viral protein cleavage, core condensation, and increased stability of the viral RNA dimer. Inactivating the viral protease prevents protein cleavage; the resulting virions lack condensed cores and contain fragile RNA dimers. Therefore, protein cleavage is linked to virion morphological change and increased stability of the RNA dimer. However, it is unclear whether protein cleavage is sufficient for mediating virus RNA maturation. We have observed a novel phenotype in a murine leukemia virus capsid mutant, which has normal virion production, viral protein cleavage, and RNA packaging. However, this mutant also has immature virion morphology and contains a fragile RNA dimer, which is reminiscent of protease-deficient mutants. To our knowledge, this mutant provides the first evidence that Gag cleavage alone is not sufficient to promote RNA dimer maturation. To extend our study further, we examined a well-defined human immunodeficiency virus type 1 (HIV-1) Gag mutant that lacks a functional PTAP motif and produces immature virions without major defects in viral protein cleavage. We found that the viral RNA dimer in the PTAP mutant is more fragile and unstable compared with those from wild-type HIV-1. Based on the results of experiments using two different Gag mutants from two distinct retroviruses, we conclude that Gag cleavage is not sufficient for promoting RNA dimer maturation, and we propose that there is a link between the maturation of virion morphology and the viral RNA dimer.

Genetic recombination increases diversity in HIV-1 populations, thereby allowing variants to escape from host immunity or antiviral therapies. In addition to the currently described nine subtypes of HIV-1, many of the circulating strains are intersubtype recombinants. In this study, we determined the recombination rate between two HIV-1 subtype C viruses and between a subtype B virus and a subtype C virus during a single round of virus replication. Although HIV-1 subtype C recombines at a high rate, similar to that of HIV-1 subtype B, the recombination rate between a subtype B virus and a subtype C virus is much lower than the intrasubtype recombination rate. A 3-nt sequence difference in the dimerization initiation signal (DIS) region between HIV-1 subtypes B and C accounts for most of the reduction of intersubtype recombination. By matching the DIS sequences, the B/C intersubtype recombination rate was elevated 4-fold; by introducing mismatches in the 3-nt sequences, the B/B intrasubtype recombination rate was reduced 4-fold. Further analyses showed that the intermolecular template-switching frequency was unaffected by the sequence identity of the DIS region. These results support the hypothesis that mismatched sequences in the DIS region alter the formation of heterozygous virions, thereby lowering the observable recombination rate. Here, we present the discovery of a major restriction in HIV-1 intersubtype recombination. These results have important implications for virus evolution, the mechanism of HIV-1 RNA packaging, high negative interference in recombination, and the generation of circulating intersubtype recombinants within the infected population.

ABSTRACT Nucleocapsid (NC) proteins in most retroviruses have a well-conserved Cys-His box(es) as well as more divergent flanking regions that are rich in basic residues. Mutations in the flanking regions can affect RNA packaging, virus assembly, and reverse transcription of the viral RNA. To gain a further understanding of the roles of NC flanking regions in the retroviral replication cycle, we generated and characterized chimeric gag-pol expression constructs derived from murine leukemia virus and spleen necrosis virus by replacing an NC flanking region from one virus with the counterpart from the other virus. We found that all four chimeras were able to generate virions, package viral RNA, and complete the viral replication cycle. Two chimeras had mild defects in virus assembly that correlated with a decrease in the isoelectric points of NCs, suggesting that the basic nature of NC is important in virus assembly. This finding indicates that, although the primary sequences of these flanking regions have little homology, the heterologous sequences are functional both as part of the Gag polyprotein and as processed NC protein.

