Functional analysis of the influenza A virus cRNA promoter and construction of an ambisense transcription system.

While influenza A viral RNA is known to act as a template for the synthesis of both viral mRNA and complementary cRNA, the latter has been observed so far only to function as an intermediate in replication and give rise to progeny vRNA molecules. Here it is shown that the cRNA promoter is also capable of initiating viral mRNA synthesis, similar to vRNA-promoted transcription adhering to the cap-snatching mode of primer recruitment. Detection of cRNA promoted transcription required an inversion of the reporter gene coding sequence plus relocation of the viral polyadenylation signal. Construction of cRNA promoter variants through RNA polymerase I reverse genetics allowed us to determine the RNA polymerase-associated, base-paired conformation in a reporter gene read-out system. It again turned out to adhere to the "corkscrew" model, similar, but slightly different in its binding interactions from the corresponding vRNA conformation. The observation of two transcription reactions, initiated in either direction from influenza vRNA and cRNA template molecules, allowed us to construct bicistronic, ambisense RNA molecules for simultaneous expression of two proteins from a single segment of viral RNA.

The packaging signal of influenza viral RNA molecules.

The packaging signal present in influenza viral RNA molecules is shown not to constitute a separate structural element, but to reside within the 5'-bulged promoter structure, as caused by the central unpaired residue A10 in its 5' branch. Upon insertion of two uridine residues in the 3' branch opposite A10, the minus-strand viral RNA (vRNA) promoter is converted into a 3'-bulged structure, whereas the plus-strand cRNA promoter instead adopts the 5'-bulged conformation. In this promoter variant it is exclusively the cRNA that is found packaged in the progeny virions. Upon insertion of only a single uridine nucleotide opposite 5'A10, the two debulged structures of the vRNA and cRNA promoters are rendered identical, and both vRNA and cRNA molecules are packaged indiscriminately, in a 1:1 ratio, but at lower rates. We propose that the binding interactions of viral polymerase with either of the two differently bulged vRNA and cRNA promoter structures result in two different conformations of the enzyme protein. Only the 5' bulged RNA-associated polymerase conformation appears to be recognized for nuclear export, which depends on nuclear matrix protein M1 and nonstructural protein NS2. And the respective wild-type vRNP- or insertion mutant cRNP complex is observed to enter the cytoplasm and hence is included in the viral encapsidation process, which takes place at the plasma membrane.

Influenza virus propagation is impaired by inhibition of the Raf/MEK/ERK signalling cascade.

Influenza A viruses are important worldwide pathogens in humans and different animal species. The functions of most of the ten different viral proteins of this negative-strand RNA virus have been well elucidated. However, little is known about the virus-induced intracellular signalling events that support viral replication. The Raf/MEK/ERK cascade is the prototype of mitogen-activated protein (MAP) kinase cascades and has an important role in cell growth, differentiation and survival. Investigation of the function of this pathway has been facilitated by the identification of specific inhibitors such as U0126, which blocks the cascade at the level of MAPK/ERK kinase (MEK). Here we show that infection of cells with influenza A virus leads to biphasic activation of the Raf/MEK/ERK cascade. Inhibition of Raf signalling results in nuclear retention of viral ribonucleoprotein complexes (RNPs), impaired function of the nuclear-export protein (NEP/NS2) and concomitant inhibition of virus production. Thus, signalling through the mitogenic cascade seems to be essential for virus production and RNP export from the nucleus during the viral life cycle.

Efficient expression of the tumor-associated antigen MAGE-3 in human dendritic cells, using an avian influenza virus vector.

Dendritic cells (DCs) are the most potent inducers of immune reactions. Genetically modified DCs, which express tumor-associated antigens (TAA), can efficiently induce antitumor immunity and thus have a high potential as tools in cancer therapy. The gene delivery is most efficiently achieved by viral vectors. Here, we explored the capacity of influenza virus vectors to transduce TAA genes. These viruses abortively infect DCs without interfering with their antigen-presenting capacity. In contrast to other viruses used for DC transduction, influenza viruses can be efficiently controlled by antiviral pharmaceuticals, lack the ability to integrate into host chromosomes, and fail to establish persistent infections. Genes encoding a melanoma-derived TAA (MAGE-3), or the green fluorescence protein (GFP), were introduced into a high-expression avian influenza virus vector. Monocyte-derived mature DCs infected by these recombinants efficiently produced GFP or MAGE-3. More than 90% of the infected DCs can express a transduced gene. Importantly, these transduced DCs retained their characteristic phenotype and their potent allogeneic T cell stimulatory capacity, and were able to stimulate MAGE-3-specific CD8(+) cytotoxic T cells. Thus influenza virus vectors provide a highly efficient gene delivery system in order to transduce human DCs with TAA, which consequently stimulate TAA-specific T cells.

