The transforming protein of Moloney murine sarcoma virus is a soluble cytoplasmic protein

The transforming gene, v-mos, of Moloney murine sarcoma virus (M-MuSV) encodes a 37,000-dalton phosphoprotein, p37mos. Since the biochemical function of this protein is unknown, we have determined the subcellular location of p37mos in M-MuSV 124-transformed cells. Using two different methods of cell lysis and fractionation, we found that newly synthesized as well as mature p37mos is a soluble cytoplasmic protein. In agreement with these results, immunofluorescent staining of cells acutely infected with M-MuSV 124, using an antiserum directed against a synthetic v-mos peptide, produced a diffuse cytoplasmic pattern. Gel filtration experiments and glycerol gradient sedimentation analysis suggest that the bulk of p37mos exists as a monomer and is not involved in a specific association with other cellular proteins. These properties of p37mos are different from those of other characterized retroviral transforming proteins.     

Detection of a transforming gene product in cells transformed by Moloney murine sarcoma virus

We identified, in cells transformed by Moloney murine sarcoma virus (M-MuSV clone 124), a protein encoded by the M-MuSV transforming gene, v-mos. An antiserum against a synthetic peptide corresponding to the C terminus of a protein predicted from the v-mos nucleotide sequence specifically recognizes a protein doublet of approximately 37,000 daltons from 35S-methionine-labeled M-MuSV 124-transformed producer cells. By peptide mapping, this protein is almost identical to the 37 kd in vitro translation product from the M-MuSV v-mos gene. Immunoprecipitates from 32P-labeled cells contain a single v-mos-specific phosphoprotein, which has at least six sites of phosphorylation containing phosphoserine. Pulse-chase experiments show that the lower band in the 35S-methionine-labeled doublet is the primary translation product, which is modified, probably by phosphorylation, to yield the upper band. A similar mos protein is immunoprecipitated from HT1-MuSV-transformed cells, but not from uninfected NIH/3T3 cells. These mos proteins are present at very low levels in transformed cell lines. Cells acutely infected with M-MuSV 124, however, transiently contain much higher levels of the mos protein. These high levels coincide with extensive cell mortality.     

Use of site-specific antipeptide antibodies to perturb the serine kinase catalytic activity of p37mos.

The mos oncogene of Moloney murine sarcoma virus encodes a protein of approximately 37,000 daltons (designated p37mos). We have detected a serine protein kinase activity which is closely associated with p37mos in immune complexes obtained with antibodies [anti-mos(37-55) serum] that were generated with a peptide containing amino acids 37 through 55 of the v-mos protein (S. A. Maxwell and R. B. Arlinghaus, Virology 143:321-333, 1985). Immune complexes that were derived with antibodies generated against peptides representing the C-terminal 8 or 12 amino acids of v-mos (anti-C2 and anti-C3 serum, respectively) exhibited very little kinase activity capable of phosphorylating p37mos. Treatment of anti-mos(37-55) complexes containing active v-mos kinase with anti-C3 or anti-C2 serum resulted in a dramatic reduction of the in vitro phosphorylation of p37mos. Antiserum blocked with the appropriate C-terminal peptide had no inhibitory effect on the phosphorylation of p37mos in anti-mos(37-55) complexes which indicated that the inhibition of v-mos kinase activity was a specific effect of these antibodies. The specific inhibition of the in vitro phosphorylation of p37mos by antibodies directed against the C terminus of the v-mos protein provides strong evidence that the v-mos gene encodes a serine protein kinase. In addition, the extreme C terminus of p37mos may be critical for an active v-mos kinase.     

Biologically active mutants with deletions in the v-mos oncogene assayed with retroviral vectors.

We have constructed retroviral expression vectors by manipulation of the Moloney murine leukemia virus genome such that an exogenous DNA sequence may be inserted and subsequently expressed when introduced into mammalian cells. A series of N-terminal deletions of the v-mos oncogene was constructed and assayed for biological activity with these retroviral expression vectors. The results of the deletion analysis demonstrate that the region of p37mos coding region upstream of the third methionine codon is dispensable with respect to transformation. However, deletion mutants of v-mos which allow initiation of translation at the fourth methionine codon have lost the biological activity of the parental v-mos gene. Furthermore, experiments were also carried out to define the C-terminal limit of the active region of p37mos by the construction of premature termination mutants by the insertion of a termination oligonucleotide. Insertion of the oligonucleotide just 69 base pairs upstream from the wild-type termination site abolished the focus-forming ability of v-mos. Thus, we have shown the N-terminal limit of the active region of p37mos to be between the third and fourth methionines, while the C-terminal limit is within the last 23 amino acids of the protein.     

