Ty3 requires yeast La homologous protein for wild‐type frequencies of transposition

The Saccharomyces cerevisiae retrovirus‐like element Ty3 inserts specifically into the initiation sites of genes transcribed by RNA polymerase III (pol III). A strain with a disruption of LHP1, which encodes the homologue of autoantigen La protein, was recovered in a screen for mutants defective in Ty3 transposition. Transposition into a target composed of divergent tRNA genes was decreased eightfold. In lhp1 mutants, Ty3 polyproteins were produced at wild‐type levels, assembled into virus‐like particles (VLPs) and processed efficiently. The amount of cDNA associated with these particles was about half the amount in a wild‐type control at early times, but approached the wild‐type level after 48 h of induction. Ty3 integration was examined at two genomic tRNA gene families and two plasmid‐borne tRNA promoters. Integration was significantly decreased at one of the tRNA gene families, but was only slightly decreased at the second tRNA gene family. These findings suggest that Lhp1p contributes to Ty3 cDNA synthesis, but might also act at a target‐specific step, such as integration.


Introduction
The current study was undertaken in order to identify host factors that participate in the life cycles of yeast retrotransposons. Ty3 is one of five retrotransposons found in Saccharomyces cerevisiae (reviewed by Boeke and Stoye, 1997;Sandmeyer et al ., 2001). Ty3 is a 5.4kbp DNA sequence composed of an internal domain flanked by long-terminal repeats (LTRs) of 340bp (Clark et al ., 1988). The element encodes Gag3p and Gag3-Pol3p polyproteins that are processed into mature proteins by the Ty3 protease (PR) (Hansen et al ., 1992;Kirchner and Sandmeyer, 1993). Gag3p is processed into major structural proteins, capsid (CA) and nucleocapsid (NC). POL3 is expressed as a Gag3p-Pol3p fusion dependent upon a + 1 frameshift at the end of GAG3 (Kirchner et al ., 1993). This polyprotein is processed into structural as well as catalytic proteins, PR, reverse transcriptase (RT) and integrase (IN) (Hansen et al ., 1992). Ty3 RT is responsible for reverse transcription of RNA into double-stranded DNA in a process facilitated by NC (Orlinsky and Sandmeyer, 1994;Gabus et al ., 1998), and IN mediates integration of DNA into the host genome (Kirchner and Sandmeyer, 1996). Ty3 integrates specifically at the transcription initiation sites of genes transcribed by RNA polymerase (pol) III (Chalker and Sandmeyer, 1990;, and pol III transcription factors have been implicated in targeting by in vitro and in vivo studies (Chalker and Sandmeyer, 1993;Kirchner et al ., 1995;Yieh et al ., 2000;Aye, 2001).
Because of limited coding capacity, retroelements rely on host cells for functions including RNA processing, nuclear import and export and translation (Coffin, 1996;Sandmeyer and Menees, 1996;Boeke and Stoye, 1997). In addition, several RNA species play quite specific roles in the retroelement life cycle, including the tRNA primer for reverse transcription and tRNA species required for frameshifting and, potentially, small RNAs involved in splicing. In the case of Ty3, tRNA iMet was shown to function as the primer for reverse transcription (Keeney et al ., 1995). Other tRNA species control the frequency of Ty3 + 1 frameshifting required for Gag3-Pol3p translation (Farabaugh et al ., 1993;Sundararajan et al ., 1999).
We identified a mutation in LHP1 in a screen for genes that affect transposition of Ty3. LHP1 encodes the yeast La homologous protein (Hendrick et al ., 1981;Yoo and Wolin, 1994), one of a number of proteins involved in the synthesis and processing of small structural RNAs. Mammalian La was first identified as a component of an autoantigen, Ro RNP, and was shown to have RNA-binding activity (Hendrick et al ., 1981;Stefano, 1984). It was initially implicated in termination of pol III transcription (Gottlieb and Steitz, 1989). Subsequently, the mammalian protein was reported to facilitate pol III transcript release and reinitiation (Maraia et al ., 1994;Maraia, 1996), although it is not absolutely required in vitro for pol III reinitiation (Kassavetis et al ., 1990;Weser et al ., 2000). Lhp1p is not essential. The role of Lhp1p as an RNA chaperone in yeast was uncovered beginning with the demonstration that a mutation in LHP1 is synthetically lethal with a mutation in the anticodon stem that destabi-lizes the structure of an essential tRNA Ser CGA encoded by a single gene (Yoo and Wolin, 1997). Since that time, Lhp1 has been implicated in a variety of RNA chaperone activities (reviewed by Wolin and Cedervall, 2002). Lhp1p was shown to stabilize pretRNAs for the endonucleolytic cleavage leading to 3 ¢ end maturation (Yoo and Wolin, 1997) and to co-operate with Gcd14p in the maturation of some tRNA precursors (Calvo et al ., 1999). In addition to binding pretRNAs, Lhp1p binds to processing intermediates of other small structural RNAs: certain U snRNP RNAs, including the pol III-transcribed U6 (Pannone et al ., 1998;2001), and pol II-transcribed U1, U2, U4 and U5 RNAs (Xue et al ., 2000); snoRNP complex RNAs (Kufel et al ., 2000); and Ro RNP complex RNAs (Pruijn et al ., 1990). In addition to its role in binding structural RNAs, a variety of other RNA-related functions have been ascribed to La. Mammalian La binds in the vicinity of the translation initiation codon of some RNAs (McBratney and Sarnow, 1996). It has been documented to promote internal translation initiation of polio virus (Meerovitch et al ., 1993;Svitkin et al ., 1994a) and hepatitis C virus (Ali et al ., 2000) RNAs. In addition, it reverses interference by the TAR element with translation of human immunodeficiency virus (HIV)-1 RNA (Svitkin et al ., 1994b). La interacts with the 5 ¢ -terminal oligopyrimidine sequence found in a collection of growth-related genes, the so-called TOP genes (Crosio et al ., 2000), and appears to enhance their translation. La also stabilizes hepatitis B virus against RNase degradation (Heise et al ., 2001).
In this study, we found that disruption of LHP1 decreased Ty3 transposition into a plasmid-borne, divergent tRNA gene target and differentially affected transposition at two chromosomal tRNA gene families. In addition, Ty3 cDNA accumulated at a slower rate in the mutant strain than in the wild-type strain. Based on these observations, we propose that Lhp1 may have two functions in Ty3 transposition: one in cDNA production and one at the target site.

