The structure of the human c-fes/fps proto-oncogene.

We have determined the complete nucleotide sequence of a human DNA fragment of approximately 13 kbp, which was shown by Southern blot analysis to contain the entire v-fes/fps cellular homolog. The v-fes/fps homologous sequences were dispersed over 11 kbp in 18 interspersed segments which were flanked by splice junctions. Fusion of these segments created a DNA fragment in which coding regions similar to those observed in the viral oncogenes v-fes of the Gardner-Arnstein (GA) and Snyder-Theilen (ST) strains of feline sarcoma virus and v-fps found in Fujinami sarcoma virus could be identified. A potential initiation site in the first exon was found. About 200 nucleotides downstream of a translational stop codon in the v-fes/fps homologous region, a poly(A) addition signal was identified. The deduced amino acid sequence has a molecular weight of 93 390 dalton resembling NCP92, the recently described human c-fes/fps product. The topography of human c-fes/fps appeared to resemble that of chicken c-fps.     

Molecular cloning of the feline c-fes proto-oncogene and construction of a chimeric transforming gene

The feline c-fes proto-oncogene, different parts of which were captured in feline leukemia virus (FeLV) to generate the transforming genes (v-fes) of the Gardner-Arnstein (GA) strain of feline sarcoma virus (FeSV) and the Snyder-Theilen strain (ST) of FeSV, was cloned and its genetic organization determined. Southern blot analysis revealed that the c-fes genetic sequences were distributed discontinuously and colinearly with the v-fes transforming gene over a DNA region of around 12.0 kb. Using cloned c-fes sequences, complementation of GA-FeSV transforming activity was studied. Upon replacement of the 3' half of v-fesGA with homologous feline c-fes sequences and transfection of the chimeric gene, morphological transformation was observed. Immunoprecipitation analysis of these transformed cells revealed expression of high Mr fusion proteins. Phosphorylation of these proteins was observed in an in vitro protein kinase assay, and tyrosine residues appeared to be involved as acceptor amino acid.     

Isolation and characterization of a human locus homologous to the transforming gene (v-fes) of feline sarcoma virus

A single locus (designated c-fes) in the human genome which exhibits homology to the transformation-specific onc gene (v-fes) of Snyder-Theilen feline sarcoma virus was identified by the Southern blot technique. Recombinant clones containing 16- to 18-kilobase inserts of human DNA including the c-fes locus were constructed. Restriction endonuclease mapping of these clones verified their identity with native human c-fes and demonstrated the presence of at least two sequences in human c-fes interrupting v-fes-homologous regions. The v-fes-homologous locus in the human genome spans about 4 kilobases. The 5'-3' orientation of the c-fes clones with respect to feline sarcoma virus proviral DNA was determined. The region of the human genome that is homologous to v-fes is proximal to the highly reiterated human Alu sequence but not to the highly reiterated human alphoid sequence.     

Human gene (c-fes) related to the onc sequences of Snyder-Theilen feline sarcoma virus.

The onc gene (v-fes) of the acutely transforming feline sarcoma virus (Snyder-Theilen strain) has homologous cellular sequences (c-fes) in all vertebrate species, including humans. We isolated from a human DNA library recombinant phages containing overlapping c-fes sequences. The human c-fes locus spans a region of 3.4 kilobases and contains 1.4 kilobases of DNA homologous to the viral onc sequence interspersed with three intervening sequences.     

Biochemical and immunological characterization of polyproteins coded for by the McDonough, Gardner-Arnstein, and Snyder-Theilen strains of feline sarcoma virus.

