application): All organisms from eubacteria to eukaryotes have mechanisms for degrading transcripts that contain a premature termination codon. The process of nonsense-mediated RNA decay (NMRD) has been best studied in the yeast S. cerevisiae, but some aspects of NMRD appear to be significantly different in mammals. In particular, as Dr. Dietz discovered (and others have subsequently confirmed) premature nonsense codons can alter nuclear mRNA splicing patterns. An attractive idea is that NMRD evolved as a surveillance mechanism to protect cells from the harmful effects of truncated proteins; differences between yeast and mammals might then reflect the abundance of introns in mammals and the paucity of introns in yeast. Dr. Dietz has obtained good evidence for this in the case of Marfan's syndrome, where carboxyterminal truncated fibrillin protein (FBN1) interferes with organization of the extracellular matrix. The mechanism by which splicing (presumably a nuclear process) can be affected by translation (presumably a cytoplasmic process) is a mystery, and many models have been invoked to explore this observation. Perhaps the key question regarding NMRD is whether the ribosome or some other device is responsible for scanning the ORFs; the ability of a suppressor tRNA to partially suppress NMRD argues for a ribosomal role, but stranger things have happened. Three yeast proteins called UPF1, 2, and 3 (for up-frameshift) have been identified genetically as essential for NMRD; Dr. Dietz has now cloned the human homologue of UPF1, and named it RENT1 (for regulation of nonsense transcripts). This application is designed to explore three aspects of RENT1 function: (1) Where, when, and how is RENT1 expressed in the mouse, and will putative dominant negative mutants of RENT1 (modeled on known dominant negatives in yeast UPF1p) stabilize reporter constructs with nonsense mutations? (2) What is the phenotype of a mouse RENT1 knockout, and does loss of NMRD in the whole animal affect expression of various natural or artificial mRNAs? As an internal control to monitor levels of NMRD, Dr. Dietz will use a spontaneous frameshift mutation known as gusmps which causes a 200-fold decrease in murine beta-glucuronidase mRNA levels. Similar controls with the R11X nonsense mutation in the murine 4-hydroxyphenylpyruvic acid dioxygenase gene are also proposed. (3) In a tissue culture or mouse model, can down regulation of NMRD by a RENT1 dominant negative rescue a disease phenotype caused by a premature nonsense codon, or even be used to detect new disease genes? The D836X nonsense mutation (or similar mutations) in the CFTR (cystic fibrosis transmembrane conductance channel) will provide a therapeutically important test case to monitor the success of this approach. In addition, using isogenic fibroblast lines that do or do not express a dominant negative RENT1 allele, Dr. Dietz will attempt to """"""""reestablish"""""""" by RDA (representational difference analysis) that a nonsense codon is responsible for the OAT defect in a gyrate atrophy patient.