Results
A genetic screen was performed in order to identify host factors that show genetic interactions with Ty3. Briefly, a library of yeast DNA fragments previously mutagenized with a mini-Tn 3 element marked with bacterial lacZ and yeast LEU2 was transformed into yeast strain YPH500 (Table1), and transformants were selected (Burns et al ., 1994;Aye, 2001). The high frequency of homologous recombination in yeast ensured that most of the transformants represented cells in which a disrupted copy of a gene has replaced the wild-type locus. A large number of isolates were screened for Ty3 transposition phenotype. We report here the characterization of one mutant. A less detailed description of Ty3 phenotypes of the larger set of mutants will be reported in a separate publication (M. Aye and S. B. Sandmeyer, manuscript in preparation).
A genetic assay in which Ty3 position-specific transposition between two divergently transcribed tRNA genes activates a suppressor tRNA gene ( sup2bo ) (Kinsey and Sandmeyer, 1995) was used in a trial screen of 2000 Leu + transformants for Ty3 transposition mutants. Leu + mutant strains and wild-type strain yMA1235, a LEU2 version of the parental strain YPH500, containing a galactose-inducible Ty3 element on low-copy plasmid pTM843 (Table2) and a tDNA target, on plasmid pPK689, were picked and arrayed in patches onto SC[glu] lacking histidine, tryptophan and leucine. Cultures were replicated to SC [gal] lacking histidine, tryptophan and leucine in order to induce Ty3 expression and select for cells retaining the plasmids. In pPK689, tDNA Val , which functions as the Ty3 target, is positioned so that it interferes with the expression of a neighbouring, divergently transcribed ochre suppressor tDNA Tyr , sup2bo . S up2bo expression is further attenuated by a tract of pyrimidines on the non-transcribed strand in the region of transcription initiation. Ty3 position-specific integration at the tDNA Val target alleviates interference between the divergent genes and changes the sequence composition upstream of sup2bo , thereby activating suppressor expression. YPH500 contains a nonsense allele, ade2o , which is suppressed by expression of sup2bo . Patches of cells induced to undergo transposition were replicated to minimal medium supplemented with uracil and lysine. On this medium, cells that have undergone transposition and activated suppressor expression are observed as papillae (data not shown).

