Search

Background Foulds defined, “Tumor progression (as a) permanent, irreversible qualitative change in one or more of its characters” (Cancer Res. 1954). Accordingly progressions, such as metastases and acquired drug-resistance, were since found to be subspecies of cancers with conserved and numerous new chromosomes. Here we ask whether cancers acquire numerous new chromosomes gradually or simultaneously in progressions. The currently prevailing theory of Nowell (Science, 1976) holds that unexplained “genetic instability” generates “variant sublines (with) changes in chromosome number” and that “clonal” progressions arise by “stepwise selection of more aggressive sublines”. The literature, however, contains many examples of “immediate” selections of progressions with numerous new chromosomes - notably experimentally initiated fusions between cancers and heterologous cells. Furthermore, the stepwise progression theory predicts intermediate sublines of cancers with multiple non-clonal additions of new chromosomes. However, the literature does not describe such intermediates. Results In view of these inconsistencies with stepwise progression we test here a saltational theory, in which the inherent variability of cancer-specific aneuploidy generates “immediate” progressions with individual clonal karyotypes, transcriptomes and phenotypes in single steps. Using cell fusion as an established controllable model of “immediate” progression, we generated seven immortal murine hybridomas by fusing immortal murine myeloma cells and normal antibody-producing B-cells with polyethylene glycol within a few minutes. These immortal hybridomas contained individual sets of 71 to 105 clonal chromosomes, compared to the 52 chromosomes of the parental myeloma. Thus the myeloma had gained 19 to 53 new clonal chromosomes in seven individual hybridomas in a single step. Furthermore, no stable intermediates were found, as would be predicted by a saltational process. Conclusions We conclude that random fusions between myelomas and normal B-cells generate clonal hybridomas with multiple, individual chromosomes in single steps. Similar single-step mechanisms may also generate the “late” clonal progressions of cancers with gains of numerous new chromosomes and thus explain the absence of intermediates. Latency would reflect the low probability of rare stochastic progressions. In conclusion, the karyotypic clonality of hybridomas and spontaneous progressions suggests karyotypic alterations as proximate causes of neoplastic progressions. Since cancer-specific aneuploidy catalyzes karyotypic variation, the degree of aneuploidy predicts the clinical risk of neoplastic progression, confirming classical predictions based on DNA content.

It has been known for over 100 years that cancers have individual karyotypes and arise only years to decades after initiating carcinogens. However, there is still no coherent theory to explain these definitive characteristics of cancer. The prevailing mutation theory holds that cancers are late because the primary cell must accumulate 3–8 causative mutations to become carcinogenic and that mutations, which induce chromosomal instability (CIN), generate the individual karyotypes of cancers. However, since there is still no proven set of mutations that transforms a normal to a cancer cell, we have recently advanced the theory that carcinogenesis is a form of speciation. This theory predicts carcinogens initiate cancer by inducing aneuploidy, which automatically unbalances thousands of genes and thus catalyzes chain-reactions of progressive aneuploidizations. Over time, these aneuploidizations have two endpoints, either non-viable karyotypes or very rarely karyotypes of new autonomous and immortal cancers. Cancer karyotypes are immortalized despite destabilizing congenital aneuploidy by clonal selections for autonomy—similar to those of conventional species. This theory predicts that the very low probability of converting the karyotype of a normal cell to that of a new autonomous cancer species by random aneuploidizations is the reason for the karyotypic individuality of new cancers and for the long latencies from carcinogens to cancers. In testing this theory, we observed: (1) Addition of mutagenic and non-mutagenic carcinogens to normal human and rat cells generated progressive aneuploidizations months before neoplastic transformation. (2) Sub-cloning of a neoplastic rat clone revealed heritable individual karyotypes, rather than the non-heritable karyotypes predicted by the CIN theory. (3) Analyses of neoplastic and preneoplastic karyotypes unexpectedly identified karyotypes with sets of 3–12 new marker chromosomes without detectable intermediates, consistent with single-step origins. We conclude that the speciation theory explains logically the long latencies from carcinogen exposure and the individuality of cancers. In addition, the theory supports the single-step origins of cancers, because karyotypic autonomy is all-or-nothing. Accordingly, we propose that preneoplastic aneuploidy and clonal neoplastic karyotypes provide more reliable therapeutic indications than current analyses of thousands of mutations.

