Amino Acid Starvation Enhances Programmed Ribosomal Frameshift in Metavirus Ty 3 of Saccharomyces cerevisiae

Ty3 is a retroviral-like element and propagates with a retroviral-like mechanism within the yeast cells. Ty3 mRNA contains two coding regions, which are GAG3 and POL3. The coding region POL3 is translated as a GAG3-POL3 fusion protein by a +1 programmed frameshift. In this study, it was shown that the Ty3 frameshift frequency is significantly increased by amino acid starvation in a Gcn2p complex dependent manner. When the yeast cells were subjected to amino acid starvation, the frameshift frequency of Ty3 increased more than 2-fold in the wild-type yeast cells, mostly independent of Gcn4p. However, Ty3 frameshift frequency remained at basal level in the gcn1, gcn20, or gcn2mutant yeast cells in amino acid starved yeasts. Gcn1p forms a complex with Gcn2p and Gcn20p and is involved in the sensing of uncharged tRNAs on the ribosomal A-site during translation. Increases in uncharged tRNA levels due to amino acid depletion lead to ribosomal pauses. These ribosomal pauses are significant actors in the regulation of Ty3 frameshift frequency. Results of this research revealed that frameshift frequency in Ty3 is regulated by the Gcn2p complex in response to amino acid starvation in yeast.


Introduction
Ty elements of the yeast Saccharomyces cerevisiae belong to the retrotransposon family of eukaryotic mobile genetic elements [1].Ty3 is one of the five different retroviral-like elements that are present in the yeast S. cerevisiae genome [2].Due to its genome organization, Ty3 has been classified as the member of the Metaviridae group of the retrovirales order [3].Ty elements transpose via an RNA intermediate using a similar strategy to the vertebrate retroviruses [4].
Ty3 genome encodes a single mRNA that contains two overlapping coding regions [2].These coding regions are named TY3A (GAG3) and TY3B (POL3), and they encode retroviral gag and pol analogs, respectively.POL3 is synthesized by +1 translational frameshifting as a fusion protein with GAG3 [5,6].Posttranslational cleavage of GAG3 generates nucleocapsid proteins that are required for the formation of TY virus like particles (Ty-VLP).Posttranslational proteolytic cleavage of POL3 gives three different proteins that show protease (PR), integrase (IN), and reverse transcriptase/RNAse H (RT/RH) activities [7].The programmed ribosomal frameshifting (PRF) in Ty3 elements is caused by the slow decoding of AGU codon by tRNA-Ser-GCU followed by the out-of-frame binding of tRNA-Val-GUU at +1 direction in the frameshift site of Ty3 [5].The rate of PRF shows variations among retroviruses and retroviral-like elements [6].The PRF rate is approximately 4-5% in Ty3, while it is 25% in Ty1 under standard growth conditions in most of the laboratory strains of S. cerevisiae (in synthetic complete medium supplemented with 2% glucose in logarithmically grown cells at 30 ∘ C) [8,9].
Several lines of evidence indicate that the PRF is not a constitutive event, and it is regulated by different cellular signals.Previously, we have shown that the programmed ribosomal frameshift (PRF) event in Ty3 is regulated by glucose signaling [9].The PRF rate in Ty3 is activated more than 3-fold in the yeast cells that are grown in low glucose medium in a Snf1p dependent manner.cAMP dependent protein kinase-A is also involved in the regulation of frameshift efficiency in Ty3 [9].In addition, we have analyzed the effects of glucose signaling on the PRF efficiency in Ty1 elements and shown that the Ty1 type +1 PRF frequency decreases several fold in glycerol-lactate-grown S. cerevisiae cells [8].Atkins and Björk [10] have also reported that mutations in certain ribosomal proteins and tRNAs also affect the PRF rate in both bacteria and yeast.
Amino acid starvation causes the transcriptional activation of more than 30 genes of 10 different pathways required for amino acid biosynthesis by the transcriptional activator Gcn4p [11].The key molecular event that activates Gcn4p synthesis is the accumulation of the uncharged tRNAs under amino acid starvation conditions.The protein kinase Gcn2p interacts with the uncharged tRNAs and initiates a cascade of events which results in the derepression of GCN4 expression [12].In addition to amino acid starvation, several metabolic signals also activate Gcn2p [13].Gcn2p is associated with two different effector proteins, namely, Gcn1p and Gcn20p [14].These effector proteins are required for the kinase function of Gcn2p during amino acid starvation conditions [14,15].It was shown that the Gcn2p, associated with Gcn1p and Gcn20p, is also present in elongating ribosomes during translation [15,16].Gcn1p has three different functional regions.These interaction sites are required for ribosome binding, for Gcn2p interactions, and for Gcn20p interactions in amino acid starved yeast cells [17].There is additional evidence indicating that Gcn2p interacts with translation elongation factors eEF1A and eEF3 in different growth conditions.It was shown that eEF1A and eEF3 deactivate Gcn2p function in nutrient replete conditions [16,18].
In this study, I have analyzed the effects of the amino acid starvation on the programmed ribosomal frameshift rate in Ty3.Results of this research indicate that programmed ribosomal frameshift efficiency is regulated by amino acid starvation in a Gcn1p-Gcn20p-Gcn2p complex dependent manner in S. cerevisiae.
Structures and constructions of Ty3 frameshift (Ty3-FS) and Ty3 frame fusion (Ty3-FF) reporter plasmids were explained previously [5].These plasmids are 2 m URA3based shuttle vectors and based on the previously described expression vector pMB38 [19].Ty3 frameshift site (GCG-AGT-T) is fused to the E. coli lacZ gene in the +1 reading frame in Ty3-frameshift reporter plasmid (TY3-FS).In this vector, translation of the lacZ fusion protein depends on the occurrence of frameshift event in the +1 direction at the 5  -GCG-AGU-U-3  sequence of mRNA.The Ty3-frame fusion plasmid does not contain the frameshift site; hence, translation of the lacZ fusion protein does not require a frameshift event (Ty3-FF) [5].Plasmids were transformed into the yeast cells as described previously using the lithium acetate, PEG method [20].These 2 m URA3-based plasmids can be stably maintained with no drastic change in copy numbers in different yeast transformants under selective growth conditions [21].