Truncation of Lhp1p reduces Ty3 transposition
Mutant 22-19 exhibited significantly less transposition than wild-type cells in the patch assay (data not shown). In order to identify the gene disrupted by insertional mutagenesis, the strain was transformed with YIp5 plasmid linearized by digestion with Pvu I. Homologous recombination between beta-lactamase gene sequences in this vector and in the mTn 3 insertion resulted in the introduction of an Escherichia coli origin of replication, allowing the region of DNA containing the insertion to be recovered by transformation. The sequence of genomic DNA at the site of the disruption was determined as described in Experimental procedures . This sequence was used in a BLASTN search of the Saccharomyces Genome Database (http://genome-http://www.stanford.edu/Saccharomyces). The mTn 3 insertion in mutant 22-19 was found to be in the middle of LHP1 (Fig.1A), the gene encoding an orthologue of human La autoantigen (Yoo and Wolin, 1994). This mutant was designated YMA1237. Sequence analysis showed that mTn 3 was inserted 447bp downstream of the LHP1 initiation codon. Thus, Lhp1p produced in the mutant strain would lack the carboxyl-terminal half of the protein, including part of the RNA recognition motif and the putative nuclear localization signal (Yoo and Wolin, 1994;Rosenblum et al., 1998) (Fig.1A). Southern blot analysis using probes against LHP1 and lacZ indicated that this was the sole insertion of mTn3 in the mutant strain ( Fig.1B; data not shown). A quantitative version of the suppressor activation assay showed that the transposition frequency in YMA1237 was ª12% of that in wildtype cells ( Fig.1C and D). Because the transformation process itself is mutagenic, two methods were used to test whether the LHP1 locus disruption corresponded to the mutation causing the defect in Ty3 transposition. First, the linearized, rescued plasmid was used to disrupt the LHP1 locus in YPH500 and YMA1322, and transposition assays were performed. Each reconstructed lhp1 mutant had patterns in Southern blot analysis indistinguishable from the predicted pattern and the pattern of the original mutant ( Fig.1B) and had a defect in transposition comparable to the original mutant ( Fig.1C and D; data not shown). Secondly, wild-type LHP1 was tested for its ability to reverse the Ty3 transposition phenotype. Overexpression of LHP1 under a galactoseinducible promoter on a high-copy plasmid, pMA1708, elevated the frequency of transposition of the lhp1 mutant relative to that of wild-type cells carrying vector alone. Wild-type cells overexpressing LHP1 also had higher frequencies of Ty3 transposition than controls ( Fig.1C and D). To investigate whether differences in transposition frequency correlated with different levels of Lhp1p in these strains, immunoblot analysis was performed on whole-cell extracts (WCE) with Lhp1 antisera (a gift from C. Yoo, Yale University). Lhp1p was readily detectable in the wild type, but not in the mutant strain with vector plasmid (Fig.1E). In addition, Lhp1p expression was considerably higher in wild-type and mutant strains in which LHP1 carried on pMA1708 was induced. Thus, expression of Lhp1p correlated well with Ty3 transposition frequency.
Because the divergent tRNA gene target is an artificial target that is selective for position specificity and orientation (Kinsey and Sandmeyer, 1995;Aye et al., 2001), it was also of interest to monitor Ty3 transposition into chromosomal tRNA gene targets. Cells carrying pTM843 were induced for Ty3 expression for 6h, a time at which transposition is still increasing in the wild-type background (data not shown). Chromosomal insertions were monitored in a  Hill (1986) polymerase chain reaction (PCR) assay in which one primer annealed to the Ty3 IN sequence upstream of the 3¢ LTR, and a second primer annealed to a conserved sequence common to 10 tDNA Gln or 14 tDNA Val genes ( Fig.2A; Table3). These tDNA families were chosen for the integration assay based on the observation that they had relatively few upstream Ty3 LTRs and no Ty3 insertions in the strain for which genomic sequence was determined. They performed better than two other such families in trial assays (data not shown). The amplified Ty3-tDNA junction fragment was visualized using a radioactive, Ty3 LTRspecific probe (Fig.2B). The amount of product produced in reactions templated on DNA from cells expressing a Ty3 IN catalytic site mutant on plasmid pMA1890 was used to estimate background in the assay. The background observed presumably arises primarily from annealing of nascent DNAs primed from endogenous tRNA gene family members with upstream Ty3 LTRs and nascent DNAs primed from the Ty3 cDNA ( Fig.2B, lanes 2 and 6). Integration of Ty3 upstream of chromosomal copies of tDNA Gln was modestly reduced (Fig.2B, compare lanes 3 and 4). In this assay, the control tDNA Gln showed significant amounts of background product (Fig.2B, lane 2). Background was less in the case of tDNA Val , and insertions associated with tDNA Val sequences were significantly reduced in the mutant strain (Fig.2B, compare lanes 7 and The location of the mTn3 insertion is indicated by an open triangle. The numbers below the boxes represent amino acid residues as described previously (Yoo and Wolin, 1994;Rosenblum et al., 1998). The La domain was determined based on similarity between the yeast and human proteins. B. LHP1 disruption in the mutant strain. Genomic DNA from the wild type (wt) or the mutant YMA1237 and the two reconstructed strains (lhp1) were cleaved with EcoRI, and Southern blot analysis was performed with a probe specific for LHP1. Lambda DNA cleaved with HindIII was used as the size marker (M). C. Overexpression of LHP1 elevates Ty3 transposition in mutant and wild-type cells. Cells were plated at a density of 10 6 cells per plate and replica plated to final selective medium as described in Experimental procedures. Growth on medium selecting for cells with Ty3 transposed into the target plasmid for wild-type (wt) and mutant (lhp1) strains transformed with either the control vector (pYES2.0) or expression plasmid (pMA1708). Plating was in triplicate, and examples are shown. D. Quantification of transposition. The average number of transposition events from triplicate plates of the type shown in (C) was determined, and the average and standard deviations are shown. The level of transposition is expressed as a percentage of the frequency in the wildtype strain transformed with vector pYES2.0. E. Expression of Lhp1p in wild-type and mutant strains. Immunoblot analysis was performed with anti-Lhp1p antibodies on -extracts from yeast strains tested for transposition. LHP1 plasmid and vector are as described above. 8). Input DNA from different cultures used to template amplification of the RAD52 locus resulted in comparable amounts of product from each sample (Fig.2B, bottom). In order to quantify the difference observed in the case of the tDNA Val gene family, amounts of genomic DNA from 1 to 25ng were used to template the reaction, and the results were evaluated using QUANTITY ONE software (Bio-Rad). Analysis showed that, within this range, the amount of Ty3-tRNA Val gene or RAD52 control product was roughly proportional to input DNA for wild type and lhp1 mutant (Fig.2C). Within this range, using 5ng of genomic DNA collected from cells induced for 6h, the amount of Ty3-tDNA Val product generated from wild-type DNA, normalized to RAD52 product, was more than fourfold the amount generated from the lhp1 mutant DNA (Fig.2C, compare lanes 4 and 7). PCR analysis of DNA from wild-type and mutant cultures collected after different times of induction showed that integration into tDNA Val genes occurred more frequently in the wild type than in the lhp1 strain (Fig.2D, compare lanes 2-5 and lanes 6-9) over a period ranging from 6h to 24h. The difference was most dramatic after 6h of induction (  A. Strategy for PCR detection of Ty3 integration into a target. Primers are specific to Ty3 (278) and to the tDNA sequences. After insertion of Ty3, a diagnostic junction fragment can be amplified in PCR from the two primers. B. Ty3 transposition into chromosomal targets. Wild-type and lhp1 mutant strains transformed with Ty3 plasmid pTM843 were grown in a synthetic medium containing galactose for induction of Ty3 expression. Total DNA was extracted from cultures, and 25ng of each was used as the template for PCR with primers specific for Ty3 and glutamine (tDNA Gln ) (lanes 1-4) or valine (tDNA Val ) tRNA gene family (lanes 5-8) genes. PCR products were subjected to Southern blot analysis with a probe specific for the Ty3 LTR (top). A Ty3 element with mutations in catalytic residues of IN (IN*) expressed from plasmid pMA1890 was used as a negative control (lanes 2 and 6). Positive control PCR (control; lane 9) was performed with pTM842, pDLC374 with a Ty3 insertion adjacent to the tDNA target (Menees and Sandmeyer, 1994 and 24h of induction were subjected to the same analysis as in (C). Times of induction for the wild-type (lanes 2-5) and lhp1 (lanes 6-9) samples are indicated. The RAD52 locus was examined as the control PCR for the input DNA (bottom). Nevertheless, in the genetic background used for the yeast knock-out collection (BY4741; Research Genetics), this assay yielded similar differences between wild type and a lhp1 null mutant (lhp1D; data not shown). Thus, the Ty3 transposition defect observed for the synthetic target extends to at least one family of chromosomal targets.

