Involvement of an initiation factor and protein phosphorylation in translational control of GCN4 mRNA.

Regulation of the GCN4 gene of Saccharomyces cerevisiae is one of the best-documented instances of gene-specific translational control in an eukaryote. Upstream open reading frames (uORFs) in GCN4 mRNA modulate the flow of scanning ribosomes to the GCN4 start codon according to the availability of amino acids. Recent results suggest that sequences at the termination codons of the uORFs, a general initiation factor, and a protein kinase all make important contributions to the proper functioning of this interesting translational-control element.

Translational control of GCN4: an in vivo barometer of initiation-factor activity.

Phosphorylation of translation initiation factor-2 (eIF-2) is an adaptive mechanism for downregulating protein synthesis under conditions of starvation and stress. The yeast Saccharomyces has evolved a sophisticated means of increasing translation of GCN4 mRNA when eIF-2 is phosphorylated, allowing the induction of an important stress-response protein when expression of most other genes is decreasing. Because translation of GCN4 mRNA is so tightly coupled to eIF-2 activity, genetic analysis of this system has provided unexpected insights into the regulation of eIF-2 and its guanine nucleotide exchange factor, eIF-2B.

A hierarchy of trans-acting factors modulates translation of an activator of amino acid biosynthetic genes in Saccharomyces cerevisiae.

The GCN4 gene encodes a positive effector of amino acid biosynthetic genes in Saccharomyces cerevisiae. Genetic analysis has suggested that GCN4 is regulated by a hierarchy of interacting positive and negative effectors in response to amino acid starvation. Results presented here for a GCN4-lacZ gene fusion support this regulatory model and suggest that the regulators of GCN4 exert their effects primarily at the level of translation of GCN4 mRNA. Both the GCN2 and GCN3 products appear to stimulate translation of GCN4 mRNA in response to amino acid starvation, because a recessive mutation in either gene blocked derepression of GCN4-lacZ fusion enzyme levels but did not reduce the fusion transcript level relative to that in wild-type cells grown in the same conditions. The GCD1 product appears to inhibit translation of GCN4 mRNA because under certain growth conditions, the gcd1-101 mutation led to derepression of the GCN4-lacZ fusion enzyme level in the absence of any increase in the fusion transcript level. In addition, the gcd1-101 mutation suppressed the low translational efficiency of GCN4-lacZ mRNA observed in gcn2- and gcn3- cells. A deletion of four small open reading frames in the 5' leader of GCN4-lacZ mRNA mimicked the effect of a gcd1 mutation and derepressed translation of the fusion transcript in the absence of either starvation conditions or the GCN2 and GCN3 products. By contrast, in a gcd1- strain, the deletion resulted in little additional increase in the translational efficiency of the fusion transcript. These results suggest that GCD1 mediates the translational repression normally exerted by the GCN4 leader sequences and that GCN2 and GCN3 antagonize these negative elements in response to amino acid starvation. The effects of the trans-acting mutations on the translation of GCN4-lacZ mRNA remained intact even when transcription of the fusion gene was placed under the control of the S. cerevisiae GAL1 transcriptional control element.

Molecular analysis of GCN3, a translational activator of GCN4: evidence for posttranslational control of GCN3 regulatory function.

GCN4 encodes a transcriptional activator of amino acid biosynthetic genes in Saccharomyces cerevisiae. The GCN3 product is a positive regulator required for increased synthesis of GCN4 protein in amino acid-starved cells. GCN3 appears to act indirectly by antagonizing GCD-encoded negative regulators of GCN4 expression under starvation conditions; however, GCN3 can also suppress the effects of gcd12 mutations under nonstarvation conditions. These results imply that the GCN3 product can promote either repression or activation of GCN4 expression depending on amino acid availability. We present a complete physical description of the GCN3 gene and its transcript, plus measurements of GCN3 expression at the transcriptional and translational levels under different growth conditions. GCN3 encodes a 305-amino-acid polypeptide with no significant homology to any other known protein sequence. GCN3 mRNA contains no leader AUG codons, and no potential GCN4 binding sites were found in GCN3 5' noncoding DNA. In accord with the absence of these regulatory sequences found at other genes in the general control system, GCN3 mRNA and a GCN3-lacZ fusion enzyme are present at similar levels under both starvation and nonstarvation conditions. These data suggest that modulation of GCN3 regulatory function in response to amino acid availability occurs posttranslationally. A gcn3 deletion leads to unconditional lethality in a gcd1-101 mutant, supporting the idea that GCN3 is expressed under normal growth conditions and cooperates with the GCD1 product under these circumstances to carry out an essential cellular function. We describe a point mutation that adds three amino acids to the carboxyl terminus of GCN3, which inactivates its positive regulatory function required under starvation conditions without impairing its ability to promote functions carried out by GCD12 under nonstarvation conditions.

