Characterization of the transcription map and Rev activity of a highly cytopathic feline immunodeficiency virus

Viral transcriptome analysis of feline immunodeficiency virus infected cells using second generation sequencing technology

Feline immunodeficiency virus (FIV) is a widespread pathogen causing immunodeficiency in domestic cats and related wild cat species. The virus genome includes the main structural genes common to all retroviruses as well as accessory genes displaying essential functions during the viral life cycle. Expression of viral genes involves transcription of provirus genomes into full-length transcripts, which are partially processed into several spliced mRNA variants for the translation of particular proteins. Among several FIV isolates derived from domestic cats, notable differences in pathogenicity could be observed leading to identification of low and high pathogenic virus isolates. This study investigates the viral transcriptome of two differentially virulent FIV strains using second generation sequencing (SGS) technology. The expression levels of viral genes as detected by SGS were additionally determined by reverse transcription quantitative PCR analysis in order to compare two methods of mRNA quantification. The different properties of both methods, especially regarding normalization between samples, had to be considered when comparing the resulting data. SGS turned out to be a suitable technique for comparing mRNA transcription between both FIV infected cell lines and the identification of spliced viral transcripts. In contrast to this, the quantification of these spliced isoforms using SGS data was impeded by the short length of sequencing reads. In summary, SGS analysis revealed very consistent mRNA levels for the majority of viral genes between the low pathogenic Petaluma and the more highly pathogenic Glasgow 8 isolate. Notable differences among the two FIV strains could be observed in the viral mRNA splicing where Glasgow 8 displays similarities to the transcription pattern seen in the early stages of natural lentivirus infections. Thus, divergences in the regulation of post-transcriptional RNA processing might represent an additional contributor to the diverse pathogenic effects of individual FIV isolates. Taken together, this study aims to investigate the viral transcriptome as one part of the complex network of virus-host interactions, which will contribute to gaining deeper insights into FIV pathogenesis.

Domain- and nucleotide-specific Rev response element regulation of feline immunodeficiency virus production

Computational analysis of feline immunodeficiency virus (FIV) RNA sequences indicated that common FIV strains contain a rev response element (RRE) defined by a long unbranched hairpin with 6 stem-loop sub-domains, termed stem-loop A (SLA). To examine the role of the RNA secondary structure of the RRE, mutational analyses were performed in both an infectious FIV molecular clone and a FIV CAT-RRE reporter system. These studies disclosed that the stems within SLA (SA1, 2, 3, 4, and 5) of the RRE were critical but SA6 was not essential for FIV replication and CAT expression. These studies also revealed that the secondary structure rather than an antisense protein (ASP) mediates virus expression and replication in vitro. In addition, a single synonymous mutation within the FIV-RRE, SA3/45, reduced viral reverse transcriptase activity and p24 expression after transfection but in addition also showed a marked reduction in viral expression and production following infection.

Evolution of the long terminal repeat and accessory genes of feline immunodeficiency virus genomes from naturally infected cougars

FIVpco is a member of the feline immunodeficiency virus family that is endemic in wild cougar populations. Virus replication is robust in FIVpco-infected cougars but there are no consequences of infection to cougar survival, fecundity or susceptibility to other infections. Unlike pathogenic lentiviruses, there is no evidence for positive selection on FIVpco gag or env. To better understand how lentivirus genomes evolve in natural infections, we evaluated the regulatory region and accessory genes from fourteen full-length FIVpco genomes, which represent the FIVpco diversity in the Northern Rockies Ecosystem. Our data demonstrate that the two sister groups of FIVpco have each acquired binding sites for different interferon response factors (IRF). The most variable gene in the FIVpco genome encodes OrfA, although there is no indication that it, or any other accessory gene, is under positive selection. There is a single-splice acceptor site for vif expression, which is conserved among all FIVpco genomes. However, there are several putative means to express rev and orfA, which differ between the phylogenetic groups of FIVpco. Our comparative study on divergent FIVpco genomes indicates that variation in potential gene regulation mechanisms, not changes in structural proteins, characterize the evolution of FIVpco in natural infections.