The McDonough (SM), Gardner-Arnstein (GA), and Snyder-Theilen (ST) strains of feline sarcoma virus (FeSV) code for high-molecular-weight polyproteins that contain varying amounts of the amino-terminal region of the FeLV gag gene-coded precursor protein and a polypeptide(s) of an as yet undetermined nature. The SM-FeSV primary translational product is a 180,000-dalton polyprotein which is immediately processed into a highly unstable 60,000-dalton molecule containing the p15-p12-p30 fragment of the FeLV gag gene-coded precursor protein and a 120,000-dalton FeSV-specific polypeptide. The GA- and ST-FeSV genomes code for polyproteins of 95,000 and 85,000 daltons, respectively, which in addition to the amino-terminal moiety (p15-12 and a portion of p30) of the FeLV gag gene-coded precursor protein also contain FeSV-specific polypeptides. However, the GA- and ST-FeSV polyproteins appear to be relatively stable molecules (half-lives of around 16 h) and are not significantly processed into smaller polypeptides. Immunological and biochemical analysis of each of the above FeSV translational products revealed that the sarcoma-specific regions of the GA- and ST-FeSV polyproteins are antigenically cross-reactive and exhibit common methionine-containing peptides. These findings favor the concept that these sarcoma-specific polypeptides are coded for by the similar subsets of cellular sequences incorporated into the GA- and ST-FeSV genomes during the generation of these transforming agents.     

McDonough feline sarcoma virus: characterization of the molecularly cloned provirus and its feline oncogene (v-fms).

The genetic structure of the McDonough strain of feline sarcoma virus (SM-FeSV) was deduced by analysis of molecularly cloned, transforming proviral DNA. The 8.2-kilobase pair SM-FeSV provirus is longer than those of other feline sarcoma viruses and contains a transforming gene (v-fms) flanked by sequences derived from feline leukemia virus. The order of genes with respect to viral RNA is 5'-gag-fms-env-3', in which the entire feline leukemia virus env gene and an almost complete gag sequence are represented. Transfection of NIH/3T3 cells with cloned SM-FeSV proviral DNA induced foci of morphologically transformed cells which expressed SM-FeSV gene products and contained rescuable sarcoma viral genomes. Cells transformed by viral infection or after transfection with cloned proviral DNA expressed the polyprotein (P170gag-fms) characteristic of the SM-FeSV strain. Two proteolytic cleavage products (P120fms and pp55gag) were also found in immunoprecipitates from metabolically labeled, transformed cells. An additional polypeptide, detected at comparatively low levels in SM-FeSV transformants, was indistinguishable in size and antigenicity from the envelope precursor (gPr85env) of feline leukemia virus. The complexity of the v-fms gene (3.1 +/- 0.3 kilobase pairs) is approximately twofold greater than the viral oncogene sequences (v-fes) of Snyder-Theilen and Gardner-Arnstein FeSV. By heteroduplex, restriction enzyme, and nucleic acid hybridization analyses, v-fms and v-fes sequences showed no detectable homology to one another. Radiolabeled DNA fragments representing portions of the two viral oncogenes hybridized to different EcoRI and HindIII fragments of normal cat cellular DNA. Cellular sequences related to v-fms (designated c-fms) were much more complex than c-fes and were distributed segmentally over more than 40 kilobase pairs in cat DNA. Comparative structural studies of the molecularly cloned proviruses of Synder-Theilen, Gardner-Arnstein, and SM-FeSV showed that a region of the feline-leukemia virus genome derived from the pol-env junction is represented adjacent to v-onc sequences in each FeSV strain and may have provided sequences preferred for recombination with cellular genes.     

Nucleotide sequence and topography of chicken c-fps. Genesis of a retroviral oncogene encoding a tyrosine-specific protein kinase

