A temperature-sensitive phenotype of avian myeloblastosis virus: determinants that influence the production of viral mRNAs

Natural infection and transmission of a retrovirus closely related to myeloblastosis-associated virus type 1 in egg-type chickens

Myeloblastosis-associated virus type 1 (MAV-1) is an exogenous avian retrovirus with oncogenic potential. MAV-1 was detected in young chicks hatching from eggs produced by an experimental genetic line of egg-type chickens. Transmissibility of MAV-1 had not been documented previously. This investigation was intended to partially characterize the virus involved and to study its transmissibility and oncogenicity in naturally and contact-infected chickens. Commercially produced white and brown layer pullets free of exogenous avian leukosis viruses were commingled at hatch with naturally MAV-1-infected chickens. The original MAV-1-infected chickens were discarded after approximately 8 wk, and the contact-exposed chickens were maintained in isolation for 36 wk. Young specific-pathogen-free (SPF) single comb white leghorn chickens were added to the group to study possible horizontal transmission of MAV-1 in young chickens. Upon weekly virus isolation attempts, MAV-1 was readily isolated from the contact-exposed white layers but not from the brown layers between 36 and 53 wk of age (18 wk in total). Three-week-old SPF chickens were readily infected with MAV-1 by contact as early as 1 wk postexposure. Throughout 22 hatches derived from the white and brown MAV-1-contact-exposed layers (between 36 and 53 wk of age), MAV-1 was frequently detected in the white layer progeny, whereas the virus was seldom isolated from the progeny produced by the brown layers during the same 18-wk period. MAV-1 induced a persistent infection in some of the SPF chickens that were exposed by contact at 3 wk of age. Gross tumors were not detected in any of the originally infected experimental chickens at 8 wk of age, in the contact-exposed brown or white layers at the termination of the study at 53 wks of age, or in the contact-exposed SPF chickens at the end of the study at 12 wk of age. Exogenous avian leukosis-related viruses may still be detected in egg-type chickens, emphasizing the importance of thorough screening before incorporation of experimental genetic material into commercial genetic lines of egg-type chickens.

Transformation by v-Myb

The v-myb oncogene of the avian myeloblastosis virus (AMV) is unique among known oncogenes in that it causes only acute leukemia in animals and transforms only hematopoietic cells in culture. AMV was discovered in the 1930s as a virus that caused a disease in chickens that is similar to acute myelogenous leukemia in humans (Hall et al., 1941). This avian retrovirus played an important role in the history of cancer research for two reasons. First, AMV was used to demonstrate that all oncogenic viruses did not contain a single cancer-causing principle. In particular, although both Rous sarcoma virus (RSV) and AMV could replicate in cultures of either embryonic fibroblasts or hematopoietic cells, RSV could transform only fibroblasts whereas AMV could transform only hematopoietic cells (Baluda, 1963; Durban and Boettiger, 1981a). Second, chickens infected with AMV develop remarkably high white counts and therefore their peripheral blood contains remarkably large quantities of viral particles (Beard, 1963). For this reason AMV was often used as a prototypic retrovirus in order to study viral assembly and later to produce large amounts of reverse transcriptase for both research and commercial purposes. Following the discovery of the v-src oncogene of RSV and the demonstration that it arose from the normal c-src proto-oncogene, a number of acute leukemia viruses were analysed by similar techniques and found to also contain viral oncogenes of cellular origin (Roussel et al., 1979). In the case of AMV, it was shown that almost the entire retroviral env gene had been replaced by a sequence of cellular origin (initially called mab or amv, but later renamed v-myb) (Duesberg et al., 1980; Souza et al., 1980). Remarkably, sequences contained in this myb oncogene were shared between AMV and the avian E26 leukemia virus, but were not contained in any other acutely transforming retroviruses. In addition, the E26 virus contained a second sequence of cellular origin (ets) that was unique. The E26 leukemia virus was first described in the 1960s and causes an acute erythroblastosis in chickens, more reminiscent of the disease caused by avian erythroblastosis virus (AEV) than by AMV (Ivanov et al., 1962).

