Showing papers by "Chantal Abergel published in 2010"

mRNA deep sequencing reveals 75 new genes and a complex transcriptional landscape in Mimivirus.

Abstract: Mimivirus, a virus infecting Acanthamoeba, is the prototype of the Mimiviridae, the latest addition to the nucleocytoplasmic large DNA viruses. The Mimivirus genome encodes close to 1000 proteins, many of them never before encountered in a virus, such as four amino-acyl tRNA synthetases. To explore the physiology of this exceptional virus and identify the genes involved in the building of its characteristic intracytoplasmic "virion factory," we coupled electron microscopy observations with the massively parallel pyrosequencing of the polyadenylated RNA fractions of Acanthamoeba castellanii cells at various time post-infection. We generated 633,346 reads, of which 322,904 correspond to Mimivirus transcripts. This first application of deep mRNA sequencing (454 Life Sciences [Roche] FLX) to a large DNA virus allowed the precise delineation of the 5' and 3' extremities of Mimivirus mRNAs and revealed 75 new transcripts including several noncoding RNAs. Mimivirus genes are expressed across a wide dynamic range, in a finely regulated manner broadly described by three main temporal classes: early, intermediate, and late. This RNA-seq study confirmed the AAAATTGA sequence as an early promoter element, as well as the presence of palindromes at most of the polyadenylation sites. It also revealed a new promoter element correlating with late gene expression, which is also prominent in Sputnik, the recently described Mimivirus "virophage." These results-validated genome-wide by the hybridization of total RNA extracted from infected Acanthamoeba cells on a tiling array (Agilent)--will constitute the foundation on which to build subsequent functional studies of the Mimivirus/Acanthamoeba system.

Identification of an l-Rhamnose Synthetic Pathway in Two Nucleocytoplasmic Large DNA Viruses

Abstract: The nucleocytoplasmic large DNA viruses (NCLDVs) are a heterogeneous group of viruses that infect several eukaryotic organisms (25, 49). They have large genomes that often encode genes not commonly found in viruses. For example, several lines of evidence indicate that Paramecium bursaria chlorella virus 1 (PBCV-1) and other chlorovirus members, such as Acanthocystis turfacea chlorella virus 1 (ATCV-1), encode at least part, if not all, of the machinery required to glycosylate their structural proteins, including glycosyltransferases (13, 21, 30, 33, 41-43). Furthermore, glycosylation occurs independently of the host endoplasmic reticulum (ER)-Golgi system (33, 42-44). The PBCV-1 major capsid protein located on the viral surface is glycosylated, and the glycan moieties contribute to virus protease resistance and antigenicity. We have previously reported that PBCV-1 and ATCV-1 encode both GDP-d-mannose 4,6-dehydratase (GMD) and GDP-4-keto-6-deoxy-d-mannose epimerase/reductase (GMER), which are involved in “de novo” GDP-l-fucose synthesis (14, 40). Because fucose is a component of the glycan portion of the PBCV-1 major capsid protein (43), the viral GMD and GMER enzymes may be necessary to provide the virus with the nucleotide sugar. The aim of the present study was to identify and analyze additional virus-encoded enzymes involved in glycan production. We included chlorella virus ATCV-1, and we extended our study to Acanthamoeba polyphaga mimivirus. Mimivirus is a giant DNA virus that infects members of the genus Acanthamoeba (6-9, 35). Its 1.2-Mb genome is the largest viral genome described so far, containing more than 900 protein-coding sequences (CDS) (36). Annotation of ATCV-1 and mimivirus genomes identified genes encoding putative enzymes involved in l-rhamnose production. This 6-deoxyhexose sugar is a common component of surface glycoconjugates such as bacterial lipopolysaccharides (LPS), where it plays an important role in pathogenicity (28, 29). l-Rhamnose also occurs in plant cell wall rhamnogalactoglucans and rhamnosides, such as flavonoids, terpenoids, and saponins (24, 38). Moreover, rhamnose is also present in the virus PBCV-1 glycan(s) attached to its major capsid protein (43). Bacteria and plants synthesize l-rhamnose from glucose by two slightly different pathways (Fig. ​(Fig.1).1). In bacteria, dTDP-d-glucose is the initial substrate for a dehydratase activity (RfbB/RmlB), which eliminates a water molecule and leads to production of an unstable intermediate compound, dTDP-4-keto-6-deoxy-d-glucose. This compound is then subjected to epimerization at C-3 and C-5 by RfbC/RmlC and finally to an NADPH-dependent reduction of C-4 by RfbD/RmlD. Three separate enzymes catalyze these steps (2, 17, 18, 20). In contrast, in the plant Arabidopsis thaliana the initial substrate is UDP-d-glucose, and the three enzymatic activities are fused into a single polypeptide, named RHM enzymes, as depicted in Fig. ​Fig.11 (34). An enzyme with epimerase/reductase activity (NSR/ER) leading to UDP-l-rhamnose was also identified in A. thaliana (44). FIG. 1. l-Rhamnose biosynthetic pathways in bacterial and plant cells. In bacteria, the three enzymatic activities are on separate polypeptides. In plants, UGD and UGER are fused into a single polypeptide (RHM isoforms) (34). In this study we identified and characterized three functional enzymes involved in UDP-l-rhamnose synthesis in both ATCV-1 and mimivirus. ATCV-1 encodes only a UDP-d-glucose 4,6-dehydratase (UGD), whereas mimivirus encodes the complete pathway, i.e., UGD and a bifunctional UDP-4-keto-6-deoxy-d-glucose 3,5-epimerase/4-reductase (UGER). The virus-encoded UDP-l-rhamnose pathway and the enzymatic properties are similar to those described for plants. Sequence and phylogenetic analyses indicate that ATCV-1 likely acquired UGD from its chlorella host through a recent horizontal gene transfer (HGT). In contrast, both UGD and UGER were transmitted much earlier, probably between mimivirus and a protozoan ancestral host. Thus, these results support the hypothesis that both ATCV-1 and mimivirus encode at least part of a host-independent glycosylation system, which may be essential for virus replication and infection.

Macromolecular crystal data phased by negative-stained electron-microscopy reconstructions.

Abstract: The combination of transmission electron microscopy with X-ray diffraction data is usually limited to relatively large particles. Here, the approach is continued one step further by utilizing negative staining, a technique that is of wider applicability than cryo-electron microscopy, to produce models of medium-size proteins suitable for molecular replacement. The technique was used to solve the crystal structure of the dodecameric type II dehydroquinase enzyme from Candida albicans (approximately 190 kDa) and that of the orthologous Streptomyces coelicolor protein.