Multiple GCD genes required for repression of GCN4, a transcriptional activator of amino acid biosynthetic genes in Saccharomyces cerevisiae.

GCN4 encodes a positive regulator of multiple unlinked genes encoding amino acid biosynthetic enzymes in Saccharomyces cerevisiae. Expression of GCN4 is coupled to amino acid availability by a control mechanism involving GCD1 as a negative effector and GCN1, GCN2, and GCN3 as positive effectors of GCN4 expression. We used reversion of a gcn2 gcn3 double mutation to isolate new alleles of GCD1 and mutations in four additional GCD genes which we designate GCD10, GCD11, GCD12, and GCD13. All of the mutations lead to constitutive derepression of HIS4 transcription in the absence of the GCN2+ and GCN3+ alleles. By contrast, the gcd mutations require the wild-type GCN4 allele for their derepressing effect, suggesting that each acts by influencing the level of GCN4 activity in the cell. Consistent with this interpretation, mutations in each GCD gene lead to constitutive derepression of a GCN4::lacZ gene fusion. Thus, at least five gene products are required to maintain the normal repressed level of GCN4 expression in nonstarvation conditions. Interestingly, the gcd mutations are pleiotropic and also affect growth rate in nonstarvation conditions. In addition, certain alleles lead to a loss of M double-stranded RNA required for the killer phenotype. This pleiotropy suggests that the GCD gene products contribute to an essential cellular function, in addition to, or in conjunction with, their role in GCN4 regulation.

Translation of the yeast transcriptional activator GCN4 is stimulated by purine limitation: implications for activation of the protein kinase GCN2.

The transcriptional activator protein GCN4 is responsible for increased transcription of more than 30 different amino acid biosynthetic genes in response to starvation for a single amino acid. This induction depends on increased expression of GCN4 at the translational level. We show that starvation for purines also stimulates GCN4 translation by the same mechanism that operates in amino acid-starved cells, being dependent on short upstream open reading frames in the GCN4 mRNA leader, the phosphorylation site in the alpha subunit of eukaryotic translation initiation factor 2 (eIF-2 alpha), the protein kinase GCN2, and translational activators of GCN4 encoded by GCN1 and GCN3. Biochemical experiments show that eIF-2 alpha is phosphorylated in response to purine starvation and that this reaction is completely dependent on GCN2. As expected, derepression of GCN4 in purine-starved cells leads to a substantial increase in HIS4 expression, one of the targets of GCN4 transcriptional activation. gcn mutants that are defective for derepression of amino acid biosynthetic enzymes also exhibit sensitivity to inhibitors of purine biosynthesis, suggesting that derepression of GCN4 is required for maximal expression of one or more purine biosynthetic genes under conditions of purine limitation. Analysis of mRNAs produced from the ADE4, ADE5,7, ADE8, and ADE1 genes indicates that GCN4 stimulates the expression of these genes under conditions of histidine starvation, and it appeared that ADE8 mRNA was also derepressed by GCN4 in purine-starved cells. Our results indicate that the general control response is more global than was previously imagined in terms of the type of nutrient starvation that elicits derepression of GCN4 as well as the range of target genes that depend on GCN4 for transcriptional activation.

Association of RAP1 binding sites with stringent control of ribosomal protein gene transcription in Saccharomyces cerevisiae.