Molecularly cloned c-mos(rat) is biologically active.

A unique rat cellular gene, c-mos(rat), homologous to the transforming sequences, v-mos, of Moloney murine sarcoma virus (M-MSV) was detected by hybridization to a v-mos specific probe. The c-mos(rat) gene was cloned together with its flanking sequences in an 11-kbp EcoRI DNA fragment inserted in vector Charon 4A. Two probes were used to investigate the position and orientation of c-mos(rat) in the clone examined ( D3e ), namely pMSV -31 which contains the sequences specific for the transforming sequences of M-MSV and pCS-1 which harbors 0.5 kbp of 5'-terminal sequences of c-mos(mouse) as well as 0.7 kbp of its flanking sequences. After ligation of a restriction fragment of clone D3e containing c-mos(rat) to a fragment containing the long terminal repeat of M-MSV and transfection of the DNA onto rat cells, we detected foci of transformed cells, thus showing that c-mos(rat) is biologically active. Using DNA framents derived from clone D3e , we studied the conservation of c-mos and of its flanking sequences in several species. c-mos(rat) as well as some of its flanking sequences appeared to be highly conserved in the species studied.     

mos gene transforming efficiencies correlate with oocyte maturation and cytostatic factor activities.

The mos proto-oncogenes from different vertebrate species transform mouse NIH 3T3 cells with markedly different efficiencies. v-mos, mouse (c-mosmu), and chicken (c-mosch) mos transform NIH 3T3 cells 10- to 100-fold more efficiently than do human (c-moshu) and Xenopus (c-mosxc) mos. The mos genes with the highest transforming activity efficiently induce maturation in Xenopus oocytes and mimic cytostatic factor (CSF) by causing mitotic cleavage arrest in embryos. Chimeric v-mos/c-moshu proteins that had high transforming efficiencies in NIH 3T3 cells were also effective in the induction of oocyte maturation and CSF cleavage arrest. We measured the in vitro autophosphorylation activities of the different mos proteins and found that the levels of kinase activity of v-mos, c-mosmu, and c-mosch were much higher than that of c-mosxc. These data indicate that mos gene transforming efficiency and the ability to induce oocyte maturation or mimic CSF activity are correlated with in vitro autophosphorylation activity and suggest that the mos protein plays a similar role in transformed cells and normal oocytes.     

Dose-dependent regulation of macrophage differentiation by mos mRNA in a human monocytic cell line.

The proto-oncogene c-mos was expressed during differentiation of the human monocytic cell line U937 into macrophages. To investigate a possible role of the mos oncogene, we introduced the v-mos gene under an inducible promoter, MT-I, into U937 cells. The v-mos transformed cells expressed mos mRNA at an amount proportional to the concentration of Zn2+ ions. The induction of the v-mos gene caused growth inhibition and macrophage differentiation in these cells. The differentiation of v-mos transformed monocytes into macrophages required continuous expression of the v-mos gene. The extent of expression of phenotypic characteristics of macrophages, such as phagocytosis, cell surface antigens and typical morphology, depends on the amount of mos mRNA present. We were therefore able to demonstrate that the expression of only one oncogene, mos, determines monocyte differentiation into macrophages.     

p37mos-associated serine/threonine protein kinase activity correlates with the cellular transformation function of v-mos.

A serine/threonine-specific protein kinase activity is closely associated with v-mos-encoded proteins. Experiments were conducted with several mutant forms of p37mos to determine whether or not the kinase function correlates with the biological activity of the mutant v-mos genes. Two mutants lacking cell transformation activity, one an arginine substitution for lysine-121 in the putative ATP-binding site and the other a 23-amino acid deletion from the C-terminal end of p37mos, had no kinase activity associated with their mutant proteins. However, a third mutant with reduced biological activity had drastically less kinase activity than the wild-type protein. The latter mutant was able to phosphorylate the kinase-inactive p37mos(Arg-121) protein in vitro. These results indicate that even though p37mos(Arg-121) can be phosphorylated in trans by other kinase molecules, it lacks the ability to phosphorylate itself in vitro. This provides a compelling argument that the protein kinase function of p37mos is an intrinsic property of the protein. Moreover, since the kinase function correlates with the cellular transformation activity of the v-mos gene, we predict that it is required for the biological activity of the v-mos gene.     