Recombinant expression and modification analysis of protein agno-1b encoded by avian polyomavirus BFDV.

Among two pairs of agnoproteins encoded in upstream positions in the late mRNAs of avian polyomavirus BFDV, either agno-1a or its splice derivative agno-1b are required for viral propagation. Out of the two proteins both of which consist of multiple electrophoretic subspecies, the smaller and less complex agno-1b has been cDNA-cloned into an influenza-virus /RNA-polymerase I expression system for production of higher amounts of this protein in infected chicken embryo fibroblasts. Fractional modification of agno-1b by phosphorylation at residues serine 51, serine 53, and threonine 73 is demonstrated through dephosphorylation by alkaline phosphatase, mass spectrometry of individual protein species isolated by strong anion exchange chromatography, and single or multiple alanine substitutions of serine or threonine residues in site-directed mutagenesis.

A novel secreted ribonuclease from Bacillus intermedius: gene structure and regulatory control.

A second secreted ribonuclease, designated binase II, has been detected in Bacillus intermedius 7P, and its structural gene was cloned and sequenced. Unlike the well-known binase I, a 109-amino acid guanyl-specific enzyme, the 292-residue binase II is closely related to the B. subtilis nuclease Bsn, in structure and in its enzymatic properties. Binase II is also insensitive to inactivation by barstar, an inhibitor protein that is specific for guanyl-specific ribonucleases. While both B. intermedius enzymes are induced upon phosphate starvation, only the gene for binase I belongs to the pho regulon system and carries pho-box elements adjacent to its promoter sequence. The gene for binase II is similar to that for Bsn in lacking such elements. The birB gene coding for binase II appears to be located next to the 3'-end of a ferric ion transport operon, with which it convergently overlaps. This would allow attenuator control over binase II expression under conditions of starvation for ferric ions.

A DNA transfection system for generation of influenza A virus from eight plasmids.

We have developed an eight-plasmid DNA transfection system for the rescue of infectious influenza A virus from cloned cDNA. In this plasmid-based expression system, viral cDNA is inserted between the RNA polymerase I (pol I) promoter and terminator sequences. This entire pol I transcription unit is flanked by an RNA polymerase II (pol II) promoter and a polyadenylation site. The orientation of the two transcription units allows the synthesis of negative-sense viral RNA and positive-sense mRNA from one viral cDNA template. This pol I-pol II system starts with the initiation of transcription of the two cellular RNA polymerase enzymes from their own promoters, presumably in different compartments of the nucleus. The interaction of all molecules derived from the cellular and viral transcription and translation machinery results in the generation of infectious influenza A virus. The utility of this system is proved by the recovery of the two influenza A viruses: A/WSN/33 (H1N1) and A/Teal/HK/W312/97 (H6N1). Seventy-two hours after the transfection of eight expression plasmids into cocultured 293T and MDCK cells, the virus yield in the supernatant of the transfected cells was between 2 x 10(5) and 2 x 10(7) infectious viruses per milliliter. We also used this eight-plasmid system for the generation of single and quadruple reassortant viruses between A/Teal/HK/W312/97 (H6N1) and A/WSN/33 (H1N1). Because the pol I-pol II system facilitates the design and recovery of both recombinant and reassortant influenza A viruses, it may also be applicable to the recovery of other RNA viruses entirely from cloned cDNA.

Molecular characterization of a Litomosoides sigmodontis protein involved in the development of the microfilarial sheath during embryogenesis.

