Identification of GCD14 and GCD15, Novel Genes Required for Translational Repression of GCN4 mRNA in Saccharomyces cerevisiae

Corresponding author: Mercedes Tamame, Instituto de Microbiología Bioquímica del C.S.I.C./Universidad de Salamanca, Edificio Departamental de Biología, Campus Miguel de Unamuno, 37007, Salamanca, Spain, tamame{at}gugu.usal.es (E-mail).

Communicating editor: A. P. MITCHELL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In Saccharomyces cerevisiae, expression of the transcriptional activator GCN4 increases at the translational level in response to starvation for an amino acid. The products of multiple GCD genes are required for efficient repression of GCN4 mRNA translation under nonstarvation conditions. The majority of the known GCD genes encode subunits of the general translation initiation factor eIF-2 or eIF-2B. To identify additional initiation factors in yeast, we characterized 65 spontaneously arising Gcd^- mutants. In addition to the mutations that were complemented by known GCD genes or by GCN3, we isolated mutant alleles of two new genes named GCD14 and GCD15. Recessive mutations in these two genes led to highly unregulated GCN4 expression and to derepressed transcription of genes in the histidine biosynthetic pathway under GCN4 control. The derepression of GCN4 expression in gcd14 and gcd15 mutants occurred with little or no increase in GCN4 mRNA levels, and it was dependent on upstream open reading frames (uORFs) in GCN4 mRNA that regulate its translation. We conclude that GCD14 and GCD15 are required for repression of GCN4 mRNA translation by the uORFs under conditions of amino acid sufficiency. The gcd14 and gcd15 mutations confer a slow-growth phenotype on nutrient-rich medium, and gcd15 mutations are lethal when combined with a mutation in gcd13. Like other known GCD genes, GCD14 and GCD15 are therefore probably required for general translation initiation in addition to their roles in GCN4-specific translational control.

THE GCN4 protein of Saccharomyces cerevisiae is a transcriptional activator of more than 50 genes involved in the biosynthesis of 12 different amino acids. Transcription of those genes is stimulated in response to amino acid or purine starvation because the rate of GCN4 synthesis increases under these conditions (HINNEBUSCH 1988 ; HINNEBUSCH and LIEBMAN 1991 ; ROLFES and HINNEBUSCH 1993 ). GCN4 expression is regulated by the intracellular levels of amino acids through a unique translational control mechanism that involves four short open reading frames (uORFs) in the GCN4 mRNA leader (MUELLER and HINNEBUSCH 1986 ; ABASTADO et al. 1991A , ABASTADO et al. 1991B ; HINNEBUSCH 1992 ) and on trans-acting factors encoded by GCN and GCD genes. Repression of GCN4 mRNA translation by the uORFs when amino acids are abundant is dependent on the GCD factors, whereas positive regulators encoded by GCN1, GCN2, and GCN3 are required to overcome translational repression by the uORFs in amino acid–starved cells (HINNEBUSCH and FINK 1983 ; HINNEBUSCH 1985 ; HARASHIMA and HINNEBUSCH 1986 ).

Recessive mutations in GCD1, GCD2 (GCD12) GCD10, GCD11, and GCD13 were isolated as suppressors of the inability of gcn2 gcn3 double mutants to derepress GCN4 and its target genes in the histidine biosynthetic pathway (HARASHIMA and HINNEBUSCH 1986 ). Based on this genetic interaction, it was concluded that these GCD factors function downstream of GCN2 and GCN3 in regulating GCN4 expression (HINNEBUSCH 1985 ). The same conclusion was reached for GCD6 and GCD7 (BUSHMAN et al. 1993A , BUSHMAN et al. 1993B ). Mutations in all the GCD genes lead to an unconditional slow-growth (Slg^-) phenotype or temperature sensitivity at 37° on nutrient-rich medium, suggesting that the GCD factors have essential functions in addition to their roles in regulating GCN4 expression (HINNEBUSCH 1992 ). This conclusion was confirmed for GCD1, GCD2, GCD6, GCD7, GCD10, and GCD11 by showing that deletions of the genes are lethal (HILL and STRUHL 1988 ; FOIANI et al. 1991 ; BUSHMAN et al. 1993A ; HINNEBUSCH and FINK 1983 ; HANNIG and HINNEBUSCH 1988 ; HANNIG et al. 1993 ; GARCIA-BARRIO et al. 1995 ). Biochemical studies of the GCD factors have shown that GCD11 is the subunit of translation initiation factor 2 (eIF2; HANNIG et al. 1993 ), and that GCD1, GCD2, GCD6, and GCD7 are four essential subunits of eIF-2B, the guanine nucleotide exchange factor for eIF2 (CIGAN et al. 1991 ; BUSHMAN et al. 1993A ; CIGAN et al. 1993 ). Mutations affecting the and ß subunits of eIF-2 (encoded by SUI2 and SUI3, respectively; DONAHUE and CIGAN 1988 ; CIGAN et al. 1989 ) also have a Gcd^- phenotype (WILLIAMS et al. 1989 ). These and other findings (TZAMARIAS et al. 1989 ) led to the idea that a reduction in eIF-2 activity is required for increased translation of GCN4 mRNA, and that GCN1, GCN2, and GCN3 would function as positive regulators of GCN4 by antagonizing eIF-2 function in amino acid–starved cells.