Background The genetic basis of metastasis is still unclear because metastases carry individual karyotypes and phenotypes, rather than consistent mutations, and are rare compared to conventional mutation. There is however correlative evidence that metastasis depends on cancer-specific aneuploidy, and that metastases are karyotypically related to parental cancers. Accordingly we propose that metastasis is a speciation event. This theory holds that cancer-specific aneuploidy varies the clonal karyotypes of cancers automatically by unbalancing thousands of genes, and that rare variants form new autonomous subspecies with metastatic or other non-parental phenotypes like drug-resistance – similar to conventional subspeciation. Results To test this theory, we analyzed the karyotypic and morphological relationships between seven cancers and corresponding metastases. We found (1) that the cellular phenotypes of metastases were closely related to those of parental cancers, (2) that metastases shared 29 to 96% of their clonal karyotypic elements or aneusomies with the clonal karyotypes of parental cancers and (3) that, unexpectedly, the karyotypic complexity of metastases was very similar to that of the parental cancer. This suggests that metastases derive cancer-specific autonomy by conserving the overall complexity of the parental karyotype. We deduced from these results that cancers cause metastases by karyotypic variations and selection for rare metastatic subspecies. Further we asked whether metastases with multiple metastasis-specific aneusomies are assembled in one or multiple, sequential steps. Since (1) no stable karyotypic intermediates of metastases were observed in cancers here and previously by others, and (2) the karyotypic complexities of cancers are conserved in metastases, we concluded that metastases are generated from cancers in one step – like subspecies in conventional speciation. Conclusions We conclude that the risk of cancers to metastasize is proportional to the degree of cancer-specific aneuploidy, because aneuploidy catalyzes the generation of subspecies, including metastases, at aneuploidy-dependent rates. Since speciation by random chromosomal rearrangements and selection is unpredictable, the theory that metastases are karyotypic subspecies of cancers also explains Foulds’ rules, which hold that the origins of metastases are “abrupt” and that their phenotypes are “unpredictable.”

Conventional genetic theories have failed to explain why cancer (1) is not heritable and thus extremely rare in newborns, (2) is caused by non-mutagenic carcinogens, (3) develops only years to decades after initiation by carcinogens, (4) follows pre-neoplastic aneuploidy, (5) is aneuploid, (6) is chromosomally and phenotypically “unstable”, (7) carries specific aneusomies, (8) generates much more complex phenotypes than conventional mutation such as multidrug resistance, (9) generates nonselective phenotypes such as metastasis (no benefit at native site) and “immortality” (not necessary for tumorigenesis), and (10) does not contain carcinogenic mutations. We propose, instead, that cancer is a chromosomal disease. Accordingly carcinogenesis is initiated by random aneuploidies, which are induced by carcinogens or spontaneously. Since aneuploidy unbalances 1000s of genes, it corrupts teams of proteins that segregate, synthesize and repair chromosomes. Aneuploidy is therefore a steady source of chromosomal variations from which, in classical Darwinian terms, selection encourages the evolution and malignant progression of cancer cells. The rates of specific chromosomal variations can exceed conventional mutations by 4–11 orders of magnitude, depending on the degrees of aneuploidy. Based on their chromosomal constitution cancer cells are new cell “species” with specific aneusomies, but unstable karyotypes. The cancer-specific aneusomies generate complex, malignant phenotypes through the abnormal dosages of 1000s of genes, just as trisomy 21 generates Down syndrome. In sum, cancer is caused by chromosomal disorganization, which increases karyotypic entropy. Thus, cancer is a chromosomal rather than a genetic disease. The chromosomal theory explains (1) non-heritable cancer because aneuploidy is not heritable, (2) non-mutagenic carcinogens as aneuploidogens, (3) long neoplastic latencies by the low probability of evolving new species, (4) nonselective phenotypes via genes hitchhiking with selective chromosomes, and (5) immortality because, through their cellular heterogeneity, cancers survive negative mutations and cytotoxic drugs via resistant subspecies.

The transforming (onc) genes of oncogenic retroviruses share most or all of their coding sequences with normal cellular genes termed proto-onc genes. The viral genes differ from proto-onc genes in virus-specific promoters and in various point mutations and substitutions of cell-derived coding regions. In view of the structural similarities between viral oncogenes and cellular proto-onc genes, the hypothesis has been advanced that proto-onc genes become cellular cancer genes if they have suffered mutations. Indeed, point mutations and substitutions have been observed in the proto-onc genes of some cancers. However, the hypothesis has been difficult to prove because mutated proto-onc genes from tumors do not transform diploid cells. Moreover, owing to the popularity of this hypothesis, even viral oncogenes are thought to derive transforming function from mutations of this cell-derived coding region. A competing hypothesis proposes that enhanced expression from retroviral promoters is necessary and sufficient for oncogenic function of proto-onc genes. To distinguish between these hypotheses we have tested tumorigenicity of RpSV, a synthetic retrovirus with the normal proto-src coding region in a vector derived from Rous sarcoma virus (RSV). In addition, we have tested the role of RSV-specific src point mutations on the tumorigenicity of RpSV. It was found that RpSV with an unmutated proto-src coding region is tumorigenic in chickens and that tumorigenicity is enhanced by RSV-specific src point mutations. It is concluded that retroviral promoters are essential for the transforming function of viral oncogenes and that certain point mutations merely supplement their transforming function. Thus retroviral onc genes are not models for the hypothesis that mutated, but transcriptionally normal, proto-onc genes of certain tumors are cancer genes.