Expression of mouse mammary tumor virus superantigen accelerates tumorigenicity of myeloma cells

To investigate whether superantigen (SAG) from endogenous mouse mammary tumor virus functions as an immunogenic or a tumorigenic factor in tumor development, the BALB/c myeloma cell line FO was transfected with the SAG gene from the 3' Mtv-50 long terminal repeat (LTR) open reading frame (ORF), the product of which was specific for Vbeta6. All five transfectants expressing Mtv-50 LTR ORF mRNA showed stimulatory activity for Vbeta6 T-cell hybridomas in vitro; this activity was inhibited by the addition of anti-Mtv-7 monoclonal antibody (MAb) or anti-major histocompatibility complex class II I-A(d) and I-E(d) MAb. All transfectants with the SAG gene grew more rapidly than did mock transfectants in BALB/c mice after subcutaneous inoculation, whereas all clones, including mock transfectants, grew equally well in athymic nude mice. A significant fraction of Vbeta6 T cells selectively expressed activation markers, including CD44(high), CD62L(low), and CD69(high), and produced large amounts of interleukin 5 (IL-5) and IL-6 in BALB/c mice inoculated with transfectants. These results suggested that the expression of viral SAG enhances the tumorigenicity of a myeloma cell line through the stimulation of SAG-reactive T cells.     

Comparison of myeloproliferative sarcoma virus with Moloney murine sarcoma virus variants by nucleotide sequencing and heteroduplex analysis.

The myeloproliferative sarcoma virus (MPSV) was derived by passage of Moloney sarcoma virus (Mo-MuSV) in adult mice. Mo-MuSV variants transform fibroblasts. However, MPSV also affects erythroid, myeloid, and hematopoietic stem cells. The MPSV proviral genome, two temperature-sensitive mutants derived from it, Mo-MuSV variant M1, and Moloney murine leukemia virus (Mo-MuLV) were compared by heteroduplex mapping. MPSV wild type was found to have 1 kilobase pair deleted from the pol gene and to contain v-mos-related sequences. The 3' end of MPSV, including the oncogene-helper junctions, the v-mos gene, and the 3' long terminal repeat, was sequenced and compared with sequences of Mo-MuLV, MSV-124, and the mouse oncogene c-mos. From these data, MPSV appears to be either closely related to the original Mo-MuSV or an independent recombinant of Mo-MuLV and c-mos. Five possible explanations of the altered specificity of MPSV are considered. (i) The MPSV mos protein has properties inherent in c-mos but lost by other Mo-MuSV mos proteins. (ii) The MPSV mos protein has altered characteristics due to amino acid changes. (iii) Due to a frameshift, MPSV codes for a mos protein truncated at the amino terminal and also a novel peptide. (iv) A second novel peptide may be encoded from the 3' env region. (v) MPSV has long terminal repeats and an enhancer sequence more like Mo-MuLV than Mo-MuSV, with a consequently altered target cell specificity.     

Recombinational junctions of variants of Moloney murine sarcoma virus: generation and divergence of a mammalian transforming gene

Different variants of Moloney murine sarcoma virus (MSV) were examined by nucleotide sequencing to compare the junctions between the acquired cellular sequence, v-mos, and the adjacent virus-derived sequences. These variants included 124-MSV, m1-MSV, and HT1-MSV and also the purportedly independent isolate Gazdar MSV. These four strains have an identical 5' junction between the murine leukemia virus env gene and the v-mos gene. This junction lies within the sixth codon of the chimeric env-mos coding region that encodes the transforming gene product. In contrast, at the 3' junction between the v-mos gene and the murine leukemia virus env gene, the three variants examined here were all different. A small deletion was found in the COOH-terminal portion of the m1-MSV env-mos coding region, indicating that the COOH terminus of this transforming gene product must be different from that of 124-MSV or HT1-MSV. The data presented here are consistent with the thesis that a virus closely related to HT1-MSV was the primordial Moloney MSV, and that all other related strains evolved from it by deletion or rearrangement. The variability observed in the Moloney MSV family is discussed in terms of possible mechanisms for the initial capture of mos sequences by the parental retrovirus and also in comparison with other transforming retrovirus families, such as Abelson murine leukemia virus and Rous sarcoma virus.     

Nucleotide sequence of the transforming gene of m1 murine sarcoma virus.

The v-mosm1 nucleotide sequence codes for a protein that is 376 amino acids long. Although the N-terminus is homologous with that of the v-mos124 protein, the C-terminus is substantially different from the C-termini of all other examined mos proteins, suggesting that this region is nonessential and perhaps cleaved. Overall, v-mosm1 has greater homology with c-mos than does v-mos124, but mutually exclusive differences between c-mos and each of the v-mos genes preclude linear descent and suggest a common ancestral murine sarcoma virus.