Expression of feline immunodeficiency virus Vif is associated with reduced viral mutation rates without restoration of replication of vif mutant viruses

The vif gene of lentiviruses has been demonstrated to be essential for efficient viral replication in many cell types. Although the Vif protein of feline immunodeficiency virus (FIV) displays limited homology to HIV-1 Vif, the role of vif in FIV replication is not known. We have examined the requirements of vif for replication of a FIV strain isolated from a non-domestic felid, Otocolobus manul (FIV-Oma). In agreement with others, we find that replication of FIV vif mutant molecular clones in CrFK cells is highly attenuated. Initial attempts to rescue vif mutant viruses in trans were limited by lack of detectable wild-type Vif expression from DNA constructs. We demonstrate that FIV-Oma Vif expression can be increased by re-synthesis of the gene to remove splice donor and acceptor sites as well as improving codon usage to a mammalian codon optimized model. Cellular localization of resynthesized Vif (Vif-RS) is cytoplasmic. Clonal stable transfectants expressing HA-tagged Vif-RS do not restore replication levels of vif mutant virus. However, in such cell lines, G-to-A mutation rates in replicating wild-type viruses are reduced.

Recombination in feline lentiviral genomes during experimental cross-species infection

Domestic cats develop an asymptomatic, productive infection with a feline immunodeficiency virus (PLV) derived from a naturally infected cougar (P. concolor). We previously demonstrated that there are extensive G to A substitutions, characteristic of host cytidine deaminase editing, and positive selection on reverse transcriptase in the PLV genome during this cross-species infection. In this study, we evaluated full-length viral genomes from each of four cats infected with PLV to determine if viral recombination occurred during this single source infection. Recombination rates were measurable in three of the four infected cats. In two of these animals, a single site in reverse transcriptase was under positive selection and there was significant topological incongruence among individual genes in the 3' half of the genomes. The break point was proximate to a splice site used for accessory gene expression. Our data indicate that recombination can facilitate lentivirus persistence in unfavorable environments such as a new host species.

Going wild: lessons from naturally occurring T-lymphotropic lentiviruses

Over 40 nonhuman primate (NHP) species harbor species-specific simian immunodeficiency viruses (SIVs). Similarly, more than 20 species of nondomestic felids and African hyenids demonstrate seroreactivity against feline immunodeficiency virus (FIV) antigens. While it has been challenging to study the biological implications of nonfatal infections in natural populations, epidemiologic and clinical studies performed thus far have only rarely detected increased morbidity or impaired fecundity/survival of naturally infected SIV- or FIV-seropositive versus -seronegative animals. Cross-species transmissions of these agents are rare in nature but have been used to develop experimental systems to evaluate mechanisms of pathogenicity and to develop animal models of HIV/AIDS. Given that felids and primates are substantially evolutionarily removed yet demonstrate the same pattern of apparently nonpathogenic lentiviral infections, comparison of the biological behaviors of these viruses can yield important implications for host-lentiviral adaptation which are relevant to human HIV/AIDS infection. This review therefore evaluates similarities in epidemiology, lentiviral genotyping, pathogenicity, host immune responses, and cross-species transmission of FIVs and factors associated with the establishment of lentiviral infections in new species. This comparison of consistent patterns in lentivirus biology will expose new directions for scientific inquiry for understanding the basis for virulence versus avirulence.

Rev response elements (RRE) in lentiviruses: an RNAMotif algorithm-based strategy for RRE prediction

Lentiviruses (a sub-family of the retroviridae family) include primate and non-primate viruses associated with chronic diseases of the immune system and the central nervous system. All lentiviruses encode a regulatory protein Rev that is essential for post-transcriptional transport of the unspliced and incompletely spliced viral mRNAs from nuclei to cytoplasm. The Rev protein acts via binding to an RNA structural element known as the Rev responsive element (RRE). The RRE location and structure and the mechanism of the Rev-RRE interaction in primate and non-primate lentiviruses have been analyzed and compared. Based on structural data available for RRE of HIV-1, a two step computational strategy for prediction of putative RRE regions in lentivirus genomes has been developed. First, the RNAMotif algorithm was used to search genomic sequence for highly structured regions (HSR). Then the program RNAstructure, version 3.6 was used to calculate the structure and thermodynamic stability of the region of approximately 350 nucleotides encompassing the HSR. Our strategy correctly predicted the locations of all previously reported lentivirus RREs. We were able also to predict the locations and structures of potential RREs in four additional lentiviruses.