The 5.2-kilobase (kb) RNA genome of avian carcinoma virus MH2 has the genetic structure 5'-delta gag (0.2 kb)- mht (1.2 kb)-myc (1.4 kb)-c (0.4 kb)-poly(A) (0.2 kb)-3'. delta gag is a partial retroviral core protein gene, mht and myc are cell-derived MH2-specific sequences, and c is the 3'-terminal retroviral vector sequence. Here we have determined the nucleotide sequence of 3.5 kb from the 3' end of delta gag to the 3' end of molecularly cloned proviral MH2 DNA, in order to elucidate the genetic structure of the virus and to compare it with other mht - and myc-containing oncogenic viruses as well as with the chicken proto-myc gene. The following results were obtained: (i) delta gag- mht forms a hybrid gene with a contiguous reading frame of 2682 nucleotides that terminates with a stop codon near the 3' end of mht . The 3' 969 nucleotides of mht up to the stop codon are 80% sequence related to the onc-specific raf sequence of murine sarcoma virus 3611 (94% homologous at the deduced amino acid level). (ii) The myc sequence is preceded by an RNA splice acceptor site shared with the cellular proto-myc gene, beyond which it is colinear up to a 3'-termination codon and 40 noncoding nucleotides with the myc sequences of avian retrovirus MC29 and chicken proto-myc. Thus, myc forms, together with a 5' retroviral exon, a second MH2-specific gene. (iii) myc is followed by the 3'-terminal c region of about 400 nucleotides, which is colinear with that of Rous sarcoma virus except for a substitution near the 5' end of the long terminal repeat. It is concluded that MH2 contains two genes with oncogenic potential, the delta gag- mht gene, which is closely related to the delta gag-raf transforming gene of MSV 3611, and the myc gene, which is related to the transforming gene of MC29. Furthermore, it may be concluded that the cellular proto-onc genes, which on sequence transduction become viral onc genes, are a small group because among the 19 known onc sequences, 5 are shared by different taxonomic groups of viruses of which the mht /raf homology is the closest determined so far.

A biologically active subgenomic DNA fragment of a polycythemia-inducing strain of the replication-defective spleen focus-forming virus (SFFV) has been molecularly cloned. The SFFV DNA fragment includes 2.0 kilobase pairs (kbp) from the 3' end of SFFV, the long terminal repeat sequences of SFFV, and 0.4 kbp from the 5' end of SFFV. The fragment contains the previously described env-related gene of SFFV. All the properties associated with SFFV can be assigned to this SFFV DNA fragment by using a two-stage DNA transfection assay with infectious helper virus DNA. The virus recovered from the transfection assays can induce erythroblastosis, splenic foci, and polycythemia in infected mice. Fibroblast cultures transfected with the SFFV DNA fragment synthesize gp52, the known intracellular product of the env-related gene of SFFV. gp52 can also be detected in spleens from diseased mice infected with the virus recovered in the two-stage transfection. The results are consistent with the hypothesis that the env-related gene sequences of SFFV and their product gp52 are required for the initiation of SFFV-induced disease.

Human DNA segments homologous to the ets region from the transforming gene of avian erythroblastosis virus, E26, were molecularly cloned and shown to be closely related to the viral equivalent by hybridization and partial sequence analysis. The transforming gene of E26 has a tripartite origin with the structure delta gag [1.2 kilobases (kb) from the viral gag gene]-myb(0.9 kb from the chicken myb gene)-ets (1.6 kb from the chicken ets gene). Human ets DNA is located on two distinct human chromosomes. The human ets-1 locus on chromosome 11 encodes a single mRNA of 6.8 kb; the second locus, ets-2 on chromosome 21, encodes three mRNAs of 4.7, 3.2, and 2.7 kb. The ets-related sequences of human DNA on chromosomes 11 and 21 are discontiguous, except for a small overlap region encoding 14 amino acids, where 12 are conserved between these two loci. By contrast, the chicken homolog has contiguous ets-1 and ets-2 sequences and is primarily expressed in normal chicken cells as a single 7.5-kb mRNA. We conclude that the ets sequence shared by the virus, the chicken, and humans is likely to contain at least two dissociable functional domains, ets-1 and ets-2. Thus, the tripartite transforming gene of E26 includes four distinct domains that may be functionally relevant for the transforming function of the virus (delta gag, myb, ets-1, and ets-2).

The oncogenic properties and RNA of the Fujinami avian sarcoma virus (FSV) and the protein it encodes were investigated and compared to those of other avian tumor viruses with sarcomagenic properties such as Rous sarcoma virus and the acute leukemia viruses MC29 and erythroblastosis virus. Cloned stocks of FSV caused sarcomas in all chickens inoculated and were found to contain a 4.5-kilobase (kb) and an 8.5-kb RNA species. The 4.5-kb RNA was identified as the genome of defective FSV because it was absent from nondefective FSV-associated helper virus and because the titer of focus-forming units increased with the ratio of 4.5-kb to 8.5-kb RNA in virus preparations. This is, then, the smallest known tumor virus RNA with a transforming function. Comparisons with other viral RNAs, based on oligonucleotide mapping and molecular hybridization, indicated that 4.5-kb FSV RNA contains a 5′ gag gene-related sequence of 1 kb, an internal specific sequence of about 3 kb that is unrelated to Rous sarcoma virus, MC29, and erythroblastosis virus, and a 3′-terminal sequence of about 0.5 kb related to the conserved C region of avian tumor viruses. The lack of some or all nucleotide sequences of the essential virion genes, gag, pol , and env , and the isolation of FSV-transformed nonproducer cell clones indicated that FSV is replication defective. A 140,000-dalton, gag -related non-structural protein was found in FSV-transformed producer and nonproducer cells and was translated in vitro from full-length FSV RNA. This protein is expected to have a transforming function both because its intracellular concentration showed a positive correlation with the percentage of transformed cells in a culture and because FSV is unlikely to code for major additional proteins since the genetic complexities of FSV RNA and the FSV protein are almost the same. It is concluded that the transforming onc gene of FSV is distinct from that of Rous sarcoma virus and other avian tumor viruses with sarcomagenic properties. Hence, multiple mechanisms exist for sarcomagenic transformation of avian cells.

