We have identified a new member of the maf oncogene family and named it mafB. This gene is expressed in a wide variety of tissues and encodes a protein of 311 amino acids containing a typical bZip motif in its carboxy-terminal region. In the bZip domain, MafB shares extensive homology not only with v-Maf but also with other Maf-related proteins. As expected from its structure, MafB forms a homodimer through its leucine repeat structure and specifically binds Maf-recognition elements (MAREs). In addition, MafB forms heterodimers with v-Maf and Fos through its zipper structure. However, unlike v-Maf, MafB fails to associate with Jun. Transient cotransfection assays revealed that both v-Maf and MafB act as transactivators for a promoter linked to MAREs, although MafB is less potent than v-Maf. As is the case for the c-maf gene, overexpression of the mafB gene induces transformation of chicken embryo fibroblasts in vitro. Through formation of numerous bZip dimers, the Maf family proteins along with the AP-1 components should provide great diversity in transcriptional regulation for a wide variety of genes.

We have molecularly cloned the provirus of the avian musculoaponeurotic fibrosarcoma virus AS42.

The maf oncogene encodes a bZip nuclear protein which recognizes sequences related to an AP-1 site either as a homodimer or as heterodimers with Fos and Jun. We describe here a novel maf-related gene, mafG, which shows extensive homology with two other maf-related genes, mafK and mafF. These three maf-related genes encode small basic-leucine zipper proteins lacking the trans-activator domain of v-Maf. Bacterially expressed small Maf proteins bind to DNA as homodimers with a sequence recognition profile that is virtually identical to that of v-Maf. As we have previously described, the three small Maf proteins also dimerize with the large subunit of NF-E2 (p45) to form an erythroid cell-specific transcription factor, NF-E2, which has distinct DNA-binding specificity. This study shows that the small Maf proteins can also dimerize among themselves and with Fos and a newly identified p45-related molecule (Ech) but not with v-Maf or Jun. Although the small Maf proteins preferentially recognize the consensus NF-E2 sequence as heterodimers with either NF-E2 p45, Ech, or Fos, these heterodimers seemed to be different in their transactivation potentials. Coexpression of Fos and small Mafs could not activate a promoter with tandem repeats of the NF-E2 site. These results raise the possibility that tissue-specific gene expression and differentiation of erythroid cells are regulated by competition among Fos, NF-E2 p45, and Ech for small Maf proteins and for binding sites.The maf oncogene was identified by structural analysis of the genome of the AS42 avian transforming retrovirus (26,40). It encodes a nuclear basic-leucine zipper (bZip) protein which can form a homodimer through its zipper structure (23). Recently, we reported that the v-Maf homodimer specifically recognizes two relatively long palindromic DNA sequences, TGCTGACTCAGCA and TGCTGACGTCAGCA, at roughly equal efficiency (24). The middle parts of the two consensus binding sequences for Maf are identical with two well-characterized binding sequences of the AP-1 transcription factor, the 12-O-tetradecanoylphorbol-13-acetate (TPA)-responsive element (TRE; TGACTCA) and the cyclic AMP-responsive element (CRE; TGACGTCA), respectively. We therefore named the two types of recognition elements TRE-type Maf recognition elements (MAREs) and CRE-type MAREs. It was also recently revealed that Maf forms heterodimers with the two major components of AP-1, Fos and Jun (24, 27, 28). These heterodimers preferably bind to asymmetric DNA sequences consisting of the two consensus binding sequences of Maf homodimer and AP-1 (24). Thus, Maf and the two AP-1 components are suggested to interact with each other in a cooperative or inhibitory way in association with their recognition sequences by forming heterodimers of altered binding specificities.Like many other proto-oncogenes, the c-maf gene is a member of a gene family. To date, four maf-related genes, mafK, mafF, mafB, and nrl, have been reported (14,22,53). Their gene products are closely related to v-Maf especially in the structur...

No abstract

Peroxisomeproliferator-activated receptor gamma (PPARc) agonists demonstrate antitumor activity likely through transactivating genes that regulate cell proliferation, apoptosis, and differentiation. The PAX8/PPARc fusion oncogene, which is common in human follicular thyroid carcinomas appears to act via dominant negative suppression of wild-type PPARc, suggesting that it may be a tumor suppressor gene in thyroid cells. We have identified a novel high-affinity PPARc agonist (RS5444) that is dependent upon PPARc for its biological activity. This is the first report of this molecule and its antitumor activity. In vitro, the IC 50 for growth inhibition is B0.8 nM while anaplastic thyroid carcinoma (ATC) tumor growth was inhibited threeto fourfold in nude mice. siRNA against PPARc and a pharmacological antagonist demonstrated that functional PPARc was required for growth inhibitory activity of RS5444. RS5444 upregulated the cell cycle kinase inhibitor, p21 WAF1/CIP1 . Silencing p21 WAF1/CIP1 rendered cells insensitive to RS5444. RS5444 plus paclitaxel demonstrated additive antiproliferative activity in cell culture and minimal ATC tumor growth in vivo. RS5444 did not induce apoptosis but combined with paclitaxel, doubled the apoptotic index compared to that of paclitaxel. Our data indicate that functional PPARc is a molecular target for therapy in ATC. We demonstrated that RS5444, a thiazolidinedione (Tzd) derivative, alone or in combination with paclitaxel, may provide therapeutic benefit to patients diagnosed with ATC.

