RNA intermediate. The resulting cDNA is incorporated into the genome of host cells. In eukaryotes, the success of long terminal repeat (LTR)–containing retroelements has led to replication mechanisms that are conserved among diverse families of retrotransposons and retroviruses. Given that LTR-retrotransposons exist in yeast, we have applied powerful techniques of yeast genetics to basic questions about the function of LTR-retroelements, specifically the molecular mechanisms of reverse transcription, transport of the retrotransposon Tf1 into the nucleus, integration of Tf1 cDNA into the host genome, and the residues that contribute to the specialized steps of reverse transcription. Our study of nuclear import is motivated by the finding that mutations in a specific factor of the nuclear pore blocks import of Tf1 virus-like particles without altering the growth of the cells. Thus, it may be possible to inhibit the infection of specific retroviruses by blocking their import into the nucleus. The process of integration is significant because of its impact on the host genome. Tf1 integration serves as a model for retroviruses and retrotransposons of humans and how they choose which host sequences to disrupt.

Specific Contribution of the Primer Grip in the RNase H of Reverse Transcriptase to Removal of the Plus Strand Primer from the 5' End of the cDNA
Atwood-Moore, Khoo, Levin
The complete process of reverse transcription requires the recognition and removal of specialized primers for both the plus and minus strands. Little is known about which domains of reverse transcriptase (RT) recognize and remove the primers from the ends of the cDNA. To uncover the amino acids that mediate the individual events or reverse transcription, we mutagenized the complete sequence of Tf1 RT and inserted it into a plasmid form of Tf1 that is used to measure transposition activity in S. pombe. We isolated 3,000 strains with mutations in RT that inhibited transposition. Thirty four versions of Tf1 with single amino acid substitutions generated high levels of cDNA products. Surprisingly, one-third of the mutants produced normal levels of full-length, double-stranded cDNA, many of which clustered within a segment of the RNase H domain of RT. A recent demonstration of the importance of this region of RNase H involved a crystallographic study of HIV RT that found that the domain of interest in the RNase H, dubbed the primer grip, interacts directly with the nucleotides of the polypurine tracts (PPT). Both the position and sequence of the primer grip residues in HIV correspond well with the domain we identified in the RNase H of Tf1. We tested the possibility that the mutations resulted in defective transposition as a consequence of altered PPT processing. The change in cDNA sequences at their termini by as little as one or two nucleotides would have a drastic impact on integration catalyzed by integrase (IN). The mutations produced five-fold increases in the levels of cDNA that could not be integrated because the PPT was not removed. Thus, our screen for mutations in RT that make full-length cDNA identified the primer grip of RNase H as contributing to the removal of the PPT RNA from the 5' end of the plus strand LTR. Surprisingly, the ability to process the PPT is specific and is not required for the other steps in the pathway of reverse transcription.

Mediation of Particle Formation and Reverse Transcription in S. pombe by the Gag Protein of Tf1
Teysset, Kim, Levin; in collaboration with Dang
The Gag proteins of LTR-retrotransposons and retroviruses assemble into the shell of virus and virus-like particles, with the formation of such particles required for reverse transcription. As a result of its role in particle formation, Gag is responsible for the size of virions and the packaging of the other components of the particle such as the mRNA, RT, and IN. Although Tf1 encodes functional protease, RT, and IN (all proteins known to be necessary, respectively, for protein processing reverse transcription, and integration), the function of the Tf1 Gag protein has yet to be determined. Unraveling that function is significant because the sequence of Tf1 Gag lacks any of the conserved motifs associated with other Gag proteins.
To test whether the Gag of Tf1 was required for particle formation, packaging of mRNA, and reverse transcription, we generated four versions of Tf1 with deletions of eight amino acids in the hydrophilic domains of the Gag. Each mutation significantly reduced transposition. While most N-terminal and C-terminal deletions did not reduce levels of Gag, RT, or IN, the two central deletions did cause severe reductions in Gag. The analysis of cDNA produced by the mutant elements demonstrated that mutants with low levels of Gag were unable to accumulate cDNA. While both the N- and C-terminal mutations generated normal levels of Gag, the C-terminal mutation significantly reduced levels of cDNA. We investigated the role of Gag in packaging Tf1 mRNA by subjecting the virus-like particles to high concentrations of nuclease. Particles produced by wild type and the element with the N-terminal deletion in Gag protected the Tf1 mRNA from degradation while the other three mutations protected the Tf1 mRNA much less, which is consistent with a role of Gag in forming a protective particle. In the case of the C-terminal mutation, the lack of RNA protection by a deletion that expresses normal levels of Gag identified residues necessary for mRNA packaging.