The ras genes of BALB and Harvey sarcoma viruses contain point mutations in codon 12 or codons 12 and 59, relative to proto-ras from normal animal and human cells. By in vitro recombination between cloned rat proto-ras and cloned BALB and Harvey sarcoma proviruses, we constructed recombinant proviruses with normal proto-ras-coding regions. These recombinant proviruses transformed mouse 3T3 cells upon transfection. However, when the transforming efficiencies of proviral DNAs were compared after transfection with helper provirus, recombinant proviruses were 2 to 30 times less efficient than the corresponding wild-type proviruses. Recombinant sarcoma viruses isolated from cells transformed by cloned proviral DNA contained the expected normal ras-coding region. They transformed rat embryo cells and induced erythroblastosis and sarcomas in newborn mice as efficiently as wild-type viruses did. We conclude that conversion of normal proto-ras genes to viral ras genes depends on truncation of normal proto-ras regulatory elements and substitution by retroviral (long terminal repeat) promoters and that the transforming function of long terminal repeat-ras genes is enhanced by point mutations.

The structure and molecular weight of the 60-70S RNA complex and the 30-40S RNA species of Rous sarcoma virus were analyzed in an electron microscope after treatment of the RNAs with the bacteriophage T4 gene-32 protein to stretch out the RNA strands. Although all RNA preparations treated with gene-32 protein showed considerable heterogeneity in length, a significant fraction of the RNA retained its original sedimentation coefficient after treatment to allow the following conclusions to be made: The 30-40S RNA was confirmed to be a linear polynucleotide with a molecular weight of about 3 × 10 6 . The 60-70S RNA exhibited a network structure with a molecular weight predominantly of about 6 × 10 6 . Therefore, the subunit hypothesis for the 60-70S RNA is confirmed. A model for the structure and molecular weight of the 60-70S RNA postulates that the complex consists of two 30-40S RNA subunits held together at many points. This model elucidates the biological observation that the infectivity of RNA tumor viruses is proportional to the amount of 30-40S RNA in a virus preparation and not to the amount of 60-70S RNA.

Envelope-specific and sarcoma-specific nucleotide sequences have been located within the 10,000 nucleotides of the RNA of nondefective Schmidt-Ruppin Rous sarcoma virus (nd SR). For this purpose, about 30 RNase-T1-resistant oligonucleotides were ordered relative to the 3'-poly(A) terminus of the RNA, to construct an oligonucleotide map of the nd SR RNA. A cluster of seven envelope-specific oligonucleotides, identified by their absence from an otherwise very similar oligonucleotide map of an envelop-defective deletion mutant (which lacks the major viral glycoprotein), mapped at a distance of 2800-5000 nucleotides from the poly(A) end of nd SR RNA. A cluster of two sarcoma-specific oligonucleotides, identified by their absence from an otherwise nearly identical oligonucleotide map of a transformation-defective deletion mutant, mapped at a distance of 1000-2000 nucleotides from the poly(A) end of nd SR RNA. The oligonucleotide maps of nd SR and of the two deletion mutants were the same from the poly(A) end up to 650 nucleotides and included one terminal oligonucleotid, termed C, which is found in all avian tumor viruses tested so far. A possible gene order consistent with our data suggests that sarcoma-specific nucleotide sequences map between envelope-specific nucleotide sequences and the poly(A) end of the RNA.

Only two avian oncogenic viruses specifically cause acute leukaemias yet do not transform chicken fibroblasts in culture: E26, which causes erythroblastosis and a low level of concomitant myeloblastosis in chickens, and avian myeloblastosis virus (AMV), which causes myeloblastosis exclusively. Both viruses are replication-defective and share a sequence termed myb (also known as amv) which is unrelated to essential virion genes and is therefore thought to be part of the transforming onc genes of these viruses. However, the genetic structure of the two viruses differs. E26 has a genomic RNA of 5.7 kilobases (kb) and encodes a 135,000 molecular weight gag-related protein (p135) with probable transforming function. We show here by in vitro translation that the 5.7-kb E26 RNA directs the synthesis of p135. Oligonucleotide analysis indicates that E26 RNA contains an internal 0.8-kb subset of the 1.2-kb AMV-related sequence (mybA), termed mybE. A 2.46-kb molecular clone prepared from cDNA transcribed in vitro from E26 RNA contained an E26 transformation-specific (ets) sequence flanked by mybE and an env-related sequence. A complete DNA sequence of this clone indicates that the 1.5-kb ets sequence extends the open reading frame of mybE for 491 amino acids. Thus, the p135 gene of E26 is a genetic hybrid of three distinct elements, approximately 1.2 kb derived from the 5' region of the retroviral gag gene, mybE and the ets sequence, linked in the order 5'-delta gag-mybE-ets-3'. The myeloid leukaemogenicity shared by E26 and AMV correlates with the common myb sequence, while the distinct erythroid leukaemogenicity of E26 correlates with ets and the E26-specific linkage of myb to delta gag.