eIF2 functions in translation initiation by forming a ternary complex with GTP and methionyl-initiator tRNA^Met (Met-tRNA_i^Met), which binds to the 40S ribosomal subunit in one of the first steps of the initiation pathway. Upon dissociation from the initiation complex, the GTP on eIF2 is hydrolyzed to GDP. The 5 subunit eIF2B complex is required to recycle eIF2.GDP to eIF2.GTP, allowing regeneration of the ternary complex for a new round of translation initiation. In mammalian cells, phosphorylation of the subunit of eIF2 on serine-51 converts it from a substrate to an inhibitor of eIF2B, and the ensuing inhibition of GDP-GTP exchange on eIF2 inhibits regeneration of the ternary complex (JACKSON 1991 ; MERRICK 1992 ). This occurs under conditions of starvation or stress as a means of downregulating total protein synthesis in mammalian cells (HERSHEY 1989 ; HINNEBUSCH 1994A ).

GCN2 is a protein kinase that phosphorylates the subunit of eIF2 on Ser-51 when yeast cells are starved of an amino acid, and this event is required for the increase in GCN4 mRNA translation under these conditions (DEVER et al. 1992 ). GCN3 is a nonessential subunit of eIF2B, and its removal from the cell uncouples GCN4 translation from the phosphorylation of eIF2 by GCN2 (HANNIG and HINNEBUSCH 1988 ; DEVER et al. 1993 ). This last finding has provided strong evidence that phosphorylation of eIF2 stimulates GCN4 translation by downregulating eIF2B function. Additional evidence for this conclusion came from the fact that overexpressing all of the subunits of eIF2B or of eIF2 also impairs derepression of GCN4 translation. Thus, GCN4 translation is inversely coupled to the level of eIF2.GTP.Met-tRNA_i^Met ternary complexes (DEVER et al. 1995 ). Presumably, gcd mutations in subunits of eIF2 or eIF2B constitutively derepress GCN4 translation because they diminish or impair ternary complex formation in the absence of eIF2 phosphorylation by GCN2.

The following model has been proposed to explain how a reduction in eIF2.GTP.Met-tRNA _i^Met ternary complex levels leads to increased translation of GCN4 mRNA (ABASTADO et al. 1991A ; DEVER et al. 1995 ). It appears that essentially all ribosomes translate the first uORF1 and continue scanning downstream in the mRNA leader (GRANT and HINNEBUSCH 1994 ; GRANT et al. 1995 ). Under conditions of amino acid sufficiency, all of these ribosomes (or more likely, 40S subunits) rebind the ternary complex before reaching uORFs 3 and 4, and reinitiate at one of these sites. After translation of uORFs 2–4, ribosomes dissociate from the mRNA and cannot reinitiate at the GCN4 AUG codon. Under starvation conditions, ternary complex levels are reduced by phosphorylation of eIF2 by GCN2. After translating uORF1, a substantial fraction of ribosomes fail to rebind the ternary complex before reaching uORFs 2–4 and, thus, cannot reinitiate at these sites. These ribosomes rebind the ternary complex after scanning past uORF4, and they reinitiate translation at the GCN4 AUG codon instead.

