New cellular interactions in transactivation by Py-LT

The large tumorantigens (LT) of SV40 and polyomavirus are pleiotropic proteins being involved in S-phase induction of the host cell and in replication of the viral genomes. They consist of several domains which bind cellular and viral targets with functional consequences. The N-terminal region carries a motif called J-domain resembling the DNA K-binding DNA J protein of E. coli which is involved in binding the chaperone HSC 70. The second important region binds the retinoblastoma protein pRB and its relatives, substantial regulators of gene expression mediated by the E2F family of transcription factors. The C-terminal part encompasses a sequence which mediates binding of LT to the origin of replication of viral DNA and enzymatic functions necessary for DNA replication, namely ATP binding, ATPase and helicase. It also carries a C2 H2 -type zinc finger, a mutation of which abolishes the formation of hexamers at the origin and thus interfers with repliation of viral DNA. This bipartite arrangement led to the suggestion that the N-terminal part of LT is responsible for transactivation of cellular genes whose products are necessary for S phase induction (DNA synthesis enzymes and cyclins E and A) while the C-terminus functions in replication of viral DNA. In our analysis of the mechanism of transactivation of cyclin A by polyomavirus large and small tumorantigens we found that mutations in the zinc finger of LT completely abolish this transactivation. This argues for an essential function of the C-terminus in transactivation. This is supported by our previous observation that the C-terminus directly binds the co-activator protein CBP/p300 requiring a region around the Zn finger and at the very C-terminus of the protein. We now found that mutation of the Zn finger not only abolishes oligomerization but also binding of p300. The same holds for a mutation of a proline residue within the ATPase region of LT. We assume that both, the mutation of the Zn finger and that of proline residues result in structural changes of the C-terminal domain of LT which interfere with oligomerization and binding of p300. Therefore these mutations do not allow us to decide wether oligomerization or binding of p300 (and/or other cellular proteins, such as E2F) or both are fundamental for transactivation. We want to study this question by producing more mutations around the zinc finger and at the C-terminal region past amino acid 650 for which we have evidence for their involvement in p300 binding. We hope to find mutants which dissociate these two properties, i.e. which are defective in binding p300 but still oligomerize (or vice versa). Such mutants might also provide evidence for other cellular targets of LT. We expect that this study will extend our knowledge about the mechanisms by which viral proteins deregulate cellular processes. We also want to lay the ground for a more detailed structural analysis of LT by creating Baculovirus constructs of wild type and mutated forms of LT which are presumably structurally altered (Zn finger and prolines). This should allow the production of larger amounts of these proteins.

Polyomaviridae are small DNA tumorviruses capable of transforming normal cells into tumorigenic ones. Members of this group are the human viruses JC and BK, the monkey virus Simian virus 40 and the mouse virus polyoma. They are very similar in structure. All code for two proteins (tumor antigens) which are important for virus multiplication. In particular one of the proteins (the large tumor antigen, LT) can activate the expression of cellular genes by interacting with key regulators of cell growth. Thereby quiescent cells are driven into the cell cycle, a condition essential for virus replication, at the same time, however, also responsible for the tumorigenic activity of the viruses. The major target of LT is a tumorsuppressor protein called retinoblastoma protein (Rb). Inactivation of Rb by mutation can cause tumors (retinoblastomas) in children. LT protein also inactivates this tumorsuppressor which consequently leads to the transcriptional activation of genes importartant for growth and cell cycle. Up to now, this inactivation was considered sufficient for this transactivation competence of LT. We show in this project that this assumption is incorrect, rather a second function of LT is essential in addition. This second function consists in the binding to LT of a transcriptional co-activator protein which functions as a histone acetyl transferase (HATs). Such enzymes modify histones, the nuclear proteins which wrap the DNA more or less tightly leading to a structure called chromatin. HATs thereby contribe to the regulation of gene expression. We found that without interaction with HAT complexes, LT can not transactivate genes, although the tumorsuppressor is inactivated. Apparently, LT brings HATs to the transcription start site of genes via its binding to the tumorsuppressor and causes not only the inactivation of the tumorsuppressor but at the same time also a modification of histone and, in fact, we could prove LT-dependent acetylation of histones in the regulatory region of one gene which is transactivated by the viral protein. Our results for the first time show that viral proteins, capable to activate gene expression, can also change the structure of chromatin. Considering the fact that human polyomaviridae are widespread in the population and that their contribution in the development of certain cancers is discussed since many years (although a definitive proof for this contribution is still lacking) our new data are a significant new step towards the clarification of the capacities of these viruses.