Reassessing the role of C-MYB in tumorigenesis

Hematopoietic tumors in both humans and mice frequently up-regulate expression of the c-myb gene, but it is unclear whether this is a cause or a consequence of the leukemic state. Recent results placing super-activation of the c-Myb protein at the bottom of a kinase-activated signal transduction pathway indicate that it may be a downstream effector of transformation induced by other oncogenes. The relationship between c-Myb and the serine-threonine kinase pim-1, its immediate activator, is discussed, together with the possibility that c-Myb, like pim-1, may be able to synergize with c-Myc to induce tumors.

Cells transformed by a v-Myb-estrogen receptor fusion differentiate into multinucleated giant cells

In order to make conditional alleles of the v-myb oncogene, we constructed and tested avian retroviruses which produce a number of different fusion proteins between v-Myb and the human estrogen receptor (ER). We found that the portion of the ER used in making these fusions profoundly affected their transcriptional activation. However, all the fusions tested were only weakly transforming in embryonic yolk sac assays and there was no direct correlation between the level of transcriptional activation and strength of oncogenic transformation. Nevertheless, transformation by a v-Myb-ER fusion was estrogen dependent, and upon withdrawal of the hormone, monocytic-lineage cells differentiated into multinucleated giant cells. Surprisingly, the withdrawal of estrogen caused a dramatic increase in the stability of the fusion protein, although it remained unable to promote cell growth or block differentiation.

The Moloney murine sarcoma virus ts110 5' splice site signal contributes to the regulation of splicing efficiency and thermosensitivity

The 5' splice site signal (5'ss) in Moloney murine sarcoma virus ts110 (MuSVts110) RNA was found to participate in the regulation of its splicing phenotype. This 5'ss (CAG/GUAGGA) departs from the mammalian consensus (CAG/GURAGU) at positions +4 and +6, both of which base pair with U1 and U6 small nuclear RNAs during splicing. A doubling in splicing efficiency and near elimination of the splicing thermosensitivity characteristic of MuSVts110 were observed in 5'ss mutants containing a U at position +6 (termed 5' A6U), even in those in which U1-5'ss complementarity had been reduced. At the permissive temperature (28 degrees C), the 5' A6U mutation increased the efficiency of the second splicing reaction, while at the nonpermissive temperature (39 degrees C), both splicing reactions were positively affected.

Branchpoint and polypyrimidine tract mutations mediating the loss and partial recovery of the Moloney murine sarcoma virus MuSVts110 thermosensitive splicing phenotype

Balanced splicing of retroviral RNAs is mediated by weak signals at the 3' splice site (ss) acting in concert with other cis elements. Moloney murine sarcoma virus MuSVts110 shows a similar balance between unspliced and spliced RNAs, differing only in that the splicing of its RNA is, in addition, growth temperature sensitive. We have generated N-nitroso-N-methylurea (NMU)-treated MuSVts110 revertants in which splicing was virtually complete at all temperatures and have investigated the molecular basis of this reversion on the assumption that the findings would reveal cis-acting elements controlling MuSVts110 splicing thermosensitivity. In a representative revertant (NMU-20), we found that complete splicing was conferred by a G-to-A substitution generating a consensus branchpoint (BP) signal (-CCCUGGC- to -CCCUGAC- [termed G(-25)A]) at -25 relative to the 3' ss. Weakening this BP to -CCCGAC- [G(-25)A,U(-27)C] moderately reduced splicing at the permissive temperature and sharply inhibited splicing at the originally nonpermissive temperature, arguing that MuSVts110 splicing thermosensitivity depends on a suboptimal BP-U2 small nuclear RNA interaction. This conclusion was supported by results indicating that lengthening the short MuSVts110 polypyrimidine tract and altering its uridine content doubled splicing efficiency at permissive temperatures and nearly abrogated splicing thermosensitivity. In vitro splicing experiments showed that MuSVts110 G(-25)A RNA intermediates were far more efficiently ligated than RNAs carrying the wild-type BP, the G(-25)A,U (-27)C BP, or the extended polypyrimidine tract. The efficiency of ligation in vitro roughly paralleled splicing efficiency in vivo [G(-25)A BP > extended polypyrimidine tract > G(-25)A,U(-27)C BP > wild-type BP]. These results suggest that MuSVts110 RNA splicing is balanced by cis elements similar to those operating in other retroviruses and, in addition, that its splicing thermosensitivity is a response to the presence of multiple suboptimal splicing signals.