A cDNA clone Ls110 was isolated from a female Litomosoides sigmodontis expression library using an antiserum raised against the microfilarial sheath. The complete cDNA encodes a protein (Ls110) of 382 amino acids. Southern and PCR analyses revealed the presence of Ls110 in L. sigmodontis as a single copy gene. The transcription of the Ls110 gene was limited to female worms. In these worms the transcription was confined to the epithelial cells of the uterus. The protein Ls110 was detected not only in the epithelial layer of the uterus but also secreted in the lumen of the uterus. All the intra-uterine embryonic stages showed the protein bound to their egg shell/sheath, except the early multicellular embryonic stages and fully developed microfilariae. The transient occurrence of Ls110 on these structures of intra-uterine stages besides the presence of a cysteine-rich N-terminal region (SXC-like domain) suggest that the protein may play a role in the formation of the microfilarial sheath during embryogenesis.

"Ambisense" approach for the generation of influenza A virus: vRNA and mRNA synthesis from one template.

We present a system for creating influenza virus by generating viral RNA (vRNA) and mRNA from one template. Recently, a system for the generation of influenza A virus entirely from cloned cDNAs was established (Neumann et al., 1999, Proc. Natl. Acad. Sci. USA 96, 9345-9350). Cells were transfected with plasmids for RNA polymerase I-driven intracellular synthesis of all eight viral RNAs, and with protein expression plasmids for the synthesis of viral structural proteins. Although this system is highly efficient in virus generation, the construction and cotransfection of 17 plasmids is cumbersome and may limit the use of this system to cell lines that can be transfected with high efficiencies. Synthesizing both vRNA and mRNA from one template would reduce the number of plasmids required for virus generation. Therefore, we generated a bidirectional transcription construct that contains cDNA encoding PB1 flanked by an RNA polymerase I (pol I) promoter for vRNA synthesis and an RNA polymerase II (pol II) promoter for mRNA synthesis. The utility of this approach is proved by the generation of virus after transfecting the pol I/pol II-promoter-PB1 construct together with vRNA- and protein-expression constructs for the remaining seven segments. Because this approach reduces the number of plasmids required for virus generation, it also reduces the work necessary for cloning, probably enhances the efficiency of virus generation, and expands the use of the reverse-genetics system to cell lines for which efficient cotransfection of 17 plasmids cannot be achieved.

Agnoprotein-1a of avian polyomavirus budgerigar fledgling disease virus: identification of phosphorylation sites and functional importance in the virus life-cycle.

The avian polyomavirus budgerigar fledgling disease virus (BFDV) encodes an unusual set of four agnoproteins in its late upstream region. Of the two pairs of these proteins, which overlap each other in two different reading frames, the p(L1)-promoted agnoprotein-1a (agno-1a) is the dominant species and is able to support virus propagation in the absence of the other three polypeptides. Viral BFDV agno-1a, and also agno-1a expressed via an influenza virus vector, consists of a complex series of electrophoretically separable subspecies that can be reduced by phosphatase action down to a primary unphosphorylated protein with an apparent molecular mass of 31 kDa. Through peptide mass spectrometry and site-directed mutagenesis, the positions of four serine and three threonine residues have been determined as phosphate-accepting groups, which are partially modified by the combined action of three different cellular kinases. Since extensively phosphorylated agno-1a is required for its intracellular function, control over VP protein expression, and unphosphorylated agno-1a is observed as an additional component in the BFDV virion, both extreme subspecies appear to be drawn from that complex mixture, which also includes the intermediate stages of phosphorylation.

Evidence for translation of VP3 of avian polyomavirus BFDV by leaky ribosomal scanning.

Due to several incomplete splicing reactions, budgerigar fledgling disease virus (BFDV) late mature mRNAs are either bicistronic or polycistronic with an agnogene located upstream of viral protein (VP) genes. While the bicistronic mRNAs code for the vast majority of VP1, the polycistronic mRNAs contain the coding sequences of VP2, VP3, and VP1 (as the most distal cistron relative to VP2 and VP3). In this work, the translation initiation mechanism of VP3 was investigated in chicken embryo fibroblast cells by transfection of a series of BFDV mutant clones and transient reporter gene chloramphenicol acetyltransferase (CAT) expression assay, leading to the conclusion that BFDV VP3 was translated by leaky ribosomal scanning. Furthermore, thanks to the high sensitivity of CAT assay experiment, we were able to demonstrate that ribosomes could reach VP1-AUG and initiate translation after scanning through 900 nucleotides on the unspliced polycistronic mRNA.

