Detection of Ki-ras proto-oncogene protein by a specific monoclonal antibody.

We have used a commercially available mouse monoclonal antibody and shown it to bind specifically to cellular Ki-ras p21 proteins and not to cellular N- and Ha-ras p21 proteins. In conjunction with electrophoresis and Western blotting, this antibody can be used, with further detailing, to assess levels of the cellular Ki-ras p21 against a background of total p21s.     

A human gastric carcinoma contains a single mutated and an amplified normal allele of the Ki-ras oncogene.

The DNA from various human tumors and tumor cell lines was screened for the presence of mutated ras oncogenes with synthetic oligonucleotide probes, as well as with the NIH/3T3 cell transfection assay. Among the various mutations found we discovered two novel Ki-ras mutations in codon 12: gly to ala and gly to ser. A gastric carcinoma was found to possess a single mutated Ki-ras allele (gly-12 to ser), as well as a 30-50 fold amplified normal allele. This implies that two activating steps must have occurred in this malignancy.     

Amplification and rearrangement of Ki-ras oncogene in human teratocarcinoma-derived cell lines

The structure of the ras gene family was analyzed in two human teratocarcinoma-derived cell lines, Tera I and Tera II, by DNA restriction enzyme digestion and Southern blot. We report here a ten-fold amplification of the c-Ki-ras-2 gene in these cell lines, whereas no structural alterations seem to occur either in c-Ha-ras-1 or N-ras. We also provide evidence indicating that no point mutation at codon 12, specifically recognized by Sac I, was detected. Moreover, DNA rearrangement, due to the loss of a Pvu II site located in the intervening sequences between the third and the fourth exon, has been found in both Tera I and Tera II.     

Oncogenicity of human N-ras oncogene and proto-oncogene introduced into retroviral vectors.

The N-ras gene is the only member of the ras family which has never been naturally transduced into a retrovirus. In order to study the in vitro and in vivo oncogenicity of N-ras and to compare its pathogenicity to that of H-ras, we have inserted an activated or a normal form of human N-ras cDNA into a slightly modified Harvey murine sarcoma virus-derived vector in which the H-ras p21 coding region had been deleted. The resulting constructions were transfected into NIH 3T3 cells. The activated N-ras-containing construct (HSN) induced 10(4) foci per microgram of DNA and was found to be as transforming as H-ras was. After infection of the transfected cells by either the ecotropic Moloney murine leukemia virus or the amphotropic 4070A helper viruses, rescued transforming viruses were injected into newborn mice. Both pseudotypes of HSN virus containing activated N-ras induced the typical Harvey disease with similar latency. However, we found that the virus which contained normal N-ras p21 (HSn) was also pathogenic and induced splenomegaly, lymphadenopathies, and sarcoma in mice after a latency of 3 to 7 weeks. In addition, Moloney murine leukemia virus pseudotypes of N-ras caused neurological disorders in 30% of the infected animals. These results differed markedly from those of previous experiments in which we had inserted the activated form of N-ras in the pSV(X) vector: the resulting SVN-ras virus was transforming on NIH 3T3 cells but was poorly oncogenic in vivo (M. Souyri, C. F. Koehne, P. V. O'Donnel, T. H. Aldrich, M. E. Furth, and E. Fleissner, Virology 158:69-78). However, similarly poor oncogenicity was also observed when the v-H-ras coding sequence was inserted in pSV(X) vector, which indicated that the vector sequences play a crucial role in the pathogenicity of a given oncogene. Altogether, these data demonstrated unequivocally that N-ras is potentially as oncogenic as H-ras and that such oncogenic effect could depend on the vector environment.     

Nucleic acid sequences of the oncogene v-rel in reticuloendotheliosis virus strain T and its cellular homolog, the proto-oncogene c-rel.