Regulation of the efficiency and thermodependence of murine sarcoma virus MuSVts110 RNA splicing by sequences in both exons

Efficient splicing of MuSVts110 RNA is restricted to temperatures of 33 degrees or lower. Previously, we have shown that this conditional splicing event is mediated, in part, by cis-acting intronic sequences. We have now examined the role of exon sequences in MuSVts110 RNA splicing. We found that deletion of all but 36 nucleotides of the gag exon (E1) yielded a transcript incapable of supporting splicing. However, inefficient, growth temperature-dependent splicing was recovered after restoration of the 300 nucleotides of E1 proximal to the 5' splice site (5' ss). Increasingly efficient splicing was observed as more E1 was restored. Hence, although MuSVts110 E1 sequences were required for splicing, they were not involved in its thermodependence. Similarly, removal of all but 88 nucleotides of the mos exon (E2) abolished splicing at the usual 3' splice site (3' ss). In contrast to E1, restoration of the 200 nucleotides of E2 adjacent to the 3' ss reactivated efficient, temperature-independent splicing. Thermodependent splicing, however, reappeared with the replacement of E2 sequences located more than 400 nucleotides distal to the 3' splice site. In MuSVts110 mutants containing the minimum amounts of both E1 and E2 which would support splicing, splicing was both far more efficient than predicted and temperature-independent, suggesting that cooperation between E1 and E2 may help to regulate MuSVts110 splicing.

Extension of the DNA binding consensus of the chicken c-Myb and v-Myb proteins

The chicken c-myb gene and the v-myb oncogene transduced by avian myeloblastosis virus (AMV) encode DNA binding transcription activators. The DNA binding domain of AMV v-Myb displays a number of amino acid changes relative to c-Myb; v-Myb proteins in which one or more of three crucial residues in the DNA binding domain are mutated to resemble the c-Myb sequence display altered transformation phenotypes. In order to establish whether the spectrum of DNA binding sites which AMV v-Myb can recognise is different from that seen by chicken c-Myb, a site selection protocol was used to derive consensus binding sequences for three variant Myb proteins made in vitro, and also using nuclear extract from the v-myb transformed cell line BM2. The results show that the original consensus binding site defined for v-Myb, YAA-CKG, can be extended to YAACKGHH, and that this new consensus holds for both v-Myb and chicken c-Myb.

A single nucleotide substitution in the 5'-untranslated region of the vaccinia N2L gene is responsible for both alpha-amanitin-resistant and temperature-sensitive phenotypes

The locus responsible for encoding resistance to alpha-amanitin was previously mapped to the vaccinia virus (VV) HindIII N fragment by using cloned wild-type VV DNA fragments to rescue the ability of an alpha-amanitin-resistant/temperature-sensitive VV mutant (alpha rts7) to replicate under nonpermissive conditions. DNA sequencing and transcriptional analyses of this region identified two leftward-reading open reading frames (ORFs), N2L and M1L, as candidates to encode the protein responsible for eliciting both phenotypes. In the present study, high-resolution marker rescue mapping and genomic sequencing techniques have been applied to identify the nature of the mutation within the HindIII N region of the alpha rts7 genome. Interestingly, a single G to T transversion mutation was noted at position -10 relative to the initiator ATG of the N2L ORF. Since transcription of the N2L gene starts at position -12/-13, this places the alpha rts7 mutation within the 5'-untranslated leader of the N2L transcript expressed early in infection and suggests that the transcriptional efficiency, mRNA stability, or translational efficiency must be altered in the mutant RNA. These results identify the N2L ORF as the gene responsible for conferring resistance to alpha-amanitin in the alpha rts7 mutant and suggest that the N2L gene product is the viral function that interacts with the host cell nucleus during VV infection.