Effect of LHP1 truncation on cellular growth and pol III transcription
The effect of LHP1 truncation on growth was tested in rich and synthetic dropout medium and on medium containing glucose, galactose or raffinose as the carbon source. These experiments, as well as growth curves (data not shown), indicated that the mutant has no significant growth defect. In addition, the lhp1 strain was not temperature sensitive as it grew as well as the wild-type strain at 37∞C (data not shown). Because Lhp1p is implicated in stabilization and endonucleolytic processing of nascent pol III transcripts (Yoo and Wolin, 1997), the possibility of gross alterations in tRNA transcripts resulting in effects on Ty3 transposition was evaluated. The amount of tRNA iMet , the Ty3 reverse transcription primer (Keeney et al., 1995) was examined first. Total RNA was prepared from the mutant and the wild-type cells grown to between A 600 =0.3 and 0.4 and induced with the addition of galactose to a final concentration of 2% for 6h. Northern blot analysis was performed with an oligonucleotide probe specific for tRNA iMet (Fig.3A). The blot was scanned, and the bands were quantified using IMAGEQUANT software. The amount of tRNA iMet in the lhp1 cultures was ª95% of that in the wild type when normalized to actin mRNA in the same preparations (Fig.4A). Thus, it seemed unlikely that differences in tRNA iMet could completely account for the differences observed in transposition. Because Ty3 inserts specifically at the transcription initiation site of genes transcribed by pol III and because mammalian La protein has been implicated in transcription reinitiation, it was possible that the lhp1 mutant was sloppy in the transcription initiation site selection, which in turn reduced the apparent efficiency of Ty3 insertion as measured by assays demanding position specificity. To investigate transcription initiation site selection by pol III in the mutant strain, primer extension analysis was performed on the RNA samples prepared from the cells carrying a plasmid with SUP2b tRNA gene (pDLC356) as described in Experimental procedures. This tRNA gene contains an intron with a unique sequence so that pre-tRNAs are preferentially detected by primer extension. Identical sizes of the primer extension products from wildtype and mutant cells indicated that the major transcription initiation site was not altered in the mutant strain (Fig.3B).