The src genes of different Rous sarcoma virus (RSV) strains have been reported to be highly conserved by some investigators using RNA-cDNA hybridization, whereas others using oligonucleotide, peptide, and serological analyses have judged src genes to be variable in 30 to 50% of the respective markers. Moreover, distinctive src oligonucleotides and peptides of so-called recovered RSVs (rRSV's) whose src genes were reported to be experimentally transduced from the cell are thought to represent specific markers of host-derived src sequences. By contrast, we have pointed out previously that these markers may represent point mutations of parental equivalents. Here we have compared the src -specific sequences of eight RSV strains and of two rRSV's to each other and to a molecular clone of the src -related chicken locus. Our comparisons are based on RNase T 1 -resistant oligonucleotides of RNA hybridized to src -specific cDNA, which was prepared by hybridizing RSV cDNA with RNA of isogenic src deletion mutants, or to a cloned cellular src -related DNA. All of the approximately 20 src -oligonucleotides of a given RSV strain were recovered by src -specific cDNA's of all other RSV strains or by cellular src -related DNA. The number of oligonucleotides varied slightly with the length of the src deletion used to prepare src -specific cDNA, thus providing a measure for src deletion mutants. Our data indicate that the src genes of all RSV strains tested, including the two reportedly transduced from the cell, are about 98% conserved and completely allelic with only scattered single nucleotide differences in certain variable regions which are subject to point mutations. Hence, based on the src oligonucleotide markers analyzed by us and others, we cannot distinguish between a cellular and viral origin of rRSV's. However, the following are not compatible with a cellular origin of rRSV's. (i) The only putative oligonucleotide marker which is exclusively shared by the two rRSV's studied and which differs from a parental counterpart in a single base was not detectable in cellular src -related DNA. (ii) The number of different allelic src markers observed by us and others in rRSV's was too large to derive from one or two known cellular src -related loci. (iii) The known absence of linkage of the cellular src -related locus with other virion sequences was extended to all non- src oligonucleotides, including some mapping directly adjacent to src . This is difficult to reconcile with the claim that transformation-defective, partial src deletion mutants of RSV which contain both, one, or, as we show here, possibly no src termini nevertheless transduce at the same frequencies, even though homologous, single or double illegitimate recombinations would be involved. Given (i) our evidence that src genes are subject to point mutation under selective conditions similar to those prevailing when rRSV's were generated and (ii) the lack of absolute evidence for the clonal purity of the transformation-defective, partial src deletion mutants of RSV used to generate rRSV's, we submit that the src genes of rRSV's could have been generated by cross-reactivation of nonoverlapping src deletions or mutation of src variants possibly present in transformation-defective, partial src deletion mutants of RSV. To prove experimental transduction, unambiguous markers need to be identified, or it would be necessary to generate rRSV's with molecularly cloned transformation-defective, partial src deletion mutants of RSV. Although our evidence casts doubt on the idea that specific src sequences of rRSV's originated by transduction, the close relationship between viral src and cellular src -related sequences argues that src genes originated at one time in evolution from the cell by events that involved illegitimate recombination and deletion of non- src sequences that interrupt the cellular src locus.

Two major RNA species were found in several clonal isolates of avian erythroblastosis virus (AEV) and avian erythroblastosis-associated helper virus (AEAV) complexes: one of 8.7 kilobases (kb), the other of 5.5 kb. The 5.5-kb species was identified as AEV RNA because ( i ) it was absent from non-transforming AEAV isolated from the same virus complex, ( ii ) it was present in complexes of AEV and different helper viruses, and ( iii ) its structure is similar to that of avian acute leukemia viruses of the MC29 group. Molecular hybridization indicated that 54% of AEV RNA is specific and 46% is related to other viruses of the avian tumor virus group, particularly to AEAV, therefore termed group-specific. The genetic structure of AEV RNA was deduced by mapping oligonucleotides representing specific and group-specific sequences and by comparing the resulting map to maps of AEAV and of other avian tumor viruses derived previously. AEV RNA contains a gag gene-related, 5′ group-specific section of 1 kb, an internal AEV-specific section of 3 kb unrelated to any other viral RNA tested, and a 3′ group-specific section of 1.5 kb. The 5′ section of AEV RNA is closely related to analogous 5′ sections of the MC29 group viruses and is homologous with a 5′ RNA section that is part of the gag gene of AEAV. The 3′ section is also shared with AEAV RNA and includes a variant C-oligonucleotide near the 3′ end that is different from the highly conserved counterparts of all other exogenous avian tumor viruses. By analogy with Rous sarcoma virus and the acute leukemia viruses of the MC29 group, the internal specific section of AEV RNA is thought to signal a third class of onc genes in avian tumor viruses. Comparisons with AEAV and the MC29 group viruses suggest that both the 5′ gag -related and the internal specific RNA sections of AEV are necessary for onc gene function.