Nickel mutagenesis: alteration of the MuSVts110 thermosensitive splicing phenotype by a nickel-induced duplication of the 3' splice site

We investigated DNA damage caused by carcinogenic metals in a murine sarcoma virus (MuSV)-based mutagenicity assay in which mutations targeted to v-mos expression can be selected. Nickel chloride treatment of NRK cells (termed 6m2 cells) infected with MuSVts110, a retrovirus conditionally defective in viral RNA splicing and cell transformation, caused the outgrowth of transformed "revertants" with changes in the MuSVts110 RNA splicing phenotype. Cadmium and chromium treatment of 6m2 cells resulted in the selection of a second class of revertants with what appeared to be frameshift mutations allowing the translation of a readthrough gag-mos protein. In both classes of metal-induced revertants, viral gene expression was distinct from that observed in revertants arising in untreated 6m2 cultures, arguing that metal treatment did not simply enhance the rate of spontaneous reversion. In one representative nickel revertant line the operative nickel-induced mutation affecting MuSVts110 RNA splicing was a duplication of 70 bases surrounding the 3' splice site. The effect of this mutation was to direct splicing to the most downstream of the duplicated 3' sites and concomitantly relax its characteristic thermosensitivity. These data establish the mutagenic potential of nickel and provide the first example of a defined nickel-induced mutation in a mammalian gene.

Ectopic expression of the erythrocyte band 3 anion exchange protein, using a new avian retrovirus vector

A retrovirus vector was constructed from the genome of avian erythroblastosis virus ES4. The v-erbA sequences of avian erythroblastosis virus were replaced by those coding for neomycin phosphotransferase, creating a gag-neo fusion protein which provides G418 resistance as a selectable marker. The v-erbB sequences following the splice acceptor were replaced by a cloning linker allowing insertion of foreign genes. The vector has been tested in conjunction with several helper viruses for the transmission of G418 resistance, titer, stability, transcription, and the transduction and expression of foreign genes in both chicken embryo fibroblasts and the QT6 quail cell line. The results show that the vector is capable of producing high titers of Neor virus from stably integrated proviruses. These proviruses express a balanced ratio of genome length to spliced transcripts which are efficiently translated into protein. Using the Escherichia coli beta-galactosidase gene cloned into the vector as a test construct, expression of enzyme activity could be detected in 90 to 95% of transfected target cells and in 80 to 85% of subsequently infected cells. In addition, a cDNA encoding the avian erythrocyte band 3 anion exchange protein has been expressed from the vector in both chicken embryo fibroblasts and QT6 cells and appears to function as an active, plasma membrane-based anion transporter. The ectopic expression of band 3 protein provides a visual marker for vector function in these cells.

Activation of cryptic splice sites in murine sarcoma virus-124 mutants

We have examined splice site activation in relation to intron structure in murine sarcoma virus (MuSV)-124 RNA. MuSV-124 contains inactive murine leukemia virus env gene splice sites (termed 5' env and 3' env) as well as cryptic sites in the gag and v-mos genes (termed 5' gag and 3' mos) which are activated for thermosensitive splicing by a 1,487-base intronic deletion in the MuSV-124 derived MuSVts110 retrovirus. To determine conditions permissive for splice site activation, we examined MuSV-124 mutants deleted in the 1,919-base intron bounded by the 5' gag and 3' mos sites. Several of these deletions activated thermosensitive splicing either at the same sites used in MuSVts110 or in a previously unreported temperature-sensitive splice event between the 5' gag and 3' env sites. These data suggested that the thermosensitive splicing phenotype characteristic of MuSVts110 required neither a specialized intron nor selection of a particular 3' splice site. The 3' env and 3' mos sites were found to compete for splicing to the 5' gag site; the more upstream 3' env site was exclusively used in MuSV-124 mutants containing both sites, whereas selection of the 3' mos site required removal of the 3' env site. Branchpoint sequences were found to have a potential regulatory role in thermosensitive splicing. Insertion of a beta-globin branchpoint sequence in a splicing-inactive MuSV-124 mutant activated efficient nonthermosensitive splicing at the 3' mos site, whereas a mutated branchpoint activated less efficient but thermosensitive splicing.