Effect of LHP1 mutation on Ty3 expression, frameshifting and proteolytic processing
Because La has been implicated in the stability of RNAs A. Northern blot analysis of tRNA iMet . Equivalent amounts of total RNA were fractionated and transferred as described in Experimental procedures. An oligonucleotide specific for mature tRNA iMet was used as the probe to perform Northern blot analysis of total RNA samples from the wild-type (wt) and mutant (lhp1) strains grown with medium containing either raffinose (raff) or galactose (gal) as the carbon source. Samples in lanes 2 and 4 were from cells induced at slightly lower densities than those in lanes 3 and 5. Samples are the same as those shown in Fig.4. B. Detection of pretRNA Tyr by primer extension. Twenty micrograms of total yeast RNA was used as the template, and 32 P-labelled SUP2b intron-specific oligonucleotide as the primer for each extension reaction. A DNA sequencing ladder was generated from a reaction templated by the SUP2b tDNA in pDLC356 using the same oligonucleotide. Plasmid pDLC315, containing a SUP2 gene with a C56G mutation in the boxB promoter element (G56), was used as a negative control. A portion of SUP2b tDNA (open box) with a boxA element (shaded box) and upstream sequences are shown to the left of the autoradiograph. Transcription initiation site, primer extension product from pretRNA and the 5¢ end-labelled, free oligonucleotide are indicated by arrows. and in translation efficiency, whether the amount of a specific Ty3 intermediate is decreased by the lhp1 mutation was investigated. Ty3 RNA, protein and cDNA were measured in wild-type and mutant strains. The wild-type and mutant strains carrying pTM843, with Ty3 under the GAL1 UAS , were grown in SC[raff] lacking tryptophan to a density of A 600 of 0.3 and 0.4. Galactose was added to these cultures to a final concentration of 2% to induce Ty3 expression. After 6h of induction, cells were harvested and proteins and nucleic acids extracted. Northern blot analysis showed that the amount of Ty3 transcripts did not differ significantly between the mutant and wild-type strains (Fig.4A, compare lanes 2 and 3 with lanes 4 and 5 respectively).
Ty3 encodes structural and catalytic proteins in GAG3 and POL3 ORFs respectively. Catalytic proteins are processed from a Gag3-Pol3p fusion protein precursor, the synthesis of which depends upon a +1 frameshift at the end of the GAG3 reading frame, which occurs at a frequency of about 10% (Kirchner et al., 1992;Farabaugh et al., 1993). Not surprisingly, disruption of the ratio of Gag3p to Gag3-Pol3p interferes with Ty3 virus-like particle (VLP) formation (Kirchner et al., 1992). In Ty3 frameshifting, low abundance of a critical, charged tRNA Ser GCU species creates a 'hungry codon' at the frameshift site, and wobble basepairing of a near cognate peptidyl-tRNA Ala IGC to the mRNA is proposed to allow +1 frameshifting of the ribosomes (Sundararajan et al., 1999). In order to monitor Ty3 frameshifting, two lacZ reporter plasmids, in which the coding region of lacZ downstream of the Ty3 frameshift sequence was in frame with or in +1 reading frame relative to the upstream HIS4 ORF (a generous gift from P. Farabaugh, University of Maryland), were used to measure b-galactosidase activity in wild-type and lhp1 cells. Comparison of the activity of these two constructs indicated that Ty3 frameshifting occurred slightly more frequently in the lhp1 mutant than in the wild-type cells (data not shown). Thus, it is possible that the loss of Lhp1 resulted in an increase in frameshifting. If the change in frameshifting was responsible for the disruption of Ty3 transposition, then the ratio of Gag3p-to Gag3-Pol3pderived proteins or VLP formation and processing would be expected to be affected. Immunoblot analysis with antibodies against Ty3 CA and IN, encoded in GAG3 and POL3, respectively, indicated that the level of Ty3 proteins in WCE and VLPs was not significantly altered by the mutation or overexpression of LHP1 (Fig.4B). Immunoblots of dilutions of extracts of cells expressing Ty3 confirmed that a twofold difference in CA or IN could have been detected and was not detected between extracts of wild-type and lhp1 cells (data not shown). We conclude that, in spite of potential differences in frameshifting, production and processing of proteins was not significantly affected by the lhp1 mutation.
Ty3 cDNA is reduced in the lhp1 mutant Subsequent to particle assembly and protein maturation, Fig. 4. Truncation of LHP1 has no detectable effect on Ty3 expression and VLP protein production. A. Northern blot analysis of Ty3 RNA. Total RNA samples from wildtype (wt) and mutant (lhp1) strains grown in medium containing raffinose (raff) or galactose (gal) were subjected to Northern blot analysis with a 32 P-labelled, internal BglII fragment of Ty3 (top) and a 32 Plabelled fragment containing the ACT1 gene (bottom). Samples from two independent inductions, one at an A 600 of 0.3 (lanes 2 and 4) and another at 0.4 (lanes 3 and 5), are shown. B. Immunoblot analysis of Ty3 proteins. Wild-type (wt) and mutant (lhp1) strains were transformed with control plasmid (pYES2.0) or LHP1 expression plasmid (pMA1708) in addition to pTM843 and grown in medium with galactose as the carbon source for preparation of extracts. VLPs were prepared as described in Experimental procedures. Immunoblots with antibodies to Ty3 CA or IN proteins were performed on 20mg each of WCE (left) and 1mg each of wild-type and lhp1 Ty3 VLPs (right). Antibodies to CA recognize mature CA protein (26kDa) and precursor Gag3p (38kDa). Antibodies to IN recognize mature IN species of 61 and 58kDa in VLP preparations. The slightly lower amount of IN in the lhp1 WCE sample observed in this experiment was not typical of the majority of experiments.
Ty3 DNA is reverse transcribed from the genomic RNA template. To investigate the effect of Lhp1p truncation on reverse transcription, expression of Ty3 was induced in the wild-type and the mutant strains, and samples were taken at different time points. Total DNA extracted from these samples was subjected to Southern blot analysis with a Ty3-specific probe (Fig.5A). The results of this experiment were measured and quantified using a phosphorimager and IMAGEQUANT software (Molecular Dynamics). This analysis showed that Ty3 cDNA, normalized to Ty3 plasmid DNA, accumulated faster in the cells where full-length Lhp1p was present. Between 12 and 36h of induction, the level of Ty3 cDNA in the mutant strain was about half the wild-type level, but continued to accumulate and approached the wild-type level by 48h, when the accumu-lation of Ty3 cDNA had reached a steady state or declined in the cells with normal or higher levels of Lhp1p (Fig.5B). Moreover, the level of cDNA accumulation in different strains showed a direct correlation with the level of Lhp1p (Figs5B and 1E), suggesting that, even in normal cells, Lhp1p may be limiting the accumulation of cDNA.
Mobilization of Ty3 was also measured using the Ty3-mhis3AI assay (Sadeghi et al., 2001) originally developed for Ty1 (Curcio and Garfinkel, 1991). Expression of Ty3 from this construct yields Ty3 RNA fused to the antisense RNA of HIS3 interrupted by an artificial intron. Intron splicing and subsequent reverse transcription produces Ty3-HIS3 cDNA, which can be integrated or recombined into the genome or the plasmid. Although Lhp1 is an RNA chaperone, its loss does not affect splicing (Xue et al., 2000). The frequency of His + prototrophy reflects recombination as well as Ty3 integration. However, in strains containing endogenous elements, the frequency of His + prototrophy appears primarily to reflect recombination (Sadeghi et al., 2001;unpublished results). In this context, the assay serves as a reporter for the delivery of Ty3 cDNA to the nucleus. The Ty3-mhis3AI assay was used to measure mobilization of a wild-type element compared with an IN mutant element in the wild-type and lhp1D background (Table4). Mobilization of the wild type and IN mutant Ty3 occurred in the wild type with a frequency of 2.47¥10 -4 and 2.45¥10 -4 per cell respectively. The wildtype Ty3 was mobilized at 1.22¥10 -4 per cell in the lhp1 background. In neither the wild-type nor the lhp1D background was the result different for a catalytic site mutant of the Ty3 IN protein. No His + cells were recovered in the absence of galactose induction of Ty3 expression in either strain (data not shown). Thus, using the Ty3-mhis3AI assay, Ty3 recombination occurred with about half the frequency in the lhp1D strain as it did in the wild-type strain. The accumulation of Ty3 cDNA showed a similar decrease in the lhp1D strain background as in the YPH500 background (data not shown). Thus, Ty3 cDNA and the recombination of that cDNA showed similar reductions in the lhp1D mutant compared with wild type.
If Lhp1p affects the amount of Ty3 reverse transcripts, it might function in association with the VLP where reverse transcription occurs. Immunoblot analysis with anti-Lhp1p A. Truncation of LHP1 reduces the level of Ty3 reverse transcripts (cDNA). Wild-type (wt) and the mutant (lhp1) strains transformed with pTM843 and control plasmid, pYES2.0, or LHP1 expression plasmid pMA1708 were grown in SC[raff] lacking tryptophan and uracil (0h). After the addition of galactose, samples were taken at the indicated times. Yeast DNA was digested with EcoRI, and Southern blot analysis was performed with a 32 P-labelled, Ty3-specific probe, which hybridizes to full-length 5.4kb cDNA, as well as to Ty3 donor plasmid (pTM843) and chromosomal Ty3 elements (unmarked bands). B. The Southern blots shown in (A) were quantified in a phosphorimager and analysed with IMAGEQUANT software (Molecular Dynamics). Ty3 cDNA level, normalized to pTM843 from the same sample and expressed in arbitrary units, was plotted against the time of Ty3 induction. Cells transformed with vector or pMA1708 are indicated by (V) or (LHP1) respectively. antibodies was performed in order to test for the presence of Lhp1p in the VLP fraction. A small amount of Lhp1p was detected in the wild-type VLPs (Fig.6A, right). However, it is possible that this Lhp1 is not physically associated with the VLPs.

