Transduction is the process by which a retrovirus captures a cellular gene. Repeated passages of an oncogene-free retrovirus on quiescent cultures of embryonic avian neuroretin cells frequently give rise to new mitogenic viruses which have transduced oncogenes of the serine / threonine kinase mil / raf family. The molecular intermediaries which follow one another during the various stages of the capture of oncogenes are:
- a chimeric transcript, comprising viral and cellular sequences fused by alternative splicing;
- a complete retrovirus whose 5 ‘end is identical to that of the preceding RNA;
- finally, a retrovirus which has acquired additional gag sequences and whose replicative capacity is increased.
The model suggested by these studies, which makes the chimeric RNA the key step in the process, makes it possible to explain the genesis of most of the transforming retroviruses isolated in vivo.
Let us first remember
The capture of cellular genes by retroviruses, called retroviral transduction, is the natural process which has allowed amputation, fusion with viral genes, overexpression) are responsible for the malignant transformation of cells infected with the recombinant virus. Transduction of an oncogene is generally accompanied by inactivation by deletion of one or more genes (called gag, pol and env) for replication of the virus. The spread of such viruses discovered oncogenes. The oncogenes transduced by retroviruses, called v-one, have undergone modifications compared to the cellular protooncogenes (c-onc) from which they are derived.
These alterations (mutation, therefore requires the presence of a helper virus, which is generally the original virus. The vast majority of viruses which have transduced an oncogene have been isolated from tumors which have arisen in various animal species (hen, mouse, cat, etc.), which appeared after multiple infections.
Consequently, the mechanisms of transduction, and more particularly the early events thereof, have only been the subject of hypotheses which are difficult to verify in these systems in vivo We have developed a biological system which has made it possible to study in cell culture the mechanisms of transduction of oncogenes, in particular by identifying the molecular intermediates involved in the genesis of a recombinant virus.
Transduction of mil / raf and R mil / 8- raf oncogenes into neuroretin cells
In our biological system, the chosen host cell is the hen embryo neuroretin (NR) cell. Neuroretin derives from the neuroectoderm and constitutes a functional entity of the central nervous system specialized in the processing of the light signal. NR cells collected on the eighth day of embryonic development and cultured quickly stop dividing and express different markers of neuronal differentiation. These cells rapidly and massively start to proliferate after infection with retroviruses carrying oncogenes.
The expression of the oncogene present in the genome of the retrovirus is necessary for the induction and maintenance of this cellular multiplication. RA V-1 (Rous-associated virus type 1), an oncogene-free avian lymphomatosis virus, can also induce the proliferation of NR cells in culture. However, this is only observed in a small number of cells and after a certain latency time. When the supernatants of induced cultures are repeatedly passed over to proliferate on new cultures, an induction of proliferation is observed more and more rapid and massive.
This phenomenon is linked to the appearance of viruses which have transduced an oncogene, reproducibly and with a high frequency. Molecular cloning of recombinant viruses (called IC for Institut Curie), which appeared during several independent experiments, made it possible to show that RA V-1 systematically transduced either the e-mil gene or the c-Rmil gene (respectively called c-raf and B-raf in mammals).
These two genes belong to the family of mil / raf oncogenes encoding serine / threonine protein kinases involved in the transmission of division or differentiation signals from the cell membrane to the nucleus. It has recently been shown that the c-raf gene product, a prototype of the family, can interact with the Ras protein and activate the “MAP kinase cascade”. As for the c-Rmil gene, it was previously unknown and was identified during these experiments.
The relationship between these two genes is found not only at the level of the protein sequence but also at the level of their genomic organization, very similar . Thus, in all IC viruses, the recombination between viral and cellular sequences took place in a region homologous to the e-mil and c-RmiL genes. The transduced portion of these genes corresponds to the catalytic domain containing the serine / threonine protein activity. kinase. Therefore, the deletion of the regulatory amino-terminal part of these proteins results in the constitutive activation of the kinase domain present in IC viruses. One of the advantages of this biological system is the possibility of analyzing the evolution of the viruses isolated at the various passages of the culture supernatants during the same experiment.
Indeed, we have characterized two types of virus: one appeared after a limited number of passages (early viruses) and the other generated after multiple passages (late viruses). Molecular cloning and the establishment of the nucleotide sequence have shown that these two types of viruses had different genomic organizations. Early viruses express the oncogene as a fusion protein comprising only the first six amino acids of the Gag protein and the amino acids encoded by the last ten exons of the emil or c-RmiL genes In contrast, late viruses express a protein comprising a greater number of amino acids of the Gag protein, fused to a shorter oncogene sequence than that present in early viruses.
Analysis of the nucleotide sequences at the 5 ‘(virusoncogene) and 3’ (oncogene-virus) recombination points between the viral and cell sequences showed that the 3 ‘junction is identical for the early virus and the late virus produced during ‘same experience. In all cases, the viral and cellular sequences present significant analogies (covering a region varying from 6 to 25 nucleotides) around the junction point. In contrast, the 5 ‘junction of the early virus differs from that of the late virus. In early virus, the leader splice donor site of RA V-1 is directly linked, respectively, to the splice acceptor site of exons 8 or 9, of the e-mil or c-Rmil genes. The leader of RAV-1 is the sequence common to the genomic (encoding the Gag and Pol proteins) and subgenomic (encoding the Env protein) transcripts (figure 1). It extends from the 5 ‘end of the R sequence (for repetition) to the donor site involved in the alternative splicing which generates the subgenomic RNA.