Recombinant expression of late genes agno-2a and agno-2b of avian polyomavirus BFDV.

Budgerigar fledgling disease virus (BFDV) genome contains two times two (two pairs) open reading frames (agnogenes) at the 5' end of the late coding region. Recombinant influenza A viruses were constructed to express the second pair of BFDV agnoproteins, agno-2a and agno-2b, with a fusion of a histidine-tag at their carboxy-termini, respectively. Specific proteins were detected in Western blot analysis using anti histidine-tag monoclonal antibody. By indirect immunofluorescence experiments agno-2a and agno-2b were shown to be located on the surface and in the perinuclear and cytoplasmic areas of infected cells. Comparisons of the expression patterns of BFDV agno-2a and agno-2b with that of simian virus 40 agnoprotein reveal high similarity, suggesting that they might have the same function(s) in polyomavirus infectious cycle.

Transient bicistronic vRNA segments for indirect selection of recombinant influenza viruses.

The 5'- and 3'-terminal regions of influenza vRNA molecules are known to constitute the promoter structure upon association with viral RNA polymerase in an activated complementary conformation. An inherent requirement for their location at the very ends of the vRNA molecules always has been implied because of that natural structure, but this study demonstrates that one or both of the promoter sequences may be relocated into vRNA-internal positions and still retain their polymerase-binding function. External extensions of vRNA molecules employed include either single-stranded RNA sequences

Generation of influenza A viruses entirely from cloned cDNAs.

We describe a new reverse-genetics system that allows one to efficiently generate influenza A viruses entirely from cloned cDNAs. Human embryonic kidney cells (293T) were transfected with eight plasmids, each encoding a viral RNA of the A/WSN/33 (H1N1) or A/PR/8/34 (H1N1) virus, flanked by the human RNA polymerase I promoter and the mouse RNA polymerase I terminator-together with plasmids encoding viral nucleoprotein and the PB2, PB1, and PA viral polymerases. This strategy yielded >1 x 10(3) plaque-forming units (pfu) of virus per ml of supernatant at 48 hr posttransfection. The addition of plasmids expressing all of the remaining viral structural proteins led to a substantial increase in virus production, 3 x 10(4)-5 x 10(7) pfu/ml. We also used reverse genetics to generate a reassortant virus containing the PB1 gene of the A/PR/8/34 virus, with all other genes representing A/WSN/33. Additional viruses produced by this method had mutations in the PA gene or possessed a foreign epitope in the head of the neuraminidase protein. This efficient system, which does not require helper virus infection, should be useful in viral mutagenesis studies and in the production of vaccines and gene therapy vectors.

Membrane-anchored incorporation of a foreign protein in recombinant influenza virions.

The RNA polymerase I system for in vivo synthesis of recombinant influenza vRNA molecules was used for the expression of a chimeric protein, consisting of the 341-amino-acid ectodomain of the glycoprotein E2 of classical swine fever virus and the 37-amino-acid C-terminal membrane anchor of the influenza virus hemagglutinin (HA). During infection with an influenza A helper virus the amplified pseudo-viral RNA was packaged into progeny virions together with influenza vRNA segments. The foreign fusion protein E2-HA was shown to be physically incorporated into the viral envelope. Incorporation of a third major glycoprotein into the envelope did not affect biological functions of HA and neuraminidase that are required for the generation of infectious virus particles. Based on mutational analyses of the cytoplasmic tail of E2-HA fusion proteins three modes of interaction during virus budding have been observed: nonspecific low-level incorporation (truncated tails), specific full-level incorporation (wild-type amino acid sequence or minor variations of it), and exclusion from incorporation (elongated tails).

Antiviral ribozymes. New jobs for ancient molecules.