Growth Conditions. Yeast cells were grown in liquid Yeast
Nitrogen Base (YNB) medium supplemented with 2% glucose and necessary amino acids as described previously [22].For amino acid starvation experiments, yeast transformants were first grown to saturation in 5 mL of liquid YNB media supplemented with 2% glucose and necessary amino acids and then diluted to OD 600 : 0.1-0.15 in 10 mL of fresh YNB liquid media.Once yeast cell density reached OD 600 : 0.5-0.6,yeast cultures were divided into two aliquots (5 mL each).Filter sterilized 3-amino 1,2,4-triazole (3-AT) was added to 10 mM final concentration to the one set of (5 mL) yeast cultures, as described previously [23].3-AT is a competitive inhibitor of HIS3 gene product and triggers amino acid starvation in the yeast cells [24].3-AT treated and untreated yeast transformants were grown for additional 5 hours and harvested for -galactosidase assays.

Enzyme Assays.
-Galactosidase assays were done in triplicate as described previously [25].-Galactosidase units are given in nmol of ONPG (2-nitrophenyl -dgalactopyranoside) cleaved per minute per mg of protein in permeabilized yeast cells.Protein concentrations in the permeabilized yeast lysates were determined by the Lowry assay as described [26].Frameshift rates were calculated as the percent of the ratio of -galactosidase activities expressed from Ty3-FS plasmid to the -galactosidase activities expressed from Ty-FF plasmid in the relevant yeast transformants.Yeast transformants were grown in triplicate and all experiments were repeated at least once under the same growth concentrations.Hence, the Ty3 FS rate values given in the tables are the mean values of at least 18 independent assays.Standard deviations for -galactosidase units were below 10% in triplicate assays.

Effects of the Amino Acid Starvation on the Ty3 Frameshift
Rate.Ty elements do not encode any known regulatory protein for their own gene expression.In this regard, they completely depend on the host-encoded transcription and translation factors for their own replication.Ty elements of S. cerevisiae cannot transpose intercellularly.Being an obligatory resident in yeast cells, genes expressed in Ty elements are also subjected to physiological changes that modulate various metabolic events in yeast cells.
In order to investigate if the frameshift event in Ty3 will be affected from physiological stresses such as amino acid starvation, yeast plasmids that contain Ty3 frameshift (Ty3-FS) and Ty3 frame fusion (Ty3-FF) expression cassettes were transformed into wild-type yeast cells.Yeast transformants were grown in minimal medium supplemented with 2% glucose and auxotrophic requirements.The programmed ribosomal frameshift frequencies for Ty3 in yeast cells grown in normal conditions were measured as 4% ± 0.2 (Table 1).
To test the effects of amino acid starvation on frameshift frequency in Ty3, aliquots of the yeast cells were subjected to amino acid starvation by adding 3-amino triazole, a competitive inhibitor for the HIS3 gene product that triggers amino acid starvation in yeast cells.The PRF frequency of Ty3 in the amino acid starved yeast strain increased more than 2fold and was measured as 9.2% ± 0.5 (Table 1).Amino acid starvation leads to the activation of the protein kinase Gcn2p and transcriptional activator Gcn4p.Transcription factor Gcn4p is a multifunctional transcriptional activator and coordinately activates the transcription of genes involved in amino acid biosynthesis [11,27].In order to test if Gcn4p has any function in the activation of Ty3 frameshift frequency in amino acid starved yeast cells, plasmids containing Ty3-FS and Ty3-FF expression cassettes were transformed into Δgcn4 mutant and its isogenic wild-type strains.Frameshift frequency of Ty3 was measured as 4% ± 0.3 in gcn4 mutant yeast cells in normal growth conditions (Table 1).Frameshift frequency in amino acid starving Δgcn4 mutant yeast cells also increased as in the wild-type strains and resulted in 7% ± 0.5 frameshift frequency (Table 1).These results indicated that Gcn4p is partially involved in the regulation of frameshift frequency in Ty3.