An amino acid limitation in bacteria elicits a global response, called stringent control, that leads to reduced synthesis of rRNA and ribosomal proteins and increased expression of amino acid biosynthetic operons. We have used the antimetabolite 3-amino-1,2,4-triazole to cause histidine limitation as a means to elicit the stringent response in the yeast Saccharomyces cerevisiae. Fusions of the yeast ribosomal protein genes RPL16A, CRY1, RPS16A, and RPL25 with the Escherichia coli lacZ gene were used to show that the expression of these genes is reduced by a factor of 2 to 5 during histidine-limited exponential growth and that this regulation occurs at the level of transcription. Stringent regulation of the four yeast ribosomal protein genes was shown to be associated with a nucleotide sequence, known as the UASrpg (upstream activating sequence for ribosomal protein genes), that binds the transcriptional regulatory protein RAP1. The RAP1 binding sites also appeared to mediate the greater ribosomal protein gene expression observed in cells growing exponentially than in cells in stationary phase. Although expression of the ribosomal protein genes was reduced in response to histidine limitation, the level of RAP1 DNA-binding activity in cell extracts was unaffected. Yeast strains bearing a mutation in any one of the genes GCN1 to GCN4 are defective in derepression of amino acid biosynthetic genes in 10 different pathways under conditions of histidine limitation. These Gcn- mutants showed wild-type regulation of ribosomal protein gene expression, which suggests that separate regulatory pathways exist in S. cerevisiae for the derepression of amino acid biosynthetic genes and the repression of ribosomal protein genes in response to amino acid starvation.

Identification of a regulatory subcomplex in the guanine nucleotide exchange factor eIF2B that mediates inhibition by phosphorylated eIF2.

Eukaryotic translation initiation factor 2B (eIF2B) is a five-subunit complex that catalyzes guanine nucleotide exchange on eIF2. Phosphorylation of the alpha subunit of eIF2 [creating eIF2(alphaP]) converts eIF2 x GDP from a substrate to an inhibitor of eIF2B. We showed previously that the inhibitory effect of eIF2(alphaP) can be decreased by deletion of the eIF2B alpha subunit (encoded by GCN3) and by point mutations in the beta and delta subunits of eIF2B (encoded by GCD7 and GCD2, respectively). These findings, plus sequence similarities among GCD2, GCD7, and GCN3, led us to propose that these proteins comprise a regulatory domain that interacts with eIF2(alphaP) and mediates the inhibition of eIF2B activity. Supporting this hypothesis, we report here that overexpression of GCD2, GCD7, and GCN3 specifically reduced the inhibitory effect of eIF2(alphaP) on translation initiation in vivo. The excess GCD2, GCD7, and GCN3 were coimmunoprecipitated from cell extracts, providing physical evidence that these three proteins can form a stable subcomplex. Formation of this subcomplex did not compensate for a loss of eIF2B function by mutation and in fact lowered eIF2B activity in strains lacking eIF2(alphaP). These findings indicate that the trimeric subcomplex does not possess guanine nucleotide exchange activity; we propose, instead, that it interacts with eIF2(alphaP) and prevents the latter from inhibiting native eIF2B. Overexpressing only GCD2 and GCD7 also reduced eIF2(alphaP) toxicity, presumably by titrating GCN3 from eIF2B and producing the four-subunit form of eIF2B that is less sensitive to eIF2(alphaP). This interpretation is supported by the fact that overexpressing GCD2 and GCD7 did not reduce eIF2(alphaP) toxicity in a strain lacking GCN3; however, it did suppress the impairment of eIF2B caused by the gcn3c-R104K mutation. An N-terminally truncated GCD2 protein interacted with other eIF2B subunits only when GCD7 and GCN3 were overexpressed, in accordance with the idea that the portion of GCD2 homologous to GCD7 and GCN3 is sufficient for complex formation by these three proteins. Together, our results provide strong evidence that GCN3, GCD7, and the C-terminal half of GCD2 comprise the regulatory domain in eIF2B.

Effect of sequence context at stop codons on efficiency of reinitiation in GCN4 translational control.