The genomic RNA and intracellular RNA of mouse hepatitis virus, strain JHM (MHV-JHM) and two plaque mutants (1a and 2c), which have been isolated from a persistently infected culture (JHM-CC), have been analyzed by T1-resistant oligonucleotide finger-printing. The genomic RNA of the virus population (JHM-CC virus) released from different passage levels of the same persistent infection has also been analyzed. The analysis shows the locations within the genomic and intracellular RNAs of more than 45 T1-resistant oligonucleotides and confirm earlier studies (J. L. Leibowitz, K. C. Wilhelmsen, and C. W. Bond (1981), Virology 114, 39-51), showing that the six subgenomic RNAs of MHV-JHM form a 3' coterminal nested set which extends for different lengths in a 5' direction. The analysis also identifies in each subgenomic RNA those large T1 oligonucleotides derived from noncontiguous regions of the genome during mRNA synthesis. Two important conclusions can be reached from analysis of the mutant viruses. First, the virus population released from the persistent infection represents a fairly constant mixture of viruses, and the fluctuating emergence of variants as predominant species in the culture does not occur. Second, the data indicate that for particular intracellular RNAs of mutant viruses the sequence rearrangements occurring during subgenomic mRNA synthesis are different from those in the corresponding intracellular RNA of wild-type virus. The result may indicate a potential flexibility in the leader/body fusion process that has not been previously recognized.

The intracellular defective RNAs generated during high-multiplicity serial passages of mouse hepatitis virus JHM strain on DBT cells were examined. Seven novel species of single-stranded polyadenylic acid-containing defective RNAs were identified from passages 3 through 22. The largest of these RNAs, DIssA (molecular weight [mw], 5.2 x 106), is identical to the genomic RNA packaged in the defective interfering particles produced from these cells. Other RNA species, DIssBl (mw, 1.9 x 106 to 1.6 x 106), DIssB2 (mw, 1.6 x 106), DIssC (mw, 2.8 x 106) DIssD (mw, 0.82 x 106), DIssE (mw, 0.78 x 106), and DIssF (mw, 1.3 x 106) were detected at different passage levels. RNase Tl-resistant oligonucleotide fingerprinting demonstrated that all these RNAs were related and had multiple deletions of the genomic sequences. They contained different subsets of the genomic sequences from those of the standard intracellular mRNAs of nondefective mouse hepatitis virus JHM strain. Thus these novel intracellular viral RNAs were identified as defective interfering RNAs of mouse hepatitis virus JHM strain. The synthesis of six of the seven normal mRNA species specific to mouse hepatitis virus JHM strain was completely inhibited when cells were infected with viruses of late-passage levels. However, the synthesis of RNA7 and its product, viral nucleoprotein, was not significantly altered in late passages. The possible mechanism for the generation of defective interfering RNAs was discussed.

After six to eight serial undiluted passages of mouse hepatitis virus (JHM strain) in DBT cell culture, a decrease in the yield of infectious virus occurred, and with further passages fluctuating yields of infectious virus were observed. The serially passaged virus interfered with the multiplication of the standard JHM virus, but not with vesicular stomatitis virus. After sucrose equilibrium centrifugation of high passage virus, a single peak contained both infectious virus and interfering activity. This virus population resembled the original JHM virus in its structural proteins, but it contained an increased proportion of a protein with a molecular weight of 65 X 10(3). Genomic RNA from standard JHM virus contained a single species of RNA with a molecular weight of 5.4 X 10(6). After five undiluted passages, however, the virion population contained two RNA species with molecular weights of 5.4 X 10(6) and 5.2 X 10(6). RNase T1 resistant oligonucleotide finger-printing of these RNAs showed that the lower molecular weight RNA had lost several oligonucleotide spots that were present in the genomic RNA of the standard JHM virus. After several serial diluted passages of passage 10 virus, a single virus population was obtained which again had only standard virus RNA with a molecular weight of 5.4 X 10(6) and lacked interfering activity. These results indicated that defective interfering particles were generated by serial undiluted passages of JHM virus.