Mutant extracts support in vitro integration of Ty3
The cDNA defect is less than the defect observed in the genetic assay using the divergent tRNA gene target, sug-gesting that Ty3 could be affected at an additional step, subsequent to cDNA synthesis in the lhp1 mutant. An in vitro assay was used to examine integration. The in vitro reaction requires, in addition to target tDNA and VLPs, TFIIIB and TFIIIC (Kirchner et al., 1995). Pol III competes with VLPs for access to the target (Connolly and Sandmeyer, 1997). Immunoblot analysis was performed to determine whether Lhp1p was present in the transcription extract fractions used in these assays. Lhp1p was detected in the S100, BR500 and TFIIIB fractions (Fig.6A,   Fig. 6. In vitro integration assays using extracts and VLPs from wild-type and lhp1 mutant strains. A. Immunoblots with anti-Lhp1p antisera were performed on S100, BR500 extracts and subfractions of BR500 (TFIIIB, TFIIIC and pol III) (left) and on VLPs (right left). Because the TFIIIB fraction is required for tDNA integration, the possibility that Lhp1p might play a role in position-specific integration of Ty3 was explored further. Transcription extracts and VLPs prepared from the wildtype and lhp1 mutant strains were compared for the ability to mediate integration into a plasmid-borne tRNA gene in vitro. Integration was detected at the tRNA gene using PCR and primers in the Ty3 and tRNA gene target (Kirchner et al., 1995). Integration activity was similar for lhp1 VLPs tested with either wild-type or lhp1 extract (Fig.6B, compare lanes 5 and 10). No difference was observed if wild-type VLPs were used in place of lhp1 VLPs in these reactions (data not shown). The effect of incubation of BR500 or VLPs with rLhp1p was also examined. Within the concentration range of rLhp1p that altered the pretRNA processing of transcripts produced from plasmid-borne tRNA gene templates (data not shown), no significant stimulation of Ty3 integration was observed in reactions with mutant or wild-type extracts (Fig.6B, lanes 5-8, 10 and 11). A small increase detected at a higher concentration, 50ng of rLhp1p, in the experiment shown (Fig.6B, lane 9) was not observed consistently. In vivo transposition of Ty3 into the target plasmid used for the in vitro assays was also examined as described previously (Menees and Sandmeyer, 1994). PCR detection of the Ty3 LTR-tDNA junction (Fig.6C) showed that integration in the lhp1 strain (Fig.6C, lane 5-7) is not dramatically decreased compared with the wild-type strain (Fig.6C, lane 2-4). Although these experiments do not exclude the possibility that Lhp1p is limiting in vivo for some chromosomal targets, we conclude that, in vitro, Lhp1p is not essential for Ty3 integration.

Discussion
In this study, we found that a disruption of the LHP1 gene reduced Ty3 transposition into synthetic and chromosomal tRNA targets. Investigation of Ty3 life cycle intermediates showed that amounts of the Ty3 RNA and mature VLP proteins were not significantly affected by LHP1 disruption. The disruption was associated with a twofold decrease in the amount of Ty3 cDNA during the first 24h of induction. Greater decreases were observed for integration into a synthetic tRNA gene target plasmid and into one of two chromosomal tRNA gene targets. These results are consistent with two models: (i) that cDNA, which could be limiting for transposition overall, is used preferentially at some targets so that a reduction in cDNA differentially affects individual targets; and (ii) that Lhp1 participates at points in the life cycle subsequent to reverse transcription. Although the LHP1 disruption mutant isolated in our study was not a complete deletion of the LHP1, it is likely that the disruption resulted in loss of Lhp1p function. The NLS of Lhp1p resides within a region that encompasses the RNA recognition motif (Yoo and Wolin, 1994;Rosenblum et al., 1998). The mTn3 insertion in our mutant strain truncated the coding region of LHP1 such that the mutant protein lacked the C-terminal half, including part of the NLS region and RNA recognition motifs. The mutant Lhp1p may simply be unstable as it was not detectable using polyclonal serum raised against a recombinant protein representing residues 1-252 of Lhp1p (Yoo and Wolin, 1997). Processing of pretRNA in the LHP1::mTn mutant extracts was similar to processing reported for a null mutant (data not shown; Yoo and Wolin, 1997). In addition, for the properties tested, a lhp1D mutant was similar to our disruption mutant: decreased cDNA, integration into a chromosomal target (data not shown) and Ty3 mobilization in a genetic assay (Table4). Finally, expression of wild-type LHP1 reversed the mutant phenotype. Although LHP1 was overexpressed under a heterologous promoter, complementation is consistent with a loss-of-function mutation.
The La protein and Lhp1p have been implicated in a broad variety of functions related to RNA-binding activity, and we speculate that this activity is also relevant in the Ty3 transposition context. As reviewed in the Introduction, these functions include pol III transcription termination, release and reinitiation, acting as a chaperone for a number of RNP RNAs, facilitating internal translation and protecting RNAs from degradation. Small RNAs and pol III transcription are directly critical to several stages of the Ty3 life cycle, including frameshifting, reverse transcription and integration targeting. It is therefore attractive to consider models in which the role of Lhp1p in the Ty3 life cycle is related to its role as a pol III transcript chaperone.
After particle assembly and protein processing, reverse transcription is initiated. tRNA iMet is the Ty3 reverse transcription primer (Keeney et al., 1995). Calvo et al. (1999) have shown that Lhp1 is important for maturation of tRNA iMet and reported decreases of about 27% in mature tRNA iMet levels and of about 77% in precursor tRNA iMet in an lhp1 null mutant. Because Lhp1p is an RNA chaperone, it is possible that it functions to facilitate proper folding or binding of tRNA iMet or inclusion of the tRNA iMet or Ty3 RNA in the VLP. Although the decrease that we observed in Ty3 cDNA levels could be consistent with a small decrease in mature tRNA iMet or in its activity as the Ty3 primer, we did not observe a comparable decrease in amounts of tRNA iMet in the lhp1 truncated strain (Fig.3A).
RNA chaperone activity might be required in unwinding and annealing of RNA-DNA duplexes for template switching events during reverse transcription. Human La was shown to possess ATP-dependent helicase activity on RNA-DNA duplexes (Bachmann et al., 1990). However, yeast Lhp1p lacks the ATP-binding motif of mammalian La (Rosenblum et al., 1998), and no helicase activity has been reported. If Lhp1p facilitates reverse transcription in conjunction with NC and RT, it would probably be associated with the VLP. Although there is no evidence for physical association with the VLPs, Lhp1p was detected in the VLP fraction of wild-type cells (Fig.6A). Despite lacking the exact mechanism of Lhp1 action, it is clear that the level of functional Lhp1 is critical for the initial accumulation of Ty3 cDNA (Figs1E and 5B) and for wild-type frequencies of transposition (Figs1C-E and 2).
The different magnitudes of effects on cDNA accumulation and transposition that we observed suggest that the role of Lhp1 in Ty3 transposition might not be exclusive for replication, and may involve a post-replication step. However, reduction in cDNA by a small magnitude could also result in a larger magnitude of defect in transposition. In the case of Ty1, host gene mutations that affect cDNA, in some instances, are also associated with disproportionate effects on transposition. Previously, disruption of DBR1, the gene for the intron-debranching enzyme, was shown to result in a ninefold reduction in Ty1 transposition with no detectable effect on RNA or proteins (Chapman and Boeke, 1991). More recently, dbr1 mutants have been reported to be deficient for Ty3 transposition (Karst et al., 2000). The defect in transposition in the case of both Ty1 and Ty3 is associated with modest decreases in reverse transcripts after 24h of induced expression (Karst et al., 2000). In addition, a strict correlation between the Ty1 cDNA level and transposition was not observed among a large set of mutants for Ty1 transposition identified recently based on a genetic assay (Scholes et al., 2001). Unfortunately, in those cases, the basis of the effect(s) is also not clear.
An alternative explanation for the discrepancy between the different magnitudes of effects of Lhp1 loss on Ty3 cDNA and transposition frequency is that Lhp1p has a role subsequent to cDNA production. A role at integration, for example, would be consistent with different effects of the lhp1 mutation observed for different targets. There was an eightfold decrease in use of the synthetic target and a more than fourfold decrease at one chromosomal target in the lhp1 mutant compared with wild type, but only modest effects at another chromosomal tRNA gene family and another synthetic plasmid target. We speculate that, if Lhp1p affects integration at the target site, the synthetic divergent target is especially sensitive to Lhp1p deficiency. The two tDNAs on this target are inefficiently transcribed (Kinsey and Sandmeyer, 1995;Aye, 2001). Because the positions of the divergent tDNAs preclude the formation of transcription initiation complexes simultaneously over both genes, they may form a target that is inherently unstable, making them particularly sensitive to conditions that destabilize the transcription complex. Although no integration defect correlated in vitro with either VLPs or factors prepared from the lhp1 mutant, the in vitro and in vivo results are not directly comparable because technical limitations of the primer combinations prevented us from specifically testing the divergent target in vitro. In fact, the target used in vitro was tested in vivo using a PCR assay and did not show the same magnitude of defect as the synthetic divergent tRNA gene target.
In this study, we showed that disruption of LHP1 resulted in a substantial decrease in Ty3 transposition into a plasmid-borne tRNA target and one chromosomal tRNA gene family, but not at another. The level of Ty3 reverse transcripts was decreased by half in the mutant strain and accumulated more slowly. Although the biochemical mechanism underlying the lhp1 effect remains to be identified, the RNA-binding function of Lhp1p seems likely to contribute to reverse transcription. The differential integration defect among target tDNAs also suggests that Lhp1p plays a post-replication role as well. Thus, Lhp1 may act at multiple points in the Ty3 life cycle.