Translational control of the GCN4 gene involves two short open reading frames in the mRNA leader (uORF1 and uORF4) that differ greatly in the ability to allow reinitiation at GCN4 following their own translation. The low efficiency of reinitiation characteristic of uORF4 can be reconstituted in a hybrid element in which the last codon of uORF1 and 10 nucleotides 3' to its stop codon (the termination region) are substituted with the corresponding nucleotides from uORF4. To define the features of these 13 nucleotides that determine their effects on reinitiation, we separately randomized the sequence of the third codon and termination region of the uORF1-uORF4 hybrid and selected mutant alleles with the high-level reinitiation that is characteristic of uORF1. The results indicate that many different A+U-rich triplets present at the third codon of uORF1 can overcome the inhibitory effect of the termination region derived from uORF4 on the efficiency of reinitiation at GCN4. Efficient reinitiation is not associated with codons specifying a particular amino acid or isoacceptor tRNA. Similarly, we found that a diverse collection of A+U-rich sequences present in the termination region of uORF1 could restore efficient reinitiation at GCN4 in the presence of the third codon derived from uORF4. To explain these results, we propose that reinitiation can be impaired by stable base pairing between nucleotides flanking the uORF1 stop codon and either the tRNA which pairs with the third codon, the rRNA, or sequences located elsewhere in GCN4 mRNA. We suggest that these interactions delay the resumption of scanning following peptide chain termination at the uORF and thereby lead to ribosome dissociation from the mRNA.

The positive regulatory function of the 5'-proximal open reading frames in GCN4 mRNA can be mimicked by heterologous, short coding sequences.

Translational control of GCN4 expression in the yeast Saccharomyces cerevisiae is mediated by multiple AUG codons present in the leader of GCN4 mRNA, each of which initiates a short open reading frame of only two or three codons. Upstream AUG codons 3 and 4 are required to repress GCN4 expression in normal growth conditions; AUG codons 1 and 2 are needed to overcome this repression in amino acid starvation conditions. We show that the regulatory function of AUG codons 1 and 2 can be qualitatively mimicked by the AUG codons of two heterologous upstream open reading frames (URFs) containing the initiation regions of the yeast genes PGK and TRP1. These AUG codons inhibit GCN4 expression when present singly in the mRNA leader; however, they stimulate GCN4 expression in derepressing conditions when inserted upstream from AUG codons 3 and 4. This finding supports the idea that AUG codons 1 and 2 function in the control mechanism as translation initiation sites and further suggests that suppression of the inhibitory effects of AUG codons 3 and 4 is a general consequence of the translation of URF 1 and 2 sequences upstream. Several observations suggest that AUG codons 3 and 4 are efficient initiation sites; however, these sequences do not act as positive regulatory elements when placed upstream from URF 1. This result suggests that efficient translation is only one of the important properties of the 5' proximal URFs in GCN4 mRNA. We propose that a second property is the ability to permit reinitiation following termination of translation and that URF 1 is optimized for this regulatory function.

GCN1, a translational activator of GCN4 in Saccharomyces cerevisiae, is required for phosphorylation of eukaryotic translation initiation factor 2 by protein kinase GCN2.

Phosphorylation of the alpha subunit of eukaryotic translation initiation factor 2 (eIF-2 alpha) by the protein kinase GCN2 mediates increased translation of the transcriptional activator GCN4 in amino acid-starved yeast cells. We show that this key phosphorylation event and the attendant translational induction of GCN4 are dependent on the product of a previously uncharacterized gene, GCN1. Inactivation of GCN1 did not affect the level of eIF-2 alpha phosphorylation when mammalian eIF-2 alpha kinases were expressed in yeast cells in place of GCN2, arguing against an involvement of GCN1 in dephosphorylation of eIF-2 alpha. In addition, while GCN1 is required in vivo for phosphorylation of eIF-2 alpha by GCN2, cell extracts from gcn1 delta strains contained wild-type levels of GCN2 eIF-2 alpha-kinase activity. On the basis of these results, we propose that GCN1 is not needed for GCN2 kinase activity per se but is required for in vivo activation of GCN2 in response to the starvation signal, uncharged tRNA. GCN1 encodes a protein of 297 kDa with an 88-kDa region that is highly similar in sequence to translation elongation factor 3 identified in several fungal species. This sequence similarity raises the possibility that GCN1 interacts with ribosomes or tRNA molecules and functions in conjunction with GCN2 in monitoring uncharged tRNA levels during the process of translation elongation.