We tested directly the role of Gag in particle formation by visualizing with electron microscopy sections of cells expressing Tf1. In cells expressing wild-type Tf1, we could occasionally detect large arrays of 30 nanometer particles, but never in cells that lacked Tf1 expression. In addition, we never observed particles in the cells with the mutations that inhibited Gag expression. The results indicate that Gag is required for particle formation. More important, Tf1 with a frameshift expressing only Gag was able to produce particles. Thus, as is the case for retroviruses, Gag of Tf1 is sufficient for the formation of particles. We conclude that the Gag of Tf1 is required for transposition and functions as the Gag of retroviruses. The central deletions showed that particles cannot form without Gag and the Tf1 mRNA cannot be packaged. The deletion near the C-terminus of Gag allowed particles to form by identified residues required for mRNA packaging. Finally, the deletion at the N-terminus of Gag did not alter protein levels or reverse transcription. Instead, the mutation likely reduced transposition because it removed the nuclear localizing signal (NLS) that is required for import of Tf1 protein and cDNA into the nucleus.

Sequences of the S. pombe Genome Selected for Integration of Tf1
Bowen, Levin; in collaboration with Singleton, Wood
Recognizing that the process of integration has the potential to disrupt genes of the host, we sought to understand the balance between the ability of the transposon to insert into the host genome versus the efforts of the host to maintain its viability. Much of what is known about the insertion sites of LTR-retrotransposons indicates that the process of integration is specifically controlled to avoid the disruption of host genes. In fact, all five transposons of Saccharomyces cerevisiae select sites for integration that lack coding sequences. Much less is known about the interactions between the genome of S. pombe and its transposons. The recent completion of the genome sequence of S. pombe allowed us to study the full set of transposon sequences and their relationship to the host genes.

We conducted a comprehensive search of the genome sequence of S. pombe for transposon-related sequences. Surprisingly, the only transposons revealed by the search were related to the Tf1/Tf2 family of LTR retrotransposons. We identified no complete copies of Tf1 and only 13 full-length copies of Tf2. Single LTRs result from the removal of full-length elements by homologous recombination. There were 182 single LTRs in the genome and those with the full-length elements constitute 0.8 percent of the S. pombe genome. The single LTRs provide important information about the transposon history of the host genome. Examination of all the LTRs revealed that each element was located within intergenic regions of sequence. Surprisingly, 96 percent of the insertions were in intergenic regions that included pol II promoters, thereby accounting for a significant bias in that pol II promoters are found in only 53 percent of the intergenics. We confirmed the association between pol II promoters and insertion sites by compiling distances between LTR insertions and the nearest ORF. Eighty-four percent of the insertions were closer to a 5' end of an ORF than a 3' end. Moreover, most of the insertions clustered within 200 nucleotides of the start codon.

Role of Interaction between the Chromo-domain of Tf1 IN and Histones with a Modification in Insertion of Tf1 cDNA Adjacent to pol II Promoters
Nam, Levin; in collaboration with Grewal
Proteins with conserved motifs dubbed chromodomains are often associated with heterochromatin. Recently, the chromodomain-containing proteins HP1, suVAR(3-9), Clr4p, and Swi6p were shown to generate heterochromatin by binding to specific nucleosomes on DNA. The binding results from an interaction between the chromodomain and histone H3, which is methylated specifically at lysine 9. Interestingly, Malik and Eickbush discovered a sequence at the C-terminus of the Tf1 IN that was similar to a chromodomain. We made mutations in the conserved residues of the putative chromodomain and found that they greatly reduced the transposition activity of Tf1. Experiments conducted by the laboratory of Shiv Grewal demonstrated that, in the S. pombe intergenic regions of pol II genes, histone H3 was not methylated at lysine 9 but instead had a high level of methylation at lysine 4, with peaks of lysine 4 methylation near pol II promoters, suggesting that the chromodomain of Tf1 IN may target insertion into intergenic sequences by associating with histone H3 methylated at lysine 4. In collaboration with this laboratory, we tested this model by measuring transposition activity in a strain that was deleted for set1, the methylransferase responsible for the methylation of H3 at lysine 4. The deletion rendered methylation of lysine 4 undetectable. More important, the transposition activity of the strain was significantly reduced compared with strains with the wild-type set1.

