Faculty Bibliography

The Ty1 retrotransposon of Saccharomyces cerevisiae is bounded by long-terminal repeats (LTRs). We have constructed a variety of Ty1 elements in which the LTR length has been increased from the normal length of 334 bp to > 2 kb. Although small insertions in the LTR have minimal effects on transposition frequency, larger insertions dramatically reduce it. Nevertheless, elements with long LTRs are incorporated into the genome at a low frequency. Most of these rare insertion events represent Ty1 tandem (head to tail) multimers.

The cDNAs and genes encoding the intron lariat-debranching enzyme were isolated from the nematode Caenorhabditis elegans and the fission yeast Schizosaccharomyces pombe based on their homology with the Saccharomyces cerevisiae gene. The cDNAs were shown to be functional in an interspecific complementation experiment; they can complement an S. cerevisiae dbr1 null mutant. About 2.5% of budding yeast S. cerevisiae genes have introns, and the accumulation of excised introns in a dbr1 null mutant has little effect on cell growth. In contrast, many S. pombe genes contain introns, and often multiple introns per gene, so that S. pombe is estimated to contain approximately 40 times as many introns as S. cerevisiae. The S. pombe dbr1 gene was disrupted and shown to be nonessential. Like the S. cerevisiae mutant, the S. pombe null mutant accumulated introns to high levels, indicating that intron lariat debranching represents a rate-limiting step in intron degradation in both species. Unlike the S. cerevisiae mutant, the S. pombe dbr1::leu1+ mutant had a severe growth defect and exhibited an aberrant elongated cell shape in addition to an intron accumulation phenotype. The growth defect of the S. pombe dbr1::leu1+ strain suggests that debranching activity is critical for efficient intron RNA degradation and that blocking this pathway interferes with cell growth.

Generalized transcriptional repression of large chromosomal regions in Saccharomyces cerevisiae occurs at the silent mating loci and at telomeres and is mediated by the silent information regulator (SIR) genes. We have identified a novel form of transcriptional silencing in S. cerevisiae in the ribosomal DNA (rDNA) tandem array. Ty1 retrotransposons marked with a weakened URA3 gene (Ty1-mURA3) efficiently integrated into rDNA. The mURA3 marker in rDNA was transcriptionally silenced in a SIR2-dependent manner. MET15 and LEU2 were also partially silenced, indicating that rDNA silencing may be quite general. Deletion of SIR4 enhanced mURA3 and MET15 silencing, but deletion of SIR1 or SIR3 did not affect silencing, indicating that the mechanism of silencing differs from that at telomeres and silent mating loci. Deletion of SIR2 resulted in increased psoralen cross-linking of the rDNA in vivo, suggesting that a specific chromatin structure in rDNA down-regulates polymerase II promoters.

The two-hybrid system was used to define regions of the Ty1 Gag protein responsible for multimerization. Gag truncations lacking the first 146 or the last 97 amino acids (Gag is 440 amino acids in length) interact. A severely C-terminally truncated molecule (lacking the last 207 amino acids) was the smallest truncation to interact, suggesting that some protein-protein interactions between Gag molecules are mediated through the first 233 amino acids. However, an internal deletion of amino acids 147 to 233 does not abolish Gag-Gag interaction, indicating that more than one region can mediate Gag interaction. Surprisingly, we found that a truncation lacking the last 97 amino acids interacts with itself but not with full-length Gag. This is apparently due to an artifact of the two-hybrid assay, since these same molecules coassemble with wild-type Gag into Ty1 virus-like particles.

We previously isolated two human L1 elements (L1.2 and LRE2) as the progenitors of disease-producing insertions. Here, we show these elements can actively retrotranspose in cultured mammalian cells. When stably expressed from an episome in HeLa cells, both elements retrotransposed into a variety of chromosomal locations at a high frequency. The retrotransposed products resembled endogenous L1 insertions, since they were variably 5' truncated, ended in poly(A) tracts, and were flanked by target-site duplications or short deletions. Point mutations in conserved domains of the L1.2-encoded proteins reduced retrotransposition by 100- to 1000-fold. Remarkably, L1.2 also retrotransposed in a mouse cell line, suggesting a potential role for L1-based vectors in random insertional mutagenesis.

Human L1 elements are highly abundant poly(A) (non-LTR) retrotransposons whose second open reading frame (ORF2) encodes a reverse transcriptase (RT). We have identified an endonuclease (EN) domain at the L1 ORF2 N-terminus that is highly conserved among poly(A) retrotransposons and resembles the apurinic/apyrimidinic (AP) endonucleases. Purified L1 EN protein (L1 ENp) makes 5'-PO4, 3'-OH nicks in supercoiled plasmids, shows no preference for AP sites, and preferentially cleaves sequences resembling L1 in vivo target sequences. Mutations in conserved amino acid residues of L1 EN abolish its nicking activity and eliminate L1 retrotransposition. We propose that L1 EN cleaves the target site for L1 insertion and primes reverse transcription.

One strategy for neutralizing retroviral infectivity is to induce the incorporation of lethal fusion proteins, such as capsid protein-nuclease fusions, into the virion during the normal viral assembly process. Genes encoding such antiviral fusion proteins must be nontoxic to the host, lethal to the virus, and must be efficiently delivered to, and expressed in, appropriate target cells.