To define conditions under which the chicken protooncogene p-myc is converted to a viral and possibly to a cellular transforming gene, we assayed transforming function of hybrid genes put together from cloned retroviral and p-myc elements and of p-myc genes isolated from spontaneous viral lymphomas. Transforming function was measured in quail embryo cells transfected with cloned myc genes. We found that only myc genes with a promoter of a retroviral long terminal repeat (LTR) located between the native p-myc promoter and the second p-myc exon have transforming function. Transforming efficiencies decreased with increasing lengths of unspliced sequences between the LTR and p-myc exon 2. p-myc DNAs with LTRs downstream of the coding region or upstream but in the opposite transcriptional orientation failed to transform embryo cells. Likewise, only those retroviral-p-myc combinations from chicken B-cell lymphomas with a LTR positioned as promoter upstream of p-myc exon 2 had transforming function. We conclude that substitution of a retroviral LTR for the promoter and for as yet poorly defined, untranscribed regulatory elements of p-myc is sufficient to convert chicken p-myc to a transforming gene. However, retroviral LTRs can only convert p-myc genes to embryo-cell-transforming genes from a limited number of positions, and not as position-independent enhancers. Further, we deduce that there are two classes of viral chicken B-cell lymphomas, those with and those without embryo-cell-transforming p-myc genes.

RNA and protein of the defective avian acute leukemia virus CMII, which causes myelocytomas in chickens, and of CMII-associated helper virus (CMIIAV) were investigated. The RNA of CMII measured 6 kilobases (kb) and that of CMIIAV measured 8.5 kb. By comparing more than 20 mapped oligonucleotides of CMII RNA with mapped and nonmapped oligonucleotides of acute leukemia viruses MC29 and MH2 and with mapped oligonucleotides of CMIIAV and other nondefective avian tumor viruses, three segments were distinguished in the oligonucleotide map of CMII RNA: (i) a 5' group-specific segment of 1.5 kb which was conserved among CMII, MC29, and MH2 and also homologous with gag-related oligonucleotides of CMIIAV and other helper viruses (hence, group specific); (ii) an internal segment of 2 kb which was conserved specifically among CMII, MC29, and MH2 and whose presence in CMII lends new support to the view that this class of genetic elements is essential for oncogenicity, because it was absent from an otherwise isogenic, nontransforming helper, CMIIAV; and (iii) a 3' group-specific segment of 2.5 kb which shared 13 of 14 oligonucleotides with CMIIAV and included env oligonucleotides of other nondefective viruses of the avian tumor virus group (hence, group specific). This segment and analogous map segments of MC29 and MH2 were not conserved at the level of shared oligonucleotides. CMII-transformed cells contained a nonstructural, gag gene-related protein of 90,000 daltons, distinguished by its size from 110,000-daltom MC29 and 100,000-dalton MH2 counterparts. The gag relatedness and similarity to the 110,000-dalton MC29 counterpart indicated that the 90,000-dalton CMII protein is translated from the 5' and internal segments of CMII RNA. The existence of conserved 5' and internal RNA segments and conserved nonstructural protein products in CMII, MC29, and MH2 indicates that these viruses belong to a related group, termed here the MC29 group. Viruses of the MC29 group differ from one another mainly in their 3' RNA segments and in minor variations of their conserved RNA segments as well as by strain-specific size markers of their gag-related proteins. Because (i) the conserved 5' gag-related and internal RNA segments and their gag-related, nonvirion protein products correlate with the conserved oncogenic spectra of the MC29 group of viruses and because (ii) the internal RNA sequences and nonvirion proteins are not found in nondefective viruses, we propose that the conserved RNA and protein elements are necessary for oncogenicity and probably are the onc gene products of the MC29 group of viruses.

Avian carcinoma virus MH2 contains two potential transforming genes, delta gag-mht and delta gag-myc. Thus, MH2 may be a model for two-gene carcinogenesis in which transformation depends on two synergistic genes. Most other directly oncogenic viruses contain single, autonomous transforming (onc) genes and are models for single-gene carcinogenesis. To determine which role each potential onc gene of MH2 plays in oncogenesis, we have prepared deletion and frameshift mutants of each of the two MH2 genes by in vitro mutagenesis of cloned proviral DNA and have tested transforming function and virus production in cultured primary quail cells. We have found that mht deletion mutants and wild-type virus transform primary cells and that myc deletion and frameshift mutants do not. The morphologies of cells transformed by the mht deletion mutants and by wild-type MH2 are similar yet vary considerably. Nevertheless, typical mutant transformed cells can often be distinguished from cells transformed by wild-type MH2. We conclude that the delta gag-myc gene transforms primary cells by itself, without the second potential onc gene. This myc-related gene is the smallest that has direct transforming function. delta gag-mht is without detectable transforming function but may affect transformation by delta gag-myc. Thus, MH2 behaves like a virus with a single onc gene, although it expresses two potential onc genes, and it appears not to be a model for two-gene carcinogenesis. Further work is necessary to determine whether the delta gag-mht gene possibly enhances oncogenic function of delta gag-myc or has independent oncogenic function in animals.

