Daniel DiMaio, MD, PhD

Bovine papillomavirus E5 protein

The laboratory is interested in determining the mechanism of action of viral oncogenes, with the belief that these studies will provide new insight into signal transduction and the control of cell proliferation. For many years, we have studied the E5 oncoprotein of bovine papillomavirus (BPV), a small DNA tumor virus closely related to the human papillomaviruses (HPV) that cause a variety of human cancers. The forty-four amino acid BPV E5 protein is the shortest protein known to cause tumorigenic transformation of cells. In transformed cells, the E5 protein exists as a dimeric transmembrane protein with a very hydrophobic central domain that spans the membranes of the Golgi apparatus and endoplasmic reticulum.

We discovered that the E5 protein employs a unique mechanism to transform cells. It specifically binds to the transmembrane domain of the platelet-derived growth factor ß receptor, a cellular receptor tyrosine kinase, thereby causing receptor activation. The E5/PDGF receptor interaction results in growth stimulation and tumorigenic transformation of cells. We also showed that direct interactions involving specific transmembrane and juxtamembrane amino acids in the E5 protein and the PDGF ß receptor result in dimerization and trans-phosphorylation of the receptor, and recruitment of cellular signaling molecules into a signal transduction complex. Thus, the E5 protein acts as a specific, intramembrane crosslinker of the PDGF ß receptor. These findings demonstrated that receptor tyrosine kinases can be activated by proteins that do not resemble their normal ligands and that specific transmembrane interactions can drive receptor activation and tumor formation. We are conducting genetic and biochemical experiments to define the structure of this unique complex, in the belief that it will provide new insight into the assembly of transmembrane protein complexes. In particular, we are attempting to determine the molecular code that allows two transmembrane proteins to specifically recognize one another.

Manipulation of cell behavior with small transmembrane proteins modeled on the E5 protein

Using the BPV E5 protein as a model, we have devised a genetic approach that allows us to construct and identify biologically active small proteins that have never existed in nature. We used a PCR-based method to construct complex retrovirus libraries in which the central transmembrane domain of the E5 protein is replaced with random hydrophobic amino acids. We then use various genetic selection or physical enrichment strategies to isolate rare clones from these libraries that express the desired biological activities. We found that a surprisingly large fraction of random transmembrane domains can support cell transformation by interacting with the PDGF ß receptor. Some of these proteins share virtually no amino acids with the wild-type viral E5 protein and display subtle differences in specificity compared to the E5 protein. We have also used this approach to construct small transmembrane proteins that activate the erythropoietin receptor and drive red blood cell formation or that down-regulate the HIV co-receptor CCR5 and dramatically inhibit HIV infection. Strikingly, the HIV inhibitors all have entirely different transmembrane sequences and display subtle differences in biological activities, suggesting that we have constructed a diverse set of proteins to probe CCR5 function and HIV infection. We are studying the mechanism of action of these novel proteins, which may serve as the basis for new strategies to support red blood cell production or to prevent or treat HIV infection. Because 30% of all cell proteins are integral membrane proteins, there are many potential targets of this approach.

Studies of tumor virus infection

We have used RNA interference and other genetic techniques to identify cellular genes required for infection by tumor viruses. We discovered that the major endoplasmic reticulum chaperone, BiP, and DNAJ co-chaperones in the ER are required for infection by SV40 and other polyomaviruses, which are widespread in the human population and can cause fatal disease. We have isolated SV40 mutants that utilize a different cellular receptor and display altered tropism and cytopathic effects on cells. We have also conducted a genome-wide siRNA screen for genes required for infection by the high-risk HPV and have identified numerous cellular factors required for this process. Bioinformatics and genetic analysis suggest for the first time that the retrograde transport pathway from the endosomes to the Golgi apparatus is important for HPV entry. These studies will uncover new aspects of infection by these viruses and may suggest new antiviral strategies. We recently showed that the HPV L2 protein binds directly to retromer, a cytoplasmic trafficking factor that delivers the virus into the retrograde pathway. To accomplish this, after the incoming virus is endocytosed and resides in the endosomal lumen, a cell-penetrating peptide on the end of L2 protrudes through the endosomal membrane into the cytoplasm so that it can bind retromer.

Studies of cervical carcinoma and senescence

The laboratory is investigating the role of the HPV E6 and E7 proteins in cervical cancer. We showed that expression of the bovine papillomavirus E2 transcription factor represses expression of the E6 and E7 genes from the integrated HPV genomes in human cervical cancer cells. This causes transient activation of the endogenous p53 and retinoblastoma tumor suppressor pathways and dramatic growth inhibition. The growth-arrested cells do not die but rather enter an irreversible non-replicating state that resembles replicative senescence, which normally blocks of the unlimited proliferation of normal cells. These results demonstrated that continued expression of the HPV oncogenes is required to maintain the proliferative state of these cancer cells and imply that manipulations that inhibit the expression or activity of the viral oncoproteins may be novel approaches to treat cancer.

Further experiments showed that repression of the E7 protein is sufficient to induce a robust senescence response that is dependent on the retinoblastoma pathway. We have also identified a new regulatory circuit in which specific cellular microRNAs inhibit the cellular B-Myb oncogene during senescence and showed that this circuit affects the senescence phenotype.