Yeast strains and plasmids
Media and standard techniques for yeast were as previously described (Sherman et al., 1986). Strains are described in Table1. Mutants were generated in the haploid yeast strain YPH500 by integration of a yeast library disrupted with insertions of mini-Tn3 carrying the yeast LEU2 gene (Burns et al., 1994). The isolated LEU2 gene excised with BamHI and NarI restriction enzymes from YEp351 was also transformed into YPH500 to obtain the leucine prototrophic strain yMA1235, which is referred to as wild type in this study. Plasmids are described in Table2. The mutant isolated as 22-19 (yMA1237) is a mutant with a reduced frequency of Ty3 transposition into a synthetic, plasmid-borne target. A derivative of YPH500, yMA1322, was used to reconstruct lhp1 mutants. Yeast strains BY4741 and its lhp1D derivative (# 3748) were acquired from Research Genetics.
The LHP1 ORF was amplified by PCR from wild-type genomic DNA, using primer oligonucleotides 477 and 478, and cloned into the pCRII vector (Invitrogen) to create pLHP1. Oligonucleotides are described in Table3. Plasmid pMA1708 was created by cloning a 0.9kb BamHI fragment containing the LHP1 coding region from pLHP1 into the BamHI site of the high-copy, galactose-inducible, yeast expression vector pYES2.0 (Invitrogen).

Yeast mutagenesis
Disruption mutagenesis was performed as described previously. A yeast library was mutagenized by subjecting it to mini-Tn3::lacZ::LEU2 transposition in bacteria (Burns et al., 1994). DNA was prepared from library pools (kindly provided by M. Snyder, Yale University), cleaved with NotI and transformed using the lithium acetate procedure into YPH500, carrying Ty3 expression vector pTM843 (Menees and Sandmeyer, 1994) and Ty3 target plasmid pPK689 (Ito et al., 1983). Disruption mutants were enriched by selection of Leu + transformants on SC medium lacking leucine, tryptophan and histidine.

Ty3 transposition assays
A genetic assay for Ty3 integration (Kinsey and Sandmeyer, 1995) was adapted to screen for Ty3 transposition mutants. Leu + mutant transformants and wild-type strain yMA1235, carrying pTM843 and pPK689, were patched onto synthetic dropout medium with glucose as a carbon source (SC[glu] lacking histidine, tryptophan and leucine). After incubation for 24h at 30∞C, each plate was replicated to SC with galactose as the carbon source (SC[gal] lacking histidine, tryptophan and leucine) for induction of Ty3 expression. After 48h of growth at 30∞C, yeast cells expressing Ty3 were replicated to minimal medium with glucose as the carbon source supplemented with uracil and lysine for detection of transposition events. This medium selected for cells in which Ty3 insertion at the synthetic target activated expression of the ochre suppressor tRNA Tyr , sup2bo, resulting in the suppression of ade2o. Cultures were incubated at 30∞C for 5days, and the number of papillae within mutant and wild-type patches of cells were compared. Cells plated onto SC[glu] lacking histidine, tryptophan and leucine were incubated for 1day at 30∞C (no induction) and replicated to minimal medium supplemented with glucose, uracil and lysine for the negative control. In a quantitative version of this assay (Kinsey and Sandmeyer, 1995), 1¥10 6 cells were plated in triplicate on SC[glu] lacking histidine, tryptophan and leucine. Cultures were replicated onto media as described for the patch assay, and the average number of colonies per plate on final selective medium was determined. Mutants isolated in this screen were retained for analysis only if they grew on selective medium when transformed with a target plasmid containing a Ty3 insertion.
A Ty3 mobilization assay using GAL-Ty3-mhis3AI elements was performed as described previously (Sadeghi et al., 2001) using BY4741 and its lhp1D derivative (#3748). Briefly, at least three transformants for each strain were patched to SC[glu]-ura plates. After 2days at 30∞C, patches were replicated to SC[gal]-ura and incubated at room temperature for 3days. Each patch was scraped off the plate and resuspended in 1ml of water. The cell suspension was serially diluted and plated to SC[glu]-ura and SC[glu]-his plates to titre the cells. After 3days at 30∞C, His + and Ura + prototrophs were counted, and the frequency of His + cells was expressed as the number of His + cells divided by the number of Ura + cells.