We isolated molecular clones of chicken DNA that carry portions of the cellular proto-oncogene c-fps and then determined the nucleotide sequence of all regions of the gene that are related to the retroviral oncogene v-fps. The homology of v-fps within c-fps resides on at least 19 interspersed segments, 17 of which represent complete exons and two of which may represent only portions of exons. Fusion of these segments reconstructs a facsimile of v-fps. The arrangement of introns and exons within c-fps differs from that of the related proto-oncogene c-src in the domains of the two genes that encode tyrosine-specific protein kinase activity. It therefore appears likely that the introns arose subsequent to the gene duplication that engendered c-src and c-fps. The data also reveal potential junctions between viral and cellular domains in the genomes of two independently isolated avian sarcoma viruses (the PRCII and Fujinami strains). The lefthand junctions can be well defined: they occur at the same position in c-fps but at different positions in the viral gene gag. The righthand junctions cannot be defined as precisely because they include a sequence of 10 to 15 nucleotides whose origin is not known. In the genome of PRCII virus, the composition of this sequence suggests that it arose from the polyadenylated 3' terminus of the c-fps messenger RNA. If this deduction proves to be correct, the data will provide direct evidence that the righthand recombination during transduction by retroviruses occurs between RNA intermediates. Irrespective of these ambiguities, both junctions are located within exons of c-fps, and both may have been formed by non-homologous recombination (although the evidence for the latter statement is not decisive). A sequence of 1020 nucleotides has been deleted from the transduced version of c-fps in the genome of PRCII virus, apparently by homologous recombination between sequences repeated within c-fps. Fujinami virus may contain the entire coding domain of c-fps, but mutations have created 26 amino acid substitutions in the viral version of the gene. By contrast, the partially deleted version of c-fps in PRCII virus contains no mutations that would alter the amino acid sequence.     

Cellular transformation by subgenomic feline sarcoma virus DNA

The genome of the Snyder-Theilen strain of feline sarcoma virus (ST-FeSV) is a 4.3-kilobase-pair (kbp) RNA molecule that contains a 1.5-kbp cellular insertion (fes gene) flanked by feline leukemia virus sequences at its 5' end (1.6 kbp) and 3' end (1.2 kbp) (Sherr et al., J. Virol. 34:200-212, 1980). DNA transfection techniques have been utilized to determine the regions of the ST-FeSV genome involved in malignant transformation. I have found that the 3.7-kbp 5'-end fragment of the ST-FeSV provirus (which corresponds to the 3.4-kbp 5'-end fragment of the viral genome) is sufficient to transform NIH/3T3 fibroblasts. Enzymes that cleave the ST-FeSV provirus DNA within the feline leukemia virus gag gene sequences or within the fes gene abolished the transforming activity. Preservation of the proviral large terminal repeats was also required for transformation. Transformed NIH/3T3 cells obtained by transfection of total or subgenomic ST-FeSV DNA expressed normal levels of the ST-FeSV gene product ST P85 and of its associated protein kinase activity. Furthermore, these cells contained high levels of phosphotyrosine residues, a biochemical marker associated with cellular transformation induced by certain retroviruses including ST-FeSV. These results, taken together, strongly support the concept that only those ST-FeSV proviral sequences necessary for ST P85 expression are involved in malignant transformation.     

Structure of a gene for rat calmodulin

The structural organization of the entire rat calmodulin gene was determined by cloning and sequencing overlapping genomic and cDNA clones from rat genomic and brain cDNA libraries. The intron/exon organization was determined by direct comparison of these sequences. Rat calmodulin gene is 9000 bases long and consisted of six exons interrupted by introns of variable sizes. The first intron separates the initiation codon (ATG) from the coding region of the protein. Three out of four intron/exon junctions in the coding region reside in the middle of calcium binding subdomains and do not correlate with the quarterly divided intramolecular homology of the protein. Their positions exactly coincide with those of the corrected version of chicken calmodulin gene. The rat calmodulin gene harbors a stretch of sequences homologous to a rat middle repetitive "identifier sequence" in the middle of the third intron. Analysis of the immediate 5' upstream region detected a TATA box (TATATATAT) and three C-G boxes (CCGCCC) but not a CAT box (CCAAT). A conserved sequence (GCGCCGCGYCYYGGGGGC) was found at -125 for rat and at -204 for chicken calmodulin genes.     

