News and Publications

Studies in Molecular and Experimental Medicine

Makoto Suzuki is using the phage-display technique to discover receptors on prostate cancer cells. This technique is based on the production of tens of millions of different peptide sequences and selection of the sequences by cultured prostate cancer cells (Fig. 1).

After 4 cycles of "panning" (absorption and desorption of phage particles on the target cells), some consensus sequences were recovered from a 12mer peptide library. No obvious linear sequence similarity was apparent between any of the selected peptides, a finding compatible with the recognition of different cell-surface molecules by different peptides. Of 18 clones sequenced, 5 were identical. Interestingly, 1 clone was detected with different methods of eluting phage from cells.

Detection of the peptides that bind to prostate cancer cells provides an opportunity to identify the receptors to which the peptides bind and to determine whether these receptors are specific to the prostate, specific to cancer, or both. If receptors of sufficient specificity are found, they might be used as targets for tumoricidal drugs; the receptors would attach drug molecules to the targeting sequences. This targeting of drugs could eventually lead to more effective treatments of prostate cancer.

In an article in the Journal of Biological Chemistry, Jennifer Johnson reports the importance of phosphorylation of the p47^phox subunit of NADPH oxidase on serine 303 or 304. The results of her studies suggest that the effect of phosphorylation is related to either the increase in hydrophilicity around serines 303 and 304 or that activation involves the formation of a metal bridge between the phosphorylated serines and another region of the protein. Dr. Johnson also showed that p47^phox-deficient B lymphoblasts that express the p47^phox S(359,370) double mutation have dramatically lower levels of enzyme activity and phosphorylation of p47^phox than do p47^phox-deficient B lymphoblasts that express wild-type p47^phox. In addition, the mutant p47^phox does not translocate to the plasma membrane when the cells are stimulated.

In contrast, normal phosphorylation and translocation occur in mutants containing aspartate or glutamate at positions 359 and 370, but oxidase activity is still greatly reduced. These results imply that a negative charge at position 359, position 370, or both is sufficient to allow the phosphorylation and translocation of p47^phox, but that features unique to a phosphorylated hydroxyamino acid are required to support production of superoxide.

Barbara Nowakowski is analyzing a fusion protein, PAX3-FKHR, associated with alveolar rhabdomyosarcoma, a soft tissue tumor in children. This protein is generated by a chromosomal rearrangement that results in the fusion of the amino-terminal DNA-binding domains of PAX3 and the carboxy-terminal transcriptional regulatory domain of the winged helix protein FKHR. PAX3-FKHR, expressed by the avian retroviral vector RCAS, induces foci of transformed cells and anchorage-independent growth in cultures of chicken embryo fibroblasts. The fusion protein is a stronger transcriptional activator than is PAX3. The gain-of-function mutation suggests a potential mechanism for oncogenesis.

Dr. Nowakowski has begun an analysis of a correlation between transactivation and transformation in chicken embryo fibroblasts. She has constructed a series of deletion mutants of PAX3-FKHR to determine the molecular domains required for transformation of the fibroblasts. Comparisons of DNA binding, transactivation, and transformation by the various mutant proteins will be important to an understanding of the method by which the fusion gene alters normal gene expression and causes tumor formation.

Makoto Nishizawa is continuing his studies of oncogenic transcription factors that belong to the BZip family of proteins. He has done a genetic analysis of Maf, a protein that plays an important role in the differentiation of hematopoietic cells. Maf has cancer-inducing potential, and Dr. Nishizawa's genetic work has uncovered mutations that activate this oncogenicity. He is currently concentrating on the role of protein dimerization and DNA binding in oncogenic transformation. For this analysis, he has constructed chimeric proteins in which the dimerization and DNA-binding domains of Maf have been replaced with corresponding domains of other, related BZip proteins such as Jun, Fos, and yeast GCN4 transcription factors. The chimeric proteins differ in their DNA targets and in their ability to form heterodimers or homodimers. Among these factors, Maf, Jun, and Fos are all oncogenic, but they induce different tumors: musculoaponeurotic fibrosarcoma, fibrosarcoma, and osteosarcoma, respectively. The yeast transcription factor GCN4 is identical with Jun in DNA-binding specificity but is not transforming. Dr. Nishizawa is now testing the transforming ability of the chimeric molecules in vitro and the specificity of tumor formation in vivo. The information obtained with these constructs will be important in identifying downstream target genes that are upregulated or downregulated by transcription factors of malignant transformation and may lead to better understanding and treatment of cancer.