It has been predicted that other initiation factors besides eIF2 and eIF2B would be involved in the process of reinitiation on GCN4 mRNA, and that mutations in these factors could be isolated on the basis of conferring a Gcn^- or Gcd^- phenotype (HINNEBUSCH 1996 ). Because the GCD genes isolated thus far are all essential and their products have a function required for general translation intitation, the identification of novel GCD genes could uncover additional components of the yeast translational apparatus. For example, it was recently suggested that gcd10 mutations reduce the ability of eIF3 to stimulate rebinding of the ternary complex, or that they impair rebinding of eIF3 itself, to the 40S ribosomal subunits scanning downstream from uORF1. This would mimic the effect of reducing ternary complex levels in allowing ribosomes to bypass uORFs 2–4 and reinitiate at GCN4 instead, thereby accounting for the Gcd^- phenotype of gcd10 mutants (GARCIA-BARRIO et al. 1995 ).

With the aim of identifying additional factors involved in reinitiation on GCN4 mRNA, we have characterized 65 new Gcd^- mutants isolated in the same genetic selection that yielded mutations in GCD10, GCD11, GCD1, GCD2, and GCD13 (HARASHIMA and HINNEBUSCH 1986 ). Analysis of these mutants has led to the identification of two novel genes, GCD14 and GCD15, that encode factors required for the repression of GCN4 translation under conditions of amino acid sufficiency. The gcd14 and gcd15 mutations lead to slow growth or thermosensitive growth at 37°, and the gcd15 mutations are lethal when combined with gcd13-501. These results suggest that the GCD14 and GCD15 products, like all other GCD factors characterized thus far, have general functions in protein synthesis initiation in addition to their roles in GCN4 translational control.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains, media, and genetic techniques:
The experiments were performed using the yeast strains described in Table 1. The 65 Gcd^- mutants (^a strains in Table 1) were obtained previously by HARASHIMA and HINNEBUSCH 1986 but have not been characterized in detail until now. Hm strains come from genetic crosses between H strains and, with the exception of Hm47 (GARCIA-BARRIO et al. 1995 ), were all constructed here. Strain F35, used as a wild-type control in the Northern blot analysis, is a derivative of TD28 and was described previously (LUCCHINI et al. 1984 ). The GCD14-integrating plasmid pRC57 was directed to integrate at the gcd14 locus of Hm316 (gcd14-2) to produce isogenic GCD14 and gcd14 strains (R. CUESTA, unpublished results). It produced a nontandem duplication consisting of either the gcd14 and GCD14 alleles or the gcd14 and gcd14 alleles that was separated by plasmid sequences and URA3. Selecting for loss of URA3 by growing the transformants on medium containing 0,1% 5-fluoroorotic acid (5-FOA; BOEKE et al. 1987 ) yielded isogenic GCD14 (Hm316G) and gcd14 (Hm316g) strains as Slg⁺/3AT^s and Slg^-/3AT^r derivatives, respectively.

Transformation of yeast strains was performed as described (ITO et al. 1983 ). Standard genetic techniques and media were conducted according to established procedures (GUTHRIE and FINK 1991 ). Sensitivity to 3-aminotriazole (3AT) and analysis of the His^- phenotype conferred by the his1-29 leaky mutation were tested as described (HINNEBUSCH and FINK 1983 ; HARASHIMA and HINNEBUSCH 1986 ).

Dominance–recessiveness and complementation analysis of the gcd mutations were conducted by replica printing colonies to medium containing 10 mM 3AT (to score the Gcd^- phenotype) and to YPD medium at 24°, 28°, and 37° [to score the temperature-sensitive (Tsm^-) or slow-growth (Slg^-) phenotypes], all as described (HARASHIMA and HINNEBUSCH 1986 ).

Segregation of the Tsm^- and Slg^- traits in tetrads generated from different crosses was scored by streaking out cells from each spore colony on YPD medium at 24°, 28°, and 37°, and comparing the sizes of colonies that arose to those produced by the congenic strains H96 or H117.

Recombination tests to establish the number of linkage groups defined by gcd mutations in H68, H74, H160, and H168 were performed by crossing these strains, or by crossing strains of the opposite mating type that contain the same gcd mutations, H160, H168, Hm296, H74, H68, and Hm325, to one another as follows: H68 ( gcd15-1) x Hm325 (a gcd15-2), H160 (a gcd14-1) x Hm296 ( gcd14-2), H68 ( gcd15-1) x H160 (a gcd14-1), H68 ( gcd15-1) x H168 (a gcd14-2), H74 ( gcd15- 2) x H160 (a gcd14-1), and H74 ( gcd15-2) x H168 (a gcd14-2).