Requirements for intercistronic distance and level of eukaryotic initiation factor 2 activity in reinitiation on GCN4 mRNA vary with the downstream cistron.

Translational control of the GCN4 gene in response to amino acid availability is mediated by four short open reading frames in the GCN4 mRNA leader (uORFs) and by phosphorylation of eukaryotic initiation factor 2 (eIF-2). We have proposed that reducing eIF-2 activity by phosphorylation of its alpha subunit or by a mutation in the eIF-2 recycling factor eIF-2B allows ribosomes which have translated the 5'-proximal uORF1 to bypass uORF2 to uORF4 and reinitiate at GCN4 instead. In this report, we present two lines of evidence that all ribosomes which synthesize GCN4 have previously translated uORF1, resumed scanning, and reinitiated at the GCN4 start site. First, GCN4 expression was abolished when uORF1 was elongated to make it overlap the beginning of the GCN4 coding region. Second, GCN4 expression was reduced as uORF1 was moved progressively closer to GCN4, decreasing to only 5% of the level seen in the absence of all uORFs when only 32 nucleotides separated uORF1 from GCN4. We additionally found that inserting small synthetic uORFs between uORF4 and GCN4 inhibited GCN4 expression under derepressing conditions, confirming the idea that reinitiation at GCN4 under conditions of diminished eIF-2 activity is proportional to the distance of the reinitiation site downstream from uORF1. While uORF4 and GCN4 appear to be equally effective at capturing ribosomes scanning downstream from the 5' cap of mRNA, these two ORFs differ greatly in their ability to capture reinitiating ribosomes scanning from uORF1. When the active form of eIF-2 is present at high levels, reinitiation appears to be much more efficient at uORF4 than at GCN4 when each is located very close to uORF1. Under conditions of reduced recycling of eIF-2, reinitiation at uORF4 is substantially suppressed, which allows ribosomes to reach the GCN4 start site; in contrast, reinitiation at GCN4 in constructs lacking uORF4 is unaffected by decreasing the level of eIF-2 activity. This last finding raises the possibility that time-dependent binding to ribosomes of a second factor besides the eIF-2-GTP-Met-tRNA(iMet) ternary complex is rate limiting for reinitiation at GCN4. Moreover, our results show that the efficiency of translational reinitiation can be strongly influenced by the nature of the downstream cistron as well as the intercistronic distance.

Ribosome association of GCN2 protein kinase, a translational activator of the GCN4 gene of Saccharomyces cerevisiae.

The GCN4 gene of the yeast Saccharomyces cerevisiae encodes a transcriptional activator of amino acid biosynthetic genes that is regulated at the translational level according to the availability of amino acids. GCN2 is a protein kinase required for increased translation of GCN4 mRNA in amino acid-starved cells. Centrifugation of cell extracts in sucrose gradients indicated that GCN2 comigrates with ribosomal subunits and polysomes. The fraction of GCN2 cosedimenting with polysomes was reduced under conditions in which polysomes were dissociated, suggesting that GCN2 is physically bound to these structures. When the association of 40S and 60S subunits was prevented by omitting Mg2+ from the gradient, almost all of the GCN2 comigrated with 60S ribosomal subunits, and it remained bound to these particles during gel electrophoresis under nondenaturing conditions. GCN2 could be dissociated from 60S subunits by 0.5 M KCl, suggesting that it is loosely associated with ribosomes rather than being an integral ribosomal protein. Accumulation of GCN2 on free 43S-48S particles and 60S subunits occurred during polysome runoff in vitro and under conditions of reduced growth rate in vivo. These observations, plus the fact that GCN2 shows preferential association with free ribosomal subunits during exponential growth, suggest that GCN2 interacts with ribosomes during the translation initiation cycle. The extreme carboxyl-terminal segment of GCN2 is essential for its interaction with ribosomes. These sequences are also required for the ability of GCN2 to stimulate GCN4 translation in vivo, leading us to propose that ribosome association by GCN2 is important for its access to substrates in the translational machinery or for detecting uncharged tRNA in amino acid-starved cells.