Nucleotide sequence of chicken myb proto-oncogene promoter region: detection of an evolutionarily conserved element.

The nucleotide sequence of the chicken myb proto-oncogene putative promoter region was determined and compared with the corresponding sequence of the mouse c-myb gene (1). 118 bp upstream from the initiation codon suggested by Gerondakis and Bishop (2) for the chicken c-myb protein, a 124-bp-long conserved element was found (92% identity in chicken and mouse sequences). Sequences homologous to this element were detected on Southern blots of restricted genomic DNAs from mouse, man, lizard, frog, and carp. No hybridization was observed with Drosophila, yeast, or Escherichia coli DNA. In human DNA, sequences homologous to this element were located at the 5' end of the c-myb gene, i.e. in the same position as in the chicken and mouse genes. Several lines of evidence suggest that the element is not a coding exon of a gene overlapping the c-myb gene. It may be of importance that one of the DNase I-sensitive sites and several c-myb mRNA cap sites localized recently in the mouse c-myb gene (3,4) lie within this region. It is suggested that this evolutionarily conserved element is involved in the regulation of myb proto-oncogene expression in vertebrates.     

Nature and distribution of feline sarcoma virus nucleotide sequences.

The genomes of three independent isolates of feline sarcoma virus (FeSV) were compared by molecular hybridization techniques. Using complementary DNAs prepared from two strains, SM- and ST-FeSV, common complementary DNA'S were selected by sequential hybridization to FeSV and feline leukemia virus RNAs. These DNAs were shown to be highly related among the three independent sarcoma virus isolates. FeSV-specific complementary DNAs were prepared by selection for hybridization by the homologous FeSV RNA and against hybridization by fline leukemia virus RNA. Sarcoma virus-specific sequences of SM-FeSV were shown to differ from those of either ST- or GA-FeSV strains, whereas ST-FeSV-specific DNA shared extensive sequence homology with GA-FeSV. By molecular hybridization, each set of FeSV-specific sequences was demonstrated to be present in normal cat cellular DNA in approximately one copy per haploid genome and was conserved throughout Felidae. In contrast, FeSV-common sequences were present in multiple DNA copies and were found only in Mediterranean cats. The present results are consistent with the concept that each FeSV strain has arisen by a mechanism involving recombination between feline leukemia virus and cat cellular DNA sequences, the latter represented within the cat genome in a manner analogous to that of a cellular gene.     

The human c-fps/fes gene product expressed ectopically in rat fibroblasts is nontransforming and has restrained protein-tyrosine kinase activity.

A 13-kilobase EcoRI genomic restriction fragment containing the human c-fps/fes proto-oncogene locus was expressed transiently in Cos-1 monkey cells and stably in Rat-2 fibroblasts. In both cases, human c-fps/fes directed synthesis of a 92-kilodalton protein-tyrosine kinase (p92c-fes) indistinguishable from a tyrosine kinase previously identified with anti-fps antiserum which is specifically expressed in human myeloid cells. Transfected Rat-2 cells containing approximately 50-fold more human p92c-fes than is found in human leukemic cells remained morphologically normal and failed to grow in soft agar. Synthesis of p92c-fes in this phenotypically normal line exceeded that of the P130gag-fps oncoprotein in a v-fps-transformed Rat-2 line. Despite this elevated expression, human p92c-fes induced no substantial increase in cellular phosphotyrosine and was not itself phosphorylated on tyrosine. In contrast, p92c-fes immunoprecipitated from these Rat-2 cells or expressed as an enzymatically active fragment in Escherichia coli from a c-fps/fes cDNA catalyzed tyrosine phosphorylation with an activity similar to that of v-fps/fes polypeptides. Thus, p92c-fes is not transforming when ectopically overexpressed in Rat-2 fibroblasts. This lack of transforming activity correlates with a restriction imposed on the kinase activity of the normal c-fps/fes product in vivo which is apparently lifted for v-fps/fes oncoproteins, suggesting that regulatory interactions within the host cell modify fps/fes protein function and normally restrain its oncogenic potential.     