We have identified a region near the C terminus of capsid (CA) of murine leukemia virus (MLV) that contains many charged residues. This motif is conserved in various lengths in most MLV-like viruses. One exception is that spleen necrosis virus (SNV) does not contain a well-defined domain of charged residues. When 33 amino acids of the MLV motif were deleted to mimic SNV CA, the resulting mutant produced drastically reduced amounts of virions and the virions were noninfectious. Furthermore, these viruses had abnormal sizes, often contained punctate structures resembling those in the cell cytoplasm, and packaged both ribosomal and viral RNA. When 11 or 15 amino acids were deleted to modify the MLV CA to resemble those from other gammaretroviruses, the deletion mutants produced virions at levels comparable to those of the wild-type virus and were able to complete one round of virus replication without detectable defects. We generated 10 more mutants that displayed either the wild-type or mutant phenotype. The distribution of the wild-type or mutant phenotype did not directly correlate with the number of amino acids deleted, suggesting that the function of the motif is determined not simply by its length but also by its structure. Structural modeling of the wild-type and mutant proteins suggested that this region forms alpha-helices; thus, we termed this motif the "charged assembly helix." This is the first description of the charged assembly helix motif in MLV CA and demonstration of its role in virus budding and assembly.

ABSTRACT Mutagenesis studies have shown that retroviral nucleocapsid (NC) protein Zn 2+ fingers (-Cys-X 2 -Cys-X 4 -His-X 4 -Cys- [CCHC]) perform multiple functions in the virus life cycle. Moloney murine leukemia virus mutants His 34→Cys (CCCC) and Cys 39→His (CCHH) were able to package their genomes normally but were replication defective. Thermal dissociation experiments showed that the CCHH mutant was not defective in genomic RNA dimer structure. Primer tRNA placement on the viral genome and the ability of the tRNA to function in reverse transcription initiation in vitro also appear normal. Some “full-length” DNA copies of the viral genome were synthesized in mutant virus-infected cells. The CCCC and CCHH mutants produced these DNA copies at greatly reduced levels. Circle junction fragments, amplified from two-long-terminal-repeat viral DNA (vDNA) by PCR, were cloned and characterized. Remarkably, it was discovered that vDNA isolated from cells infected with mutant virions had a wide variety of abnormalities at the site at which the two ends of the linear precursor had been ligated to form the circle (i.e., the junction between the 5′ end of U3 and the 3′ end of U5). In some molecules, bases were missing from regions corresponding to the U3 and U5 linear vDNA termini; in others, the viral sequences extended either beyond the U5 sequences into the primer-binding site and 5′ leader or beyond the U3 sequences into the polypurine tract into the env coding region. Still other molecules contained nonviral sequences between the linear vDNA termini. Such defective genomes would certainly be unsuitable substrates for integration. Thus, strict conservation of the CCHC structure in NC is required for infection events prior to and possibly including integration.

Retroviruses contain a dimeric RNA consisting of two identical molecules of plus-strand genomic RNA. The structure of the linkage between the two monomers is not known, but they are believed to be joined near their 5' ends. Darlix and coworkers have reported that transcripts of retroviral RNA sequences can dimerize spontaneously in vitro (see, for example, E. Bieth, C. Gabus, and J. L. Darlix, Nucleic Acids Res. 18:119-127, 1990). As one approach to identification of sequences which might participate in the linkage, we have mapped sequences derived from the 5' 378 bases of Harvey sarcoma virus (HaSV) RNA which can dimerize in vitro. We found that at least three distinct regions, consisting of nucleotides 37 to 229, 205 to 272, and 271 to 378, can form these dimers. Two of these regions contain nucleotides 205 to 226; computer analysis suggests that this region can form a stem-loop with an inverted repeat in the loop. We propose that this hypothetical structure is involved in dimer formation by these two transcripts. We also compared the thermal stabilities of each of these dimers with that of HaSV viral RNA. Dimers of nucleotides 37 to 229 and 205 to 272 both exhibited melting temperatures near that of viral RNA, while dimers of nucleotides 271 to 378 are quite unstable. We also found that dimers of nucleotides 37 to 378 formed at 37 degrees C are less thermostable than dimers of the same RNA formed at 55 degrees C. It seems possible that bases from all of these regions participate in the dimer linkage present in viral RNA.