Identification of genomic loci disrupted by mTn3
Genes disrupted in each mutant were identified by recovering the disrupted locus in E. coli. Each candidate mutant was transformed with PvuI-linearized, YIp5 plasmid to integrate an E. coli origin of replication into the mTn3 construct. Genomic DNA prepared from each Ura + Leu + transformant was cleaved with NsiI or BglII and treated with T4 DNA ligase to circularize DNA fragments. The ligation mixture was transformed into E. coli strain HB101, and transformants with LEU2-marked plasmids were selected. This selection was facilitated by the ability of the yeast LEU2 gene to complement growth of HB101, which is a leuB mutant. Plasmids with the characteristic restriction pattern of the mTn3 construct were sequenced with M13 (-40) primer, which anneals to proximal lacZ sequence. Each sequence generated was compared with complete S. cerevisiae genomic DNA using the BLASTN search of the Saccharomyces Genome Database (http://genome-http://www.stanford.edu/Saccharomyces/) to identify the site of mTn3 insertion. The recovered LHP1 insertion plasmid was cleaved with BglII to linearize the plasmid and transformed into YPH500 and yMA1322 to reconstruct this disruption. Transformants were selected on SC[glu] lacking uracil and leucine; cells that lost the YIp5 URA3 marker by recombination were selected on SC medium containing 5-fluoroorotic acid (5-FOA). Gene disruptions were confirmed by Southern blotting with probes specific for LHP1 and lacZ. These isolates were retested for Ty3 transposition.

Assays for Ty3 chromosomal transposition
Ty3 transposition into chromosomal loci was measured using PCR to amplify a diagnostic fragment generated with primers annealed to Ty3 and to the tDNA. Yeast strains transformed with pTM843 were grown in SC[raff] lacking tryptophan and histidine to an A 600 between 0.2 and 0.4. Galactose was added to a final concentration of 2% to induce Ty3 expression. After 6h at 30∞C, the cells were harvested, and total DNA extracted. The concentration of yeast DNA was quantified by fluorometry in a TKO 100 (Hoefer Scientific Instruments); 1-25ng of DNA was used as the template for PCR with primer oligonucleotides 278 (Ty3) and 676 (14 valine tDNAs) or 677 (10 glutamine tDNAs). PCR conditions were essentially the same as described previously (Menees and Sandmeyer, 1994), except that the Mg 2+ concentration was 2mM and the annealing temperature was 60∞C. In order to visualize exclusively PCR products that included a Ty3 LTR, as expected for junction fragments of interest, 10 ml of 75ml PCR was separated on an agarose gel, and Southern blot analysis was performed with a labelled, NdeI-XhoI fragment of Ty3 LTR. To control for input DNA, the RAD52 locus was amplified by PCR using primers 712 and 713 under the same PCR conditions, except that 20-25 cycles of amplification were used instead of 40.

Immunoblot analysis
In order to examine the expression of Ty3 proteins, wholecell extracts (WCE) and VLPs were prepared as described previously (Menees and Sandmeyer, 1994). Twenty micrograms of WCE or 1mg of concentrated VLP protein was fractionated by SDS-PAGE, transferred to nitrocellulose membranes (Hybond ECL; Amersham) and incubated with rabbit polyclonal antibodies against CA and IN (Menees and Sandmeyer, 1994). Secondary antibodies to rabbit IgG were detected by the ECL system (Amersham). In a separate set of experiments, twofold dilutions of whole-cell extracts from wild-type and lhp1 cells expressing Ty3 were compared showing that assays sensitive to twofold differences did not detect differences in amounts of Ty3 proteins between the two cell types (data not shown). Immunoblot analysis of Lhp1p was performed using antisera against Lhp1p (a gift from C. Yoo, Yale University) as described previously (Yoo and Wolin, 1994).

DNA analysis
Ty3 cDNA was visualized as described previously (Menees and Sandmeyer, 1994). RNA-free total yeast DNA (1mg) was digested with EcoRI to linearize the Ty3 plasmid and separated by electrophoresis on an agarose gel, transferred to a nylon membrane (Duralon UV; Stratagene) and immobilized by UV cross-linking in a Stratalinker 1800 (Stratagene). Hybridization was performed with a 32 P-labelled, internal BglII fragment of Ty3, which hybridizes with full-length cDNA of 5.4kbp, as well as Ty3 donor plasmid and chromosomal Ty3 elements, but not with LTRs. To confirm that mTn3::lacZ::LEU2 was inserted into LHP1, yeast DNA was digested with EcoRI and processed as described above. Hybridization was performed with 32 P-labelled, LHP1 coding sequence amplified by PCR or a 32 P-labelled, BamHI fragment of lacZ. Lambda DNA digested with HindIII was 32 P labelled and used as the probe to visualize the lambda size markers for Southern blots.

RNA analysis
In order to examine the effect of lhp1 on Ty3 RNA levels, yeast cells transformed with pTM843 were grown for 6h with 2% raffinose or 2% galactose as the carbon source, harvested and total RNA extracted (Clark et al., 1988). RNA samples denatured by glyoxylation and fractionated by agarose gel electrophoresis were transferred to a GeneScreen membrane (Stratagene) and probed with the 32 P-labelled, internal BglII fragment of Ty3. For detection of tRNA iMet , total RNA was separated in 8% polyacrylamide-8.3M urea gel by electrophoresis, transferred to a GeneScreen Plus membrane (Stratagene) and probed with 32 P-labelled oligonucleotide 83, specific for mature tRNA iMet .