Homologous segments in three subunits of the guanine nucleotide exchange factor eIF2B mediate translational regulation by phosphorylation of eIF2.

eIF2B is a five-subunit guanine nucleotide exchange factor that is negatively regulated by phosphorylation of the alpha subunit of its substrate, eIF2, leading to inhibition of translation initiation. To analyze this regulatory mechanism, we have characterized 29 novel mutations in the homologous eIF2B subunits encoded by GCD2, GCD7, and GCN3 that reduce or abolish inhibition of eIF2B activity by eIF2 phosphorylated on its alpha subunit [eIF2(alphaP)]. Most, if not all, of the mutations decrease sensitivity to eIF2(alphaP) without excluding GCN3, the nonessential subunit, from eIF2B; thus, all three proteins are critical for regulation of eIF2B by eIF2(alphaP). The mutations are clustered at both ends of the homologous region of each subunit, within two segments each of approximately 70 amino acids in length. Several mutations alter residues at equivalent positions in two or all three subunits. These results imply that structurally similar segments in GCD2, GCD7, and GCN3 perform related functions in eIF2B regulation. We propose that these segments form a single domain in eIF2B that makes multiple contacts with the alpha subunit of eIF2, around the phosphorylation site, allowing eIF2B to detect and respond to phosphoserine at residue 51. Most of the eIF2 is phosphorylated in certain mutants, suggesting that these substitutions allow eIF2B to accept phosphorylated eIF2 as a substrate for nucleotide exchange.

Mutations in the GCD7 subunit of yeast guanine nucleotide exchange factor eIF-2B overcome the inhibitory effects of phosphorylated eIF-2 on translation initiation.

Phosphorylation of the alpha subunit of eukaryotic translation initiation factor 2 (eIF-2 alpha) impairs translation initiation by inhibiting the guanine nucleotide exchange factor for eIF-2, known as eIF-2B. In Saccharomyces cerevisiae, phosphorylation of eIF-2 alpha by the protein kinase GCN2 specifically stimulates translation of GCN4 mRNA in addition to reducing general protein synthesis. We isolated mutations in several unlinked genes that suppress the growth-inhibitory effect of eIF-2 alpha phosphorylation catalyzed by mutationally activated forms of GCN2. These suppressor mutations, affecting eIF-2 alpha and the essential subunits of eIF-2B encoded by GCD7 and GCD2, do not reduce the level of eIF-2 alpha phosphorylation in cells expressing the activated GCN2c kinase. Four GCD7 suppressors were shown to reduce the derepression of GCN4 translation in cells containing wild-type GCN2 under starvation conditions or in GCN2c strains. A fifth GCD7 allele, constructed in vitro by combining two of the GCD7 suppressors mutations, completely impaired the derepression of GCN4 translation, a phenotype characteristic of deletions in GCN1, GCN2, or GCN3. This double GCD7 mutation also completely suppressed the lethal effect of expressing the mammalian eIF-2 alpha kinase dsRNA-PK in yeast cells, showing that the translational machinery had been rendered completely insensitive to phosphorylated eIF-2. None of the GCD7 mutations had any detrimental effect on cell growth under nonstarvation conditions, suggesting that recycling of eIF-2 occurs efficiently in the suppressor strains. We propose that GCD7 and GCD2 play important roles in the regulatory interaction between eIF-2 and eIF-2B and that the suppressor mutations we isolated in these genes decrease the susceptibility of eIF-2B to the inhibitory effects of phosphorylated eIF-2 without impairing the essential catalytic function of eIF-2B in translation initiation.

Dimerization by translation initiation factor 2 kinase GCN2 is mediated by interactions in the C-terminal ribosome-binding region and the protein kinase domain.