Catalytic RNAs are a genetic property not only of some particular viroids or viruses, but also are more common naturally among eukaryotes and even prokaryotes than earlier expected. However, the major interest in ribozymes results from their potential for development of "tailor-made" cDNA constructions designed to be transcribed into catalytic RNAs that will recognize by hybridization and destroy by specific cleavage their cellular or viral RNA targets. The efficiency of an antiviral ribozyme is determined by both the accessibility and sequence conservation of the target region, as well as the design of the ribozyme: its type, size, and composition of flanking sequences; expression rates; and cellular compartment localization. Until now the most frequently selected viral target is the human immunodeficiency virus, where an up to a 10(4)-fold inhibition in its progeny production has been achieved. Although the first generation ribozymes focused on improvements in basic design and expression rates, more recently the efficiency of antiviral catalytic activity has been increased by employing polyribozymes and/or multitarget ribozymes, as well as special constructions to enhance the cellular co-compartmentation of the ribozyme with its viral RNA target.

Expression of the microfilarial sheath protein 2 (shp2) of the filarial parasites Litomosoides sigmodontis and Brugia malayi.

The microfilarial sheaths of the filarial parasites Brugia malayi, Brugia pahangi, and Litomosoides sigmodontis consist of several parasite proteins, probably ranging between 7 and 10. The gene encoding sheath protein 2 (shp2), which is the object of this study, is transcribed in embryos and in the uterine epithelium; at least in B. malayi, it is translated in both tissues. Apparently, shp2 is synthesized as a monomer, exported by the respective cells, and integrated into the microfilarial sheath. In the sheath, it exists as a highly polymerized molecule cross-linked by cysteine formation and other covalent bonds, presumably epsilon-(gamma-glutamyl)-lysine links.

Structural deviations in a bovine low expression lysozyme-encoding gene active in tissues other than stomach.

Lysozyme-encoding genes (Lys) constitute a gene-family in ruminants. While several of these genes are highly expressed in stomach (sLys), few other copies are weakly expressed in other tissues, notably in polymorphnuclear granulocytes and macrophages (mLys). Searching an understanding for these grossly different levels of expression, we isolated the bovine variant of the gene being expressed in granulocytes and characterized it by sequencing, together with its promoter. Spanning about 9 kb of genomic DNA, the gene is found to be segmented into four exons, in common with all other Lys, as known from vertebrates. Sequence homologies between all bovine sLys-variants exceeds 70% over much of the entire coding sequence and promoter region. This indicates (i) that bovine lysozymes expressed either in stomach or granulocyte originate from a common ancestral gene and (ii) also excludes the possibility that the observed weak expression of the mLys gene is due to major structural rearrangements within the promoter segment. However, primer extension analysis based on RNA isolated from kidney locates the transcription startpoint (tsp of that gene) 44 nt further upstream than observed in both, bovine stomach lysozyme RNA or any of the homologous genes in mice and man. The observed weak expression of this bovine mLys gene appears to be a consequence of both the presence of an extra ATG codon in the extended 5'-UTR, and a severe down mutation of the ancestral TATA-box which is only partially compensated for by the presence of another mutation further upstream resulting in a weak substitute promoter sequence.

Promoter elements in the influenza vRNA terminal structure.

The role of the partially double-stranded influenza vRNA terminal structure and its constitutive elements as a promoter signal was studied in vivo in a series of nucleotide substitution and insertion derivatives. A series of single and complementary double exchanges restoring intrastrand base pairing shows that a distal promoter element consists of a six-base pair double-stranded RNA rod in long-range complementary interaction. Within the distal element, all base pair positions are freely exchangeable, and hence no nucleotide-specific recognition could be identified. The proximal promoter element consists of nine partially complementary nucleotides at the vRNA 5' and 3' end. The nine plus six base pair panhandle rod of protein-free vRNA is interrupted by a central third element, a single unpaired nucleotide: adenosine 10 or various substitute residues, which appears to cause a bulged conformation in the overall structure. Mutagenization studies in the promoter proximal element indicate that, upon binding to polymerase, nucleotides at positions 2 and 3 interact with positions 9 and 8 within each branch (5' or 3') in short-range base pairing. In this conformation, the intermediate positions 4-7 are exposed as a single-stranded tetra-loop, which includes invariant guanosine residue 5 in the top conformational position of the 5' segment loop. Altogether, the three base paired segments in angular conjunction to each other adopt a conformation that is described in a "corkscrew model" for an activated stage of vRNA/polymerase interaction.