Reticuloendotheliosis virus strain T (Rev-T) is a highly oncogenic replication-defective retrovirus which contains the oncogene v-rel. It is thought that Rev-T arose when a virus similar to Rev-A, the helper virus of Rev-T, infected a turkey and recombined with c-rel from that turkey. There is one large c-rel locus in the turkey genome which contains all of the sequences homologous to v-rel (K. C. Wilhelmsen and H. M. Temin, J. Virol. 49:521-529, 1984). We have sequenced v-rel and its flanking sequences, each of the regions of the c-rel locus from turkey that are homologous to v-rel and their flanking sequences, and the coding sequence for env and part of pol of Rev-A. The v-rel coding sequences can be translated into a 503-amino acid env-v-rel-out-of-frame-env fusion polypeptide. We have not detected any sequences in the Los Alamos or University of California-San Diego data bases that are more significantly related to the amino acid or nucleic acid sequence of v-rel than to the randomized sequence of v-rel. Comparison of Rev-A, Rev-T, and c-rel indicates that the v-rel sequences may have been transduced from the c-rel (turkey) locus by a novel mechanism. There are sequences in Rev-A and c-rel that are similar to splicing signals, indicating that the 5' virus-rel junction of Rev-T may have been formed by cellular RNA splicing machinery. Eight presumed introns have presumably been spliced out of c-rel to generate v-rel. There are also short imperfect regions of homology between sequences at the boundaries of v-rel and sequences in Rev-A and c-rel (turkey), indicating that c-rel may have been transduced by homologous recombination. There are many differences between the amino acid sequences of the predicted translational products of v-rel and c-rel which may account for their difference in transformation potential. These sequence differences between v-rel and c-rel include 10 missense transitions, four missense transversions, and three places where Rev-T has a small in-frame deletion of sequences relative to c-rel. Most of the coding sequence differences between c-rel and v-rel are nonconservative amino acid changes.     

Human proto-oncogene c-mos maps to 8q11.

The c-mos proto-oncogene is the cellular counterpart of the viral oncogene v-mos isolated from Moloney murine sarcoma virus. The c-mos gene locus has previously been assigned to human chromosome 8. By both in situ hybridization and molecular hydridization to sorted chromosome DNA (using a c-mos probe) we have localized the c-mos gene to band 8q11. This regional localization is at variance with the one previously reported at 8q22 and may explain why no rearrangement of c-mos has been found in acute leukaemia with the chromosomal translocation t(8;21)(q22;q22).     

Mitogenic effects of the proto-oncogene and oncogene forms of c-H-ras DNA in human diploid fibroblasts.

Nuclear microinjection of c-H-ras DNA induced DNA synthesis in reversibly nonproliferating quiescent human cells. The proto-oncogene and oncogene forms were equally effective inducers. In contrast, c-H-ras DNA either alone or in combination with the adenovirus E1A gene did not cause terminally nondividing senescent cells to synthesize DNA.     

Human recipient cell for oncogene transfection studies.

We used human oncogene DNA to transform the nontumorigenic, revertant, human osteosarcoma cell line HOS TE-85 clone 5 (ATCC CRL 1543) to tumorigenicity in athymic nude mice with latency periods as short as 3 weeks. These cells were also transformed by genetic markers in genomic DNA samples. Because of their low rate of spontaneous tumor formation and the simplicity of culturing them, HOS cells provide a human cell alternative to NIH 3T3 murine fibroblasts for oncogene transfection studies.     

Structure of the feline c-fes/fps proto-oncogene: genesis of a retroviral oncogene.

The nucleotide sequence of the feline c-fes/fps proto-oncogene was analyzed. Comparison with v-fes and v-fps revealed that all v-fes/fps homologous sequences were dispersed over 11 kilobase pairs in 19 interspersed segments. All segments, numbered exon 1 to exon 19 as in the chicken and human loci, were flanked by consensus splice junctions. The putative promoter region contained a CATT sequence and three CCGCCC motifs which were also found in the human locus at similar positions. About 200 nucleotides downstream of a translational stop codon in exon 19, a putative poly(A) addition signal was identified. Using the putative translation initiation codon in exon 2, a 93,000-molecular-weight protein could be deduced. This protein resembled very well the putative protein of the human c-fes/fps proto-oncogene (94% overall homology) and, although less well, the putative protein of the chicken c-fes/fps proto-oncogene (70% overall homology). As far as the feline c-fes/fps proto-oncogene sequences transduced to the Gardner-Arnstein (GA) and Snyder-Theilen (ST) strains of feline sarcoma virus (FeSV) are concerned, homology in deduced amino acid sequences between the GA- and ST-v-fes viral oncogenes and the proto-oncogene was 99%. Analysis of the recombination junctions between feline leukemia virus and v-fes sequences in GA- and ST-FeSV proviral DNA revealed for the left-hand junction the involvement of homologous recombination, presumably at the DNA level. The right-hand junction, which appeared identical in the GA-FeSV and ST-FeSV genomes, could have been the result of a site-specific recombination at the RNA level.