Gcn2p Is Essential for the Regulation of Ty3 Frameshift
Frequency.Programmed ribosomal frameshift events take place during the translational elongation stage.Gcn2p is a protein kinase that has multiple functions in the sensing and signaling processes of amino acid starvation in yeast.Recent evidence also indicates that Gcn2p associates with ribosomes and may also be involved in translational elongation [16,18].Hence, the effects of Gcn2p on frameshift frequency in Ty3 were analyzed in the wild-type and gcn2 mutant yeast cells.Frameshift frequencies of Ty3 in the wild-type yeast cells were measured as 4% ± 0.2 and 9.2% ± 0.5 in normal and in amino acid starved yeast cells, respectively (Table 2).Frameshift frequency of Ty3 was measured as 4% ± 0.4 in gcn2 mutant cells that were grown under normal growth conditions (Table 2).When the gcn2 mutant yeast cells were subjected to amino acid starvation, frameshift frequency remained approximately at the same levels (5% ± 0.5).These results indicated that functional Gcn2p is essential for the regulation of the frameshift event in amino acid starved yeast cells.

Gcn1p and Gcn20p are Involved in the Regulation of Ty3
Frameshift Frequency.Previous results suggest a model in which Gcn1p and Gcn20p associate with Gcn2p in elongating ribosomes [16,18].After showing that Gcn2p is involved in the activation of frameshift during amino acid starvation, the effects of Gcn1p and Gcn20p on the frameshift frequency of Ty3 were also analyzed in the wild-type and isogenic gcn1, gcn20, and gcn1/gcn20 double mutant yeast strains.Yeast plasmids that have Ty3-FS and Ty3-FF expression cassettes were transformed into wild-type and isogenic gcn1, gcn20, and gcn1/gcn20 double mutant yeasts.Transformants were grown under normal growth conditions and were subjected to amino acid starvation by adding 3-AT to the growth mediums.
Frameshift frequency in the wild-type gcn1, gcn20, and gcn1/gcn20 double mutants remained at approximately the same basal levels (approximately 4% ± 0.5) as in the yeast cells under standard growth conditions (Table 3).As expected, frameshift frequency of Ty3 in the wild-type yeast strain increased more than 2-fold (from 4% to 9%) in amino acid starved yeast cells (Table 3).However, when the gcn1, gcn20, and gcn1/gcn20 mutants were subjected to amino acid starvation to activate the Gcn2p pathway, frameshift frequency of Ty3 did not increase in these gcn mutants and remained essentially at the same basal levels (approximately 4% ± 0.5) (Table 3).These results suggest that the functional Gcn2p complex which also contains Gcn1p and Gcn20p is involved in the regulation of frameshift frequency of Ty3 in amino acid starved yeast cells.Moreover, it is clear from the gcn1/gcn20 double mutant that Gcn1p and Gcn2p do not have any cumulative effects on the frameshift efficiency in Ty3 during amino acid starvation.