Replication-defective acute leukemia viruses E26 and myeloblastosis virus (AMV) cause distinct leukemias although they belong to the same subgroup of oncogenic avian tumor viruses based on shared transformation-specific ( onc ) RNA sequences. E26 causes predominantly erythroblastosis in chicken and in quail, whereas AMV induces a myeloid leukemia. However, upon cultivation in vitro for >1 month, a majority of surviving hemopoietic cells of E26-infected animals bear myeloid markers similar to those of AMV-transformed cells. We have analyzed the genetic structure and gene products of E26 virus for a comparison with those of AMV. An E26/helper virus complex was found to contain two RNA species: a 5.7-kilobase (kb) RNA that hybridizes with cloned AMV-specific proviral DNA and hence is probably the E26 genome; and an 8.5-kb RNA that is unrelated to AMV and represents helper virus RNA. Thus, E26 RNA is smaller than 7.5-kb AMV RNA. Hybridization of size-selected poly(A)-terminating E26 RNA fragments with AMV-specific DNA indicated that the shared specific sequences are located in the 5′ half of the E26 genome as opposed to a 3′ location in AMV RNA. In nonproducer cells transformed in vitro by E26, a gag -related nonstructural 135,000-dalton protein (p135) was found. No gag (Pr76) or gag-pol (Pr180) precursors of essential virion proteins, which are present in AMV nonproducer cells, were observed. p135 was also found in cultured E26 virus producing cells of several leukemic chickens, and its intracellular concentration relative to that of the essential virion proteins encoded by the helper virus correlates with the ratio of E26 to helper RNA in virions released by these cells. p135 is phosphorylated but not glycosylated; antigenically it is not related to the pol or env gene products. It appears to be coded for by a partial gag gene and by E26-specific RNA sequences, presumably including those shared with AMV. Hence, AMV and E26 appear to use different strategies for the expression of related onc sequences: AMV is thought to encode a transforming protein via a subgenomic mRNA, whereas E26 codes for a gag -related polyprotein via genomic RNA. It is speculated that differences in the oncogenic properties of E26 and AMV are due to differences in their genetic structures and gene products.

The RNase-T1-resistant oligonucleotides of two Prague Rous sarcoma viruses with temperature-sensitive (ts) DNA polymerases (DNA nucleotidyltransferases), termed ts LA 337 and 335 of one leukosis virus, RAV-6, and 20 of their recombinant progeny have been mapped relative to the 3' poly (A) terminus of the viral RNA. The resulting oligonucleotide maps have been ocrrelated with markers of the four known viral genetic elements encoded in the RNA of 10,000 nucleotides. In accord with previous results recombinant RNAs contained (i) oligonucleotides characteristic of the src gene, coding for sarcoma formation, between the poly(A) end and 2000 nucleotides and (ii) olignucleotides characteristic of the env gene, coding for the envelope glycoprotein, between 2500 and 5000 nucleo tides from the poly(A) end. (iii) A cluster of four oligonucleotides that mapped between 6000 and 8000 nucleotides from the 3' poly(A) end of each RNA was shared by both parental viruses and all recombinants. Since all other map segments of our recombinants failed to segregate with the ts- or wild-type markers of the parental DNA polymerase gene (pol), it was concluded that the ts pol lesion maps in this RNA segment. (iv) The 5' segment of each recombinant RNA contained a cluster of four to five oligonucleotides whose parental origin correlated with an electrophoretic marker of one of the parental virion proteins, p27, a major product of the viral gag gene. The gene order 5'-gag-pol-env-src-poly(A) is consistent with our data.

Several clones of Fujinami sarcoma virus (FSV) isolated from a laboratory stock or from mutagenized virus were temperature sensitive ( ts ) in transformation of cells in culture. When shifted from the permissive (37°C) to the nonpermissive (41.5°C) temperature, the cellular phenotype reverted to normal within 2 h, but it required about 48 h at 37°C to revert back to the transformed morphology. A temperature-resistant ( tr ) FSV clone was isolated from a tumor of an animal. All ts mutants were tumorigenic in animals but induced tumors only after latent periods of 12 to 25 days, compared to 5 to 6 days with tr virus. The ts lesions of the FSV mutants affected 90% of the phosphorylation of the nonstructural, gag -related 140,000-kilodalton phosphoprotein coded by FSV (p140), but did not affect virus replication or the synthesis of p140. Upon shifting from the permissive to the nonpermissive temperature, p140 was 90% dephosphorylated with an approximate 32 P half-life of 20 min. When shifted back to the permissive temperature, the preexisting p140 was rephosphorylated in the absence of protein synthesis within a 90-min test period. Likewise, most of the phosphate of fully phosphorylated p140 was exchanged at the permissive temperature within 30 to 90 min even when protein synthesis was inhibited. However, the protein structure of p140 had a half-life of 5 h at both temperatures. These results prove p140 to be a substrate of reversible phosphorylation. Superinfection and transformation of ts FSV-infected cells maintained at the nonpermissive temperature with acute leukemia virus MC29 failed to phosphorylate p140. It would follow that in vivo phosphorylation of ts p140 is controlled by an FSV-specific mechanism and is a prerequisite, not a consequence, of transformation. p140 of ts FSV recovered from cells maintained at 41.5°C with anti- gag serum was over 10 times less phosphorylated by associated kinase than the same protein recovered from cells at 37°C if assayed in vitro at 20°C. This kinase activity associated with or dissociated from p140 with a half-life of less than 30 min during temperature shifts of ts FSV-infected cells. However, p140 recovered from ts FSV-infected cells maintained at 37°C was phosphorylated by associated kinase in vitro not only at 20°C but also, and essentially at the same level, at 41.5°C. This suggests that the kinase associated with the immunocomplex of p140 of ts FSV is not temperature sensitive. p140 translated in vitro from ts and tr FSV RNA lacked kinase activity. We conclude that a fully phosphorylated p140 is necessary for the maintenance of transformation by FSV. This is consistent with the notion that other highly oncogenic viruses also code for nonstructural phosphoproteins with probable transforming function. A model which postulates that p140 is a substrate of reversible phosphorylation and that the lesion of the ts FSV clones described herein affects association of p140 with a cellular kinase rather than a hypothetical intrinsic kinase activity of the protein is most compatible with our data.