Tf1 Residues in Gag that Regulate Nuclear Localization
Kim, Levin
As in the case for HIV, Tf1 must translocate into the nucleus the proteins and cDNA of its preintegration complex (PIC) before it integrates the cDNA into the host genome. In previous experiments, we identified a component of the nuclear pore that is required for import of Tf1 into the nucleus. The function of the protein Nup124p shows specificity in that its absence disrupts import of Tf1 but does not reduce import of other proteins. Interestingly, the nuclear localization of Gag occurs only when cultures reach stationary phase.

We tested whether residues adjacent to the NLS of Gag regulate the NLS activity. We made five mutant transposons, each with sequential tracts of four amino acids downstream of the NLS replaced with stretches of four alanines. All five versions of Tf1 transposed significantly less than the wild type, but all five mutants made normal amounts of Gag. We showed that two of the mutants (position IV and V) did not complete reverse transcription, indicating that residues in the Nterminus of Gag contributed to reverse transcription. Deletions at positions I, II, and III did not reduce levels of reverse transcription, and the defects in transposition caused by deletions I, II, and III were not caused by changes in protein levels or reduced reverse transcription. Thus, we investigated the possibility that the mutations altered the function of the NLS. In four of the five alanine mutants (mutant positions I through IV), induced for transposition, localization of Gag in the nucleus was significantly reduced, indicating that the residues deleted at positions I, II, and III do contribute to the function of the NLS. Surprisingly, the mutation at position V caused a defect in the regulation of import that allowed Gag to localize in the nucleus even during log phase growth. One explanation is that the mutation V disrupts the Gag-Gag interactions necessary for particle formation, which, in turn, could allow import of a nonparticle form of Gag by a process different from the stationary phase mechanism. Alternatively, deletion V could remove a site of post-translational modification that inhibits import.

Role of Tf1 t-SNARE Syntaxin in Transport into the Nucleus
Nam, Peters, Kim, Levin
The life cycle of Tf1 is very similar to that of retroviruses and has therefore been used as a model system for studying the mechanism of retrovirus propagation. To understand the host functions required for the propagation of LTR elements, we mutagenized a culture of S. pombe and screened for mutants that were severely defective for Tf1 transposition. Previous characterization of the strains revealed that mutations in the nuclear pore factor Nup124p and the histone deacetylation factor Sin3p decreased transposition frequencies ten-fold, a reduction that is attributable to defects in the transport of Tf1 into the nucleus. We have now identified a mutation in a newly identified host gene, psy1, that reduces Tf1 transposition more than 40-fold. The gene encodes a t-SNARE protein with 65 percent sequence similarity to the homologs of syntaxin in S. cerevisiae. Syntaxin proteins reside in membrane structures and promote vesicle trafficking by mediating the fusions of vesicles with target membranes. It is therefore likely that Psy1p mediates some form of vesicle traffic in S. pombe. Recent work by Kanamura and Shimoda demonstrated that Psy1p in S. pombe is a plasma membrane t-SNARE.

We investigated the role of Psy1p in transposition by testing the effect of the psy1-1 mutation on the intermediates of Tf1. The mutation did not reduce the levels of production of either Tf1 proteins or cDNA. The data indicate that the defect in transposition occurs after reverse transcription. We tested whether the Tf1 cDNA produced by the mutant strain was transported into the nucleus. The psy1-1 allele caused a dramatic reduction in the homologous recombination that occurred between Tf1 cDNA and a plasmid copy of Tf1, suggesting that, although cDNA is produced, it is not transported into the nucleus. To test whether the mutation in psy1-1 altered the import of Tf1 protein, we determined the cellular localization of Gag. While Gag localized in the nucleus of wild-type cells, the psy1-1 mutation caused Gag to remain in the cytoplasm, indicating that a late step in transposition, potentially transport to the nuclear envelope, requires a specific class of vesicle traffic.