V(D)J recombination coding junction formation without DNA homology: processing of coding termini.

Coding junction formation in V(D)J recombination generates diversity in the antigen recognition structures of immunoglobulin and T-cell receptor molecules by combining processes of deletion of terminal coding sequences and addition of nucleotides prior to joining. We have examined the role of coding end DNA composition in junction formation with plasmid substrates containing defined homopolymers flanking the recombination signal sequence elements. We found that coding junctions formed efficiently with or without terminal DNA homology. The extent of junctional deletion was conserved independent of coding ends with increased, partial, or no DNA homology. Interestingly, G/C homopolymer coding ends showed reduced deletion regardless of DNA homology. Therefore, DNA homology cannot be the primary determinant that stabilizes coding end structures for processing and joining.     

Structure, origin, and transforming activity of feline leukemia virus-myc recombinant provirus FTT.

A myc-containing recombinant feline leukemia provirus, designated FTT, was molecularly cloned from the cat T-cell lymphoma line F422. Its transforming activity, as well as the nucleotide sequence of the 3' 2.7 kilobases of FTT, including v-myc, was determined. The predicted v-myc protein differs from feline c-myc by three amino acid changes and is truncated by two amino acids at the carboxyl terminus. Comparison with feline leukemia virus (FeLV), feline c-myc, and other FeLV proviruses indicates that recombination junctions involved in the generation of FeLV-onc viruses occur at preferred locations within the virus. They usually follow or occur within the sequence ACCCC at 5' junctions and may result from homologous recombination between sequences of marked purine-pyrimidine strand bias, especially at 3' junctions. Some recombination sites also resemble recombinase recognition sequences utilized in immunoglobulin and T-cell receptor variable-region joining. Transfection of primary rat embryo fibroblasts and subsequent in vivo analysis revealed that morphologic and tumorigenic transformation require cotransfection of FTT with human EJ-ras DNA; neither gene alone is sufficient. FTT v-myc is expressed in these transformed rat cells as a 3.0-kilobase subgenomic RNA; however, in contrast to the depressed level of c-myc expression in v-myc-involved feline tumors, steady-state levels of rat c-myc RNA and protein are apparently unaltered.     

Study of chromosome rearrangements associated with the trpE26 mutation of Bacillus subtilis

Chromosome rearrangements involved in the formation of merodiploid strains in the Bacillus subtilis 168-166 system were explained by postulating the existence of intrachromosomal homology regions. This working hypothesis was tested by analysing sequences and restriction patterns of the, as yet uncharacterized, junctions between chromosome segments undergoing rearrangements in parent, 168 trpC2 and 166 trpE26, as well as in derived merodiploid strains. Identification, at the Ia/Ib chromosome junction of both parent strains, of a 1.3 kb segment nearly identical to a segment of prophage SPbeta established the existence of one of the postulated homology sequences. Inspection of relevant junctions revealed that a set of different homology regions, derived from prophage SPbeta, plays a key role in the formation of so-called trpE30, trpE30+, as well as of new class I merodiploids. Analysis of junctions involved in the transfer of the trpE26 mutation, i.e. simultaneous translocation of chromosome segment C and rotation of the terminal relative to the origin moiety of the chromosome, did not confirm the presence of any sequence suitable for homologous recombination. We propose a model involving simultaneous introduction of four donor DNA molecules, each comprising a different relevant junction, and their pairing with the junction regions of the recipient chromosome. The resolution of this structure, resting on homologous recombination, would confer the donor chromosome structure to the recipient, achieving some kind of 'transstamping'. In addition, a rather regular pattern of inverse and direct short sequence repeats in regions flanking the breaking points could be correlated with the initial, X-ray-induced, rearrangement.