The protein kinase GCN2 stimulates translation of the transcriptional activator GCN4 in yeast cells starved for amino acids by phosphorylating translation initiation factor 2. Several regulatory domains, including a pseudokinase domain, a histidyl-tRNA synthetase (HisRS)-related region, and a C-terminal (C-term) segment required for ribosome association, have been identified in GCN2. We used the yeast two-hybrid assay, coimmunoprecipitation analysis, and in vitro binding assays to investigate physical interactions between the different functional domains of GCN2. A segment containing about two thirds of the protein kinase (PK) catalytic domain and another containing the C-term region of GCN2 interacted with themselves in the two-hybrid assay, and both the PK and the C-term domains could be coimmunoprecipitated with wild-type GCN2 from yeast cell extracts. In addition, in vitro-translated PK and C-term segments showed specific binding in vitro to recombinant glutathione S-transferase (GST)-PK and GST-C-term fusion proteins, respectively. Wild-type GCN2 could be coimmunoprecipitated with a full-length LexA-GCN2 fusion protein from cell extracts, providing direct evidence for dimerization by full-length GCN2 molecules. Deleting the C-term or PK segments abolished or reduced, respectively, the yield of GCN2-LexA-GCN2 complexes. These results provide in vivo and in vitro evidence that GCN2 dimerizes through self-interactions involving the C-term and PK domains. The PK domain showed pairwise in vitro binding interactions with the pseudokinase, HisRS, and C-term domains; additionally, the HisRS domain interacted with the C-term region. We propose that physical interactions between the PK domain and its flanking regulatory regions and dimerization through the PK and C-term domains both play important roles in restricting GCN2 kinase activity to amino acid-starved cells.

Identification of positive-acting domains in GCN2 protein kinase required for translational activation of GCN4 expression.

GCN4 is a transcriptional activator of amino acid-biosynthetic genes in the yeast Saccharomyces cerevisiae. GCN2, a translational activator of GCN4 expression, contains a domain homologous to the catalytic subunit of eucaryotic protein kinases. Substitution of a highly conserved lysine residue in the kinase domain abolished GCN2 regulatory function in vivo and its ability to autophosphorylate in vitro, indicating that GCN2 acts as a protein kinase in stimulating GCN4 expression. Elevated GCN2 gene dosage led to derepression of GCN4 under nonstarvation conditions; however, we found that GCN2 mRNA and protein levels did not increase in wild-type cells in response to amino acid starvation. Therefore, it appears that GCN2 protein kinase function is stimulated posttranslationally in amino acid-starved cells. Three dominant-constitutive GCN2 point mutations were isolated that led to derepressed GCN4 expression under nonstarvation conditions. Two of the GCN2(Con) mutations mapped in the kinase domain itself. The third mapped just downstream from a carboxyl-terminal segment homologous to histidyl-tRNA synthetase (HisRS), which we suggested might function to detect uncharged tRNA in amino acid-starved cells and activate the adjacent protein kinase moiety. Deletions and substitutions in the HisRS-related sequences and in the carboxyl-terminal segment in which one of the GCN2(Con) mutation mapped abolished GCN2 positive regulatory function in vivo without lowering autophosphorylation activity in vitro. These results suggest that sequences flanking the GCN2 protein kinase moiety are positive-acting domains required to increase recognition of physiological substrates or lower the requirement for uncharged tRNA to activate kinase activity under conditions of amino acid starvation.

Genetic evidence that ferric reductase is required for iron uptake in Saccharomyces cerevisiae.

The requirement for a reduction step in cellular iron uptake has been postulated, and the existence of plasma membrane ferric reductase activity has been described in both procaryotic and eucaryotic cells. In the yeast Saccharomyces cerevisiae, there is an externally directed reductase activity that is regulated by the concentration of iron in the growth medium; maximal activity is induced by iron starvation. We report here the isolation of a mutant of S. cerevisiae lacking the reductase activity. This mutant is deficient in the uptake of ferric iron and is extremely sensitive to iron deprivation. Genetic analysis of the mutant demonstrates that the reductase and ferric uptake deficiencies are due to a single mutation that we designate fre1-1. Both phenotypes cosegregate in meiosis, corevert with a frequency of 10(-7), and are complemented by a 3.5-kilobase fragment of genomic DNA from wild-type S. cerevisiae. This fragment contains FRE1, the wild-type allele of the mutant gene. The level of the gene transcript is regulated by iron in the same was as the reductase activity. The ferrous ion product of the reductase must traverse the plasma membrane. A high-affinity (Km = 5 microM) ferrous uptake system is present in both wild-type and mutant cells. Thus, iron uptake in S. cerevisiae is mediated by two plasma membrane components, a reductase and a ferrous transport system.