The last eighteen nucleotides of the leader encode the first six amino acids of the Gag protein and the Env protein, and provide the initiator methionine for the proteins expressed by early viruses. The location of the splicing donor site downstream of the initiation codon of the Gag and Env proteins is characteristic of avian lymphomatosis viruses. In the late virus, the 5 ′ recombination takes place at the level of zones of analogies (extending over ten nucleotides) between the gag sequences of the RA V-1 and the sequences of the oncogene present on the early virus . In summary, the mechanisms at the origin of the 5 ‘junction are different for early and late viruses while their 3’ junction is identical. All of these data suggest that there is a filiation between early and late viruses: in other words, late viruses result from recombination between an early virus and RAV-1. This hypothesis was verified in experiments of cotransfection of an early virus and RA V-1 in NR cells.
Indeed, the repeated passages of the supernatants of these cultures systematically lead to the genesis of a virus having acqws additional gag sequences. The genesis of early viruses, meanwhile, involves a splicing mechanism between viral and cellular sequences. This recombination would therefore take place at the level of the RNA, by an alternative splicing.
The synthesis of a chimeric virus-oncogene RNA is a key step in transduction
Analysis of early viruses suggested that the 5 ‘junction occurs through splicing at an RNA molecule. We therefore sought the presence of chimeric transcripts comprising the leader sequence of RA V-1 and sequences of the e-mil or c-Rmil genes, in cultures of NR cells induced to proliferate by RAV-1.
A DNA (complementary DNA) containing leader sequences fused to sequences of the R mil oncogene was cloned by the technique of amplification by PCR (polymerase chain reaction) associated with reverse transcription. Its sequence has been compared with that of the early virus ICl l and that of the eDNA of the c-Rmil gene. The corresponding RNA, called R5Rmil, therefore has the leader-Rmil-polyA structure, the 5 ‘junction taking place, as for the IC ll virus, with exon 9.
R5Rmil therefore has an open reading phase comprising the codons for the first six residues of the Gag protein, fused to the codons for the last 387 residues of the c-Rrnil protein. Analysis of this molecule demonstrates that there is indeed an alternative splicing between viral sequences and cellular sequences in cells induced to proliferate by RAV-I. Similar experiments indicate that such chimeric transcripts comprising sequences either of the e-mil proto-oncogene or of the c-Rmi l protooncogene are present in all cultures infected with RAV-1, even before the appearance of cell division foci. This suggests that leader-mif: .polyA or leader-Rmil-polyA transcripts could be involved in the induction of NR cell division and in the genesis of early viral forms. To verify these two points, we transfected a construct capable of generating the R5Rmil transcript in NR cell cultures.
These cultures are rapidly induced to proliferate, which demonstrates the mitogenic property of the transcript. One of these cultures was infected with RA V-1 and then its supernatant used to infect new cultures. These are massively induced to proliferate and the analysis of their genomic content shows the presence of a complete virus carrying Rmil sequences. The leader-Rmil-polyA chimeric RNA can, therefore, recombine with the RA V-1 to generate a full length virus. In the transduction process, such a transcript therefore represents a key molecule that our system has made it possible to isolate.
Mecanisms of oncogene transduction by retroviruses
The results presented in the two preceding paragraphs have enabled us to propose a mechanism for the transmission of oncogenes. It is generally accepted that the integration of the provirus close to the oncogene is the first step in the activation and capture of oncogenes. Thus, in the NR cell, the RAV-1 would integrate into the e-mil or c-Rmil locus, in the same transcriptional orientation as the protooncogene. The transcription of RAV-1, which begins in the 5 ‘LTR (long terminal repeat) region, would be the source of certain RNAs, called readthroughs, comprising viral sequences followed by cellular sequences located downstream of the provirus.
Indeed, Herman et al. described that about 15% of RAV-1 transcripts escape the cleavage and polyadenylation which normally takes place in the 3 ‘LTR sequence. Alternative splicing of such a readthrough transcript between the RAV-1 donor site and an e-mil or c-Rmil oncogene acceptor site would generate a chimeric leader-ontrpolyA molecule similar to that described in the previous paragraph. Some viruses which have transduced another oncogene, the src gene, have a 5 ‘junction resulting from splicing, similar to that of early IC viruses.
Furthermore, virus-one-polyA transcripts, resulting from alternative splicing, have been isolated in avian and murine systems, although in the latter the synthesis of retroviruses from readthrough transcripts has not been demonstrated. Therefore, splicing between a viral donor site and an oncogene acceptor site could represent a general mechanism of formation of the 5 ‘junction during transduction. Obtaining a complete retrovirus from a chimeric transcript requires the acquisition of a 3 ‘retroviral end including an LTR sequence.
We have shown that for all IC viruses there are analogies between viral and cellular sequences at the junction point. This suggests that 3 ‘recombination results from a jump in the reverse transcriptase template, from RAV-1 RNA to the chimeric transcript. Indeed, the two transcripts, provided to the leader sequence (which contains the encapsidation signal , are known sceptibles be coencaps1des, this qw allow a start of reverse transcription so that the two molecules are still associated. Experiments involving in play viral constructs which carry genes for resistance to antibiotics as selection markers, have confirmed that reverse transcription is a process favorable to recombinations between molecules exhibiting sequence analogies.
Our system has also made it possible to to characterize the viral forms which follow one another during the transduction of oncogenes We have shown that the late viruses result from the recombination between the early virus and the RAV-1 at the level of sequence analogies between the oncogene and the gag gene These late viruses therefore have a selective advantage thanks to the increased replicative and transforming properties conferred on them by the acquisition of s additional gag equations. The vast majo The nature of the transforming retroviruses expresses a Gag-one fusion protein, thus suggesting that our model may account for the transduction of other proto-oncogenes.
In conclusion, the work presented here made it possible to dissect the stages of transduction of the emil and c-Rmil genes. The crucial step in the process is the alternative splicing between viral and cellular sequences, giving rise to hybrid RNAs, themselves at the origin of the early viral forms. The same events are probably involved in the transduction of many oncogenes .