Discussion
Programmed ribosomal frameshift is one of the regulatory mechanisms in retroviral gene expression [6].It occurs during the translational elongation stage.The molecular mechanisms of the PRF in Ty3 have been uncovered previously [5,28].The core nucleotide sequence of the frameshift site on Ty3 is GCG-AGU-U.Frameshift is stimulated by the ribosomal pause resulting from the slow decoding of the AGU codon when it arrives in the A-site of elongating ribosomes.Peptidyl-tRNA Ala which decodes the GCG codon associates with the P-site of the ribosome.Frameshift occurs at +1 direction without tRNA slippage by out-of-frame binding of valyl-tRNA to the GUU codon at the A-site of the ribosome [28].Previously, it was reported that the frameshift frequency in Ty3 can be modulated by growth conditions [9].It is activated by nonfermentable carbon sources up to 3-fold in a Snf1p dependent manner.
In this study, the effects of amino acid starvation caused by 3-AT on the frameshift frequency in Ty3 were analyzed.3-AT inhibits the histidine biosynthesis by acting on the product of the HIS3 gene, imidazole glycerol-phosphate dehydratase [23,24].Presence of 3-AT also increases the amount of uncharged tRNA within cytoplasm, which results in the activation of Gcn2p kinase activity [11,12].
It is found that amino acid starvation increases frameshift frequency more than 2-fold in a Gcn1p dependent manner.Gcn1p, which also interacts with Gcn20p, is the effector protein for the activation of the kinase function of Gcn2p [29].Gcn1p-Gcn20p-Gcn2p complex associates with ribosomal A-site in amino acid starved yeast cells [17,29].It can be proposed that the interaction of this complex with ribosomal A-site is crucial for the induction of the Gcn2p kinase activity [29].Ribosomal pause is one of the critical reasons for the frameshift in Ty3.It was suggested that the Gcn1p-Gcn2p complex delivers uncharged tRNAs to ribosomal A-site during amino acid starvation.Hence, it is reasonable to propose that when the A-site is occupied by uncharged tRNAs, it leads to an increase in ribosomal pause in amino acid starved yeast cells.This prolonged pause further increases the frameshift frequency in Ty3 in wildtype GCN + yeast cells.Longer pause durations increase the likelihood of out-of-frame binding of valyl-tRNA to the GUU codon at the A-site of the ribosome.Frameshift frequency in any one of the gcn1, gcn20, or in gcn2 mutant strains did not change in amino acid starvation conditions and remained at amino acid replete levels.It is also clear that the lack of both Gcn1p and Gcn20p in the double mutant strain did not have any cumulative effect on the frameshift frequency, which indicates that all of these factors are involved in the increase of frameshift frequency in Ty3 in amino acid starved yeast cells.
Gcn4p is a multifunctional transcriptional activator and its translation is also activated by amino acid starvation in a Gcn2p dependent manner [30].Gcn4 activates the expression of purine biosynthetic genes [31].Gcn4p is also involved in the regulation of the genes encoding ribosomal proteins [32].To test if Gcn4p affects frameshift frequency indirectly, frameshift frequency of Ty3 was also analyzed in the gcn4 mutant strain.Results of this study showed that Gcn4p has a partial effect on the regulation of frameshift frequency in the amino acid starved yeast cells.
Programmed ribosomal frameshift is a good model to use in the analysis of the regulation of translation elongation stage.PRF in Ty3 can be used to analyze the regulatory events during the translation elongation of protein synthesis.Another question is how this change will affect the replicative cycle of Ty3 within the yeast.It was suggested that yeast Ty elements are under continuous selective pressure for the repression of transposition and for elimination from the yeast cell.Ty elements must overcome host pressure and elimination and still coexist with the yeast genome [33].It has already been shown that the alterations in frameshift rate dramatically affect the transposition events in the yeast Ty1 elements [34].Reduction of TYB protein levels, which are similar to retroviral pol products, resulted in a decreased rate of transposition in the yeast [35].Kirchner et al. [35] also found that the perturbation of GAG3 and POL3 ratios also affects the transposition rate of Ty3 in S. cerevisiae.

Conclusion
It is known that environmental factors such as various stresses can increase transposable elements' mobility to generate genetic diversity in the host organisms.Amino acid starvation and the glucose limitations are metabolic stress conditions that affect many cellular processes and the pathways in the yeast cells.Results presented in this study indicate that the programmed frameshift level in the metavirus Ty3 can be regulated by nutrient limiting conditions through the protein kinase Gcn2p complex.Increases in the frameshift rate in response to these stresses also lead to the biosynthesis of more TY3B polypeptide which might lead to increases in the Ty3 propagation in the host cells.

Table 1 :
Amino acid starvation increases PRF efficiency in a partially Gcn4p dependent manner in Ty3.% frameshift rates were calculated as the percent of the ratio of galactosidase activities expressed from the Ty3-frameshift plasmid to the galactosidase activities expressed from the Ty3-frame fusion plasmid in the relevant yeast transformants.

Table 2 :
Gcn2p is required for the increase of PRF efficiency in response to amino acid starvation.

Table 3 :
Gcn1p and Gcn20p are required for the increase of PRF efficiency in response to amino acid starvation.
* % frameshift rates were calculated as the percent of the ratio of galactosidase activities expressed from the Ty3-frameshift plasmid to the galactosidase activities expressed from the Ty3-frame fusion plasmid in the relevant yeast transformants.±SD: standard deviations.