The first and fourth upstream open reading frames in GCN4 mRNA have similar initiation efficiencies but respond differently in translational control to change in length and sequence.

The third and fourth AUG codons in GCN4 mRNA efficiently repress translation of the GCN4-coding sequences under normal growth conditions. The first AUG codon is approximately 30-fold less inhibitory and is required under amino acid starvation conditions to override the repressing effects of AUG codons 3 and 4. lacZ fusions constructed to functional, elongated versions of the first and fourth upstream open reading frames (URFs) were used to show that AUG codons 1 and 4 function similarly as efficient translational start sites in vivo, raising the possibility that steps following initiation distinguish the regulatory properties of URFs 1 and 4. In accord with this idea, we observed different consequences of changing the length and termination site of URF1 versus changing those of URFs 3 and 4. The latter were lengthened considerably, with little or no effect on regulation. In fact, the function of URFs 3 and 4 was partially reconstituted with a completely heterologous URF. By contrast, certain mutations that lengthen URF1 impaired its positive regulatory function nearly as much as removing its AUG codon did. The same mutations also made URF1 a much more inhibitory element when it was present alone in the mRNA leader. These results strongly suggest that URFs 1 and 4 both function in regulation as translated coding sequences. To account for the phenotypes of the URF1 mutations, we suggest the most ribosomes normally translate URF1 and that the mutations reduce the number of ribosomes that are able to complete URF1 translation and resume scanning downstream. This effect would impair URF1 positive regulatory function if ribosomes must first translate URF1 in order to overcome the strong translational block at the 3'-proximal URFs. Because URF1-lacZ fusions were translated at the same rate under repressing and derepressing conditions, it appears that modulating initiation at URF1 is not the means that is used to restrict the regulatory consequences of URF1 translation to starvation conditions.

Truncated protein phosphatase GLC7 restores translational activation of GCN4 expression in yeast mutants defective for the eIF-2 alpha kinase GCN2.

GCN2 is a protein kinase in Saccharomyces cerevisiae that is required for increased expression of the transcriptional activator GCN4 in amino acid-starved cells. GCN2 stimulates GCN4 synthesis at the translational level by phosphorylating the alpha subunit of eukaryotic translation initiation factor 2 (eIF-2). We identified a truncated form of the GLC7 gene, encoding the catalytic subunit of a type 1 protein phosphatase, by its ability to restore derepression of GCN4 expression in a strain containing the partially defective gcn2-507 allele. Genetic analysis suggests that the truncated GLC7 allele has a dominant negative phenotype, reducing the level of native type 1 protein phosphatase activity in the cell. The truncated form of GLC7 does not suppress the regulatory defect associated with a gcn2 deletion or a mutation in the phosphorylation site of eIF-2 alpha (Ser-51). In addition, the presence of multiple copies of wild-type GLC7 impairs the derepression of GCN4 that occurs in response to amino acid starvation or dominant-activating mutations in GCN2. These findings suggest that the phosphatase activity of GLC7 acts in opposition to the kinase activity of GCN2 in modulating the level of eIF-2 alpha phosphorylation and the translational efficiency of GCN4 mRNA. This conclusion is supported by biochemical studies showing that the truncated GLC7 allele increases the level of eIF-2 alpha phosphorylation in the gcn2-507 mutant to a level approaching that seen in wild-type cells under starvation conditions. The truncated GLC7 allele also leads to reduced glycogen accumulation, indicating that this protein phosphatase is involved in regulating diverse metabolic pathways in yeast cells.