Kirsten murine sarcoma-leukemia virus (Ki-MSV[MLV]) was found to contain less RNase H per unit of viral DNA polymerase than avian Rous sarcoma virus (RSV). Upon purification by chromatography on Sephadex G-200 and subsequent glycerol gradient sedimentation the avian DNA polymerase was obtained in association with a constant amount of RNase H. By contrast, equally purified DNA polymerase of Ki-MSV(MLV) and Moloney [Mo-MSV(MLV)] lacked detectable RNase H if assayed with two homopolymer and phage fd DNA-RNA hybrids as substrates. On the basis of picomoles of nucleotides turned over, the ratio of RNase H to purified avian DNA polymerase was 1:20 and that of RNase H to purified murine DNA polymerase ranged between

Several natural RNAs were compared with respect to their template activities for the DNA polymerase of Rous Sarcoma Virus during a 2-hr incubation period. 60-70S viral RNA was found to be a 5- to 10-fold better template than heat-dissociated Rous viral RNA, influenza virus RNA, tobacco mosaic virus RNA, or ribosomal RNA. Denatured salmon DNA is a little better, and poly(dAT) is 2-4 times better as a template for the enzyme than is 60-70S Rous viral RNA. The 60-70S RNAs of different strains of avian tumor viruses have very similar template activities for a given avian tumor virus DNA polymerase. Oligo(dT) or oligo(dC) were found to enhance the template activity of heat-dissociated Rous viral RNA 20- to 30-fold, and that of other natural RNAs tested one- to several-fold. DNA syntheses of 1-24% were obtained during a 2-hour incubation of the enzyme with the above RNA templates. The results suggest that the enzyme prefers partially doublestranded or hybrid regions of RNAs for optimal DNA synthesis, but certain regions of single-stranded RNA can also serve as templates. Poly(dAT) competes with viral RNA for purified DNA polymerase during DNA synthesis, as would be expected if RNA- and DNA-dependent DNA synthesis was performed by at least one common active site of the same enzyme.

Preinfection of chick embryo fibroblasts with an avian tumor virus does not eliminate the requirement of early DNA synthesis for other avian tumor viruses superinfecting the same cells several days after the primary inoculum. The function served by the early DNA in avian tumor virus infection either fails to persist into late infection or is genetically specific for each viral type.

Background In 1952 Papanicolaou et al. first diagnosed and graded cervical carcinomas based on individual “abnormal DNA contents” and cellular phenotypes. Surprisingly current papilloma virus and mutation theories of carcinomas do not mention these individualities. The viral theory holds that randomly integrated, defective genomes of papilloma viruses, which are often untranscribed, cause cervical carcinomas with unknown cofactors 20–50 years after infection. Virus-free carcinomas are attributed to mutations of a few tumor-suppressor genes, especially the p53 gene. But the paradox of how a few mutations or latent defective viral DNAs would generate carcinomas with endless individual DNA contents, degrees of malignancies and cellular phenotypes is unsolved. Since speciation predicts individuality, we test here the theory that cancers are autonomous species with individual clonal karyotypes and phenotypes. This theory postulates that carcinogens induce aneuploidy. By unbalancing mitosis genes aneuploidy catalyzes chain reactions of karyotypic evolutions. Most such evolutions end with non-viable karyotypes but a few become new cancer karyotypes. Despite congenitally unbalanced mitosis genes cancer karyotypes are stabilized by clonal selections for cancer-specific autonomy. Results To test the prediction of the speciation theory that individual carcinomas have individual clonal karyotypes and phenotypes, we have analyzed here the phenotypes and karyotypes of nine cervical carcinomas. Seven of these contained papilloma virus sequences and two did not. We determined phenotypic individuality and clonality based on the morphology and sociology of carcinoma cells in vitro. Karyotypic individuality and clonality were determined by comparing all chromosomes of 20 karyotypes of carcinomas in three-dimensional arrays. Such arrays list chromosome numbers on the x-axis, chromosome copy numbers on the y-axis and the number of karyotypes arrayed on the z-axis. We found (1) individual clonal karyotypes and phenotypes in all nine carcinomas, but no virus-specific markers, (2) 1-to-1 variations between carcinoma-specific karyotypes and phenotypes, e.g. drug-resistance and cell morphology, (3) proportionality between the copy numbers of chromosomes and the copy numbers of hundreds of over- and under-expressed mRNAs, (4) evidence that tobacco-carcinogens induce cervical carcinomas via aneuploidy, consistent with the speciation theory. Conclusions Since the individual clonal karyotypes of nine carcinomas correlated and co-varied 1-to-1 with complex individual transcriptomes and phenotypes, we have classical genetic and functional transcriptomic evidence to conclude that these karyotypes encode carcinomas - much like the clonal karyotypes that encode conventional species. These individual karyotypes explain the individual “DNA contents”, the endless grades of malignancies and the complex individual transcriptomes and phenotypes of carcinomas.