Functional domains in cyclin D1: pRb-kinase activity is not essential for transformation.

Although cyclin D1 plays a major role during cell cycle progression and is involved in human tumourigenesis, its domain structure is still poorly understood. In the present study, we have generated a series of cyclin D1 N- and C-terminal deletion constructs. These mutants were used to define the domains required for transformation of rat embryonal fibroblasts (REF) in cooperation with activated Ha-ras and and to establish correlations with defined biochemical properties of cyclin D1. Protein binding and REF assays showed that the region of the cyclin box required for the interaction with CDK4 as well as C-terminal sequences determining protein stability were crucial for transformation. Surprisingly, however, the N-terminal deletion of 20 amino acids which impaired pRb kinase activity did not affect the transforming ability of cyclin D1. Likewise, no effect on transformation was observed with mutants defective in p21CIP interaction. These observations argue against a crucial role of pRb inactivation or p21CIP squelching in cyclin D1-mediated transformation.

Immune responses to viral antigens versus transgene product in the elimination of recombinant adenovirus-infected hepatocytes in vivo.

Human adenoviruses have been developed as an attractive vehicle for in vivo liver-directed gene therapy. Problems with the application of first generation recombinant adenoviruses to liver-directed gene therapy have been transient expression of the recombinant gene and development of hepatitis. Previous studies in mouse models of gene transfer to liver and lung suggested that MHC class I-restricted cytotoxic T lymphocytes (CTLs) to viral antigens may be effectors in the elimination of transgene expression. The goal of this study was to evaluate the importance of viral antigens versus transgene product in inducing CTL mediated hepatocyte destruction in vivo. Immunization of C57BL/6 mice with a lacZ-expressing adenovirus elicited CTL responses to both viral antigens and the transgene product, beta-galactosidase (beta-gal). Adoptive transfer experiments, as well as studies involving lacZ-transgenic mice (ROSA-26) revealed that CTLs to viral antigens are sufficient to destroy virus-infected hepatocytes, indicating that CTLs to beta-gal can not solely account for the observed hepatocyte destruction that has characterized the use of first generation viruses. In addition, we confirmed that B cell-mediated events do not participate in destruction of hepatocytes in vivo, despite the production of virus- and beta-gal-specific antibodies. These data confirm the hypothesis that viral gene expression elicits host responses that contribute to the problem of transgene instability. Recombinant adenoviruses must be redesigned to diminish viral gene expression if they are to be used in the treatment of chronic diseases.

Deregulation of genes encoding microfilament-associated proteins during Fos-induced morphological transformation.

The mechanism of Fos-induced transformation is still poorly understood. In the present study, we have asked whether genes whose products play a role in determining cell morphology might become deregulated in the course of Fos-induced transformation. A clear up-regulation in Fos-transformed rat fibroblasts was seen with ezrin, as well as tropomyosin (TM) -3 and -5B, while TM-1 was down-regulated. Significantly, the same genes were deregulated in a very similar, but hormone-inducible way in cells expressing a Fos-estrogen receptor fusion protein. In agreement with these results, Fos-expressing cells showed decreased levels of two TM isoforms of 36 and 38 kDa, and showed an impaired TM network. The significance of these observations is strengthened by the fact that the deregulation of TM expression has been shown to contribute to morphological transformation in other experimental systems. Deregulation of the TM and ezrin genes preceeds the induction of morphological transformation suggesting that this deregulation is not merely a consequence of transformation. On the other hand, deregulation follows the induction of direct Fos target genes. We therefore propose that a cascade of regulatory events is triggered by Fos oncoproteins which eventually leads to the deregulation of genes encoding cytoskeleton-associated proteins.

An autonomous N-terminal transactivation domain in Fos protein plays a crucial role in transformation.

To date, three functional domains have been defined in c-Fos and v-Fos proteins and have been shown to play a role in transactivation: the leucine zipper mediating hetero-dimerization, the basic DNA contact site, and a C-terminally located transactivation domain (C-TA) harbouring the HOB1 and HOB2 motifs. While the bZip region, consisting of the leucine zipper and the DNA contact site, is indispensable for transformation, the C-TA domain is not required and is actually altered by internal deletions in the FBR-MuSV. We now show that the N-terminal regions of c-Fos and v-Fos contain a second transactivation domain (N-TA). A functionally crucial motif within the N-TA domain, termed NTM, was pinpointed to a approximately 25 amino acid stretch around positions 60-84 which is highly conserved in FosB. Analysis of LexA fusion proteins showed that the N-TA domains of both c-Fos and FosB function in an autonomous fashion in both fibroblasts and yeast. Most importantly, deletion of the NTM motif impairs the transforming properties of v-Fos. Apart from the bZip region, the N-TA domain is the only functional domain required for transformation by v-Fos, at least when its expression is driven by the strong FBR-MuSV-LTR promoter.

Mutual transrepression of Fos and the glucocorticoid receptor: involvement of a functional domain in Fos which is absent in FosB.

In this study, we show that Fos protein can repress transactivation by the glucocorticoid receptor (GR). In addition, we demonstrate that GR is capable of inhibiting, in a hormone-dependent fashion, Fos-mediated transactivation of AP-1 dependent transcription. Moreover, repression of the serum response element by Fos is abolished by the GR in the presence of hormone. Transrepression of glucocorticoid mediated induction involves a region of Fos, located between amino acids 40 and 111, to which no function has been previously assigned, and which is poorly conserved among Fos, FosB and Fra-1. In agreement with this finding, FosB is not capable of transrepressing GR activation of transcription, representing the first functional difference between Fos and FosB. We have mapped the domain of the GR which is required for repression of AP-1 dependent transcription, to the region of central DNA binding domain. Our results suggest that Fos and the GR may form transcriptionally inactive complexes and point to a regulatory interrelationship between different signal transduction pathways.