Valerie Pasquetto is studying hepatitis B virus (HBV). After antigen recognition, HBV-specific cytotoxic T lymphocytes (CTLs) induce a necroinflammatory liver disease in HBV transgenic mice. An early event in this process is the activation of apoptosis by the CTLs in a fraction of the hepatocytes that are positive for the HBV surface antigen. In addition, the CTLs inhibit HBV replication by noncytopathically eliminating viral nucleocapsids from the cytoplasm of the remaining viable hepatocytes by a process that depends on IFN-Γ and TNF-α. Dr. Pasquetto's studies were undertaken to determine the fate of cytoplasmic nucleocapsids and their cargo of replicative intermediates in the apoptotic hepatocytes, because the elimination of HBV during viral hepatitis traditionally was thought to require the destruction of infected cells.

Primary cultures of HBV transgenic mouse hepatocytes were incubated in vitro in the presence or absence of cells from a CD8+ CTL clone specific for cells bearing the HBV surface antigen. The cultures were harvested at various intervals to monitor cell survival and capsid-associated HBV replication. Concomitant with the induction of apoptosis in less than 1 hour and the destruction of all hepatocytes in less than 4 hours, the capsid-associated replicative intermediates were shifted from the viable adherent hepatocytes to the apoptotic debris and then to the culture supernatant in an orderly, time-dependent manner. The capsid-associated replicative intermediates were stable in the apoptotic cellular debris and culture supernatant for at least 5 days after the addition of the CTLs, at levels comparable to those detected in the starting hepatocytes.

These results (1) indicate that HBV capsids and their associated replicative intermediates survive the apoptotic destruction of the cells that produce them and (2) confirm that downregulation of HBV replication by CTLs and other cytokine-inducing stimuli requires viable hepatocytes to be effective. The replication competence, potential infectivity, and ultimate fate of these capsids in the presence of macrophages and nucleocapsid-specific antibodies in vivo are under investigation.

Randolph S. Piotrowicz is focusing on the roles of the 27-kD heat-shock protein (HSP27) in vascular endothelial cell behaviors relevant to angiogenesis in breast tumors. In particular, he has investigated the effects of HSP27 expression and phosphorylation on the regulation of cell motility. He showed that enhanced expression of HSP27 in endothelial cells resulted in greater motility. Expression of a nonphosphorylation mutant HSP27 retarded cell migration. HSP27 was localized to dynamic F-actin structures of migrating cells, suggesting that HSP27 regulates the formation of lamellipodial F-actin and thus regulates cell motility.

In addition, the release of the angiogenic hormone fibroblast growth factor 2 (FGF-2) was studied. A novel mechanism for the release of FGF-2 and the capacity for high molecular weight forms of this hormone to inhibit cell migration have been determined. The inhibition is unique because it requires the presence of the estrogen receptor for transduction of the inhibitory signal induced by high molecular weight forms of FGF-2. This work suggests that coupling of peptide and steroid hormone receptors in the transduction of signals by an angiogenic factor can occur in endothelial cells. This possibility is being investigated. These studies may be relevant to the spread of breast cancer in women.

Anna Demina studied a mutation in UGT1A1, a gene important in the disposal of bilirubin, the yellow bile pigment that is the cause of jaundice. The promoter of this gene may have 6 or 7 thymine-adenine (ta) repeats. The 7-repeat promoter is less efficient, and persons who inherit this mutation tend to have a mild form of jaundice known as Gilbert syndrome. Dr. Demina studied the population genetics of this mutation to determine whether the lower frequency of jaundice in Africans was due to a lack of the 7-repeat mutation. However, she found that persons of African ancestry include individuals with 5 or 8 repeats as well as individuals with 6 or 7 repeats. Even though bilirubin levels are lower in Africans than in Europeans, the promoters with lower activity were more common. An incidental finding was a polymorphism at nucleotide 106, tightly linked to the (ta)₅ haplotype. Among persons of Asian ancestry, in whom jaundice is more common, the deficient (ta)₇ form was, surprisingly, less common. Thus, within the Caucasian ethnic group a strong correlation exists between the number of repeats in the promoter and bilirubin level; that is, promoters with higher numbers of repeats are associated with higher levels of bilirubin. However, between ethnic groups, lower numbers of repeats are associated with higher levels of bilirubin.

Serum levels of bilirubin are influenced by many factors, both genetic and environmental. These data suggest that the unstable UGT1A1 polymorphism may "fine tune" the plasma level of bilirubin within population groups, maintaining a level that is high enough to provide protection against oxidative damage but sufficiently low enough to prevent kernicterus in infants. The differences in bilirubin levels that still exist between racial groups, despite changes in the number of promoter repeats, imply that the changes have occurred sufficiently recently that full compensation has not yet been achieved.