Recombination of gcd14 or gcd15 with gcd1, gcd2, gcd6, gcd7, and gcd10 alleles was analyzed by crossing H168 (a gcd14-2) or Hm320 (a gcd15-1) with H56 ( gcd1-501), H63 ( gcd2-502), H1792 ( gcd6-1), H1794 ( gcd7-201), and H55 ( gcd10-502), respectively. Recombination with gcd11 and gcd13 was determined in crosses of Hm295 ( gcd14-1) or H68 ( gcd15-1) with H236 (a gcd11-505) and H248 (a gcd13-502), respectively. Meiotic segregation of the 3AT^r and Slg^- and Tsm^- traits in tetrads obtained from all these crosses was scored as described above.

Plasmids:
All plasmids used in this study are described in Table 2. Cloned genes used in complementation tests of Gcd^- mutants were contained in low copy number yeast vectors YCp50 (JOHNSTON and DAVIS 1984 ) or pRS316 (SIKORSKI and HIETER 1989 ). The GCN4-lacZ fusion on plasmid p180 contains the wild-type leader with all four uORFs intact. Plasmid p227 is identical to p180, except that point mutations remove the ATG codons of all four uORFs; plasmid p226 is identical to p227 except that uORF4 remains intact (HINNEBUSCH 1985 ; MUELLER and HINNEBUSCH 1986 ). Plasmid DNA purification and restriction enzyme digestion, as well as the method of random priming for radiolabeling of DNA, were conducted according to established procedures (SAMBROOK et al. 1989 ). Isolation of DNA fragments from low melting agarose gels was carried out using the Wizard PCR Preps System (Promega, Madison, WI) following the vendor's protocol.

RNA analysis:
Preparation of yeast total RNA and RNA blot hybridization were carried out as described previously (HINNEBUSCH 1985 ). A 500-bp NdeI-PstI fragment harboring the 3' end of the HIS3 gene was used as the radiolabeled probe for HIS3mRNA, and a 6.7-kbp HindIII fragment containing the entire pyruvate kinase–coding sequence (PYK1) was used as the probe for PYK1 mRNA. A 450-bp KpnI-MluI fragment was used as the probe for GCN4 mRNA.

Enzymatic assays of HIS4-lacZ and GCN4-lacZ expression:
Expression of ß-galactosidase activity from an integrated HIS4-lacZ allele or from plasmid-borne GCN4-lacZ constructs was assayed as described previously (LUCCHINI et al. 1984 ; MOEHLE and HINNEBUSCH 1991 ). For repressing conditions, saturated cultures were diluted 1:50 and harvested in midlogarithmic phase after 6 h of growth. For derepressing conditions, cultures were grown for 2 h under repressing conditions and then for 6 h after the addition of 3-AT to 10 mM. The values shown are the averages of two independent determinations. ß-Galactosidase activities are expressed as nanomoles of o-nitrophenyl-ß-D-galactopyranoside hydrolized per minute per milligram of protein.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genetic characterization of Gcd^- mutants:
A genetic analysis of 65 Gcd^- mutants was carried out to identify novel GCD genes. These mutants were isolated as spontaneous revertants of the Gcn^- phenotype of two strains of opposite mating type (H96 and H117, Table 1) bearing the mutations his1-29, gcn2-101, and gcn3-101 (HARASHIMA and HINNEBUSCH 1986 ). Mutations in GCN2 and GCN3 prevent increased expression of GCN4 and the amino acid biosynthetic genes under its control in response to amino acid limitation. Consequently, Gcn^- mutants are hypersensitive to 3-AT, an inhibitor of the histidine biosynthetic enzyme encoded by HIS3 (HINNEBUSCH 1988 ). his1-29 is a leaky mutation that only partially impairs the histidine biosynthetic enzyme encoded by HIS1. In GCN strains, derepression of his1-29 transcription by GCN4 provides sufficient HIS1 function to allow growth on medium lacking histidine (His⁺). Because transcription of his1-29 cannot be derepressed in Gcn^- mutants, gcn his1-29 double mutants are His^- (WOLFNER et al. 1975 ). The Gcd^- mutants were isolated as spontaneous revertants of both the His^- and the 3AT^s phenotypes of strains H96 and H117. The presence of two gcn mutations in these strains diminished the risk of obtaining GCN back-mutations rather than unlinked suppressors.

The strength of the Gcd^- phenotype of each mutant was assessed by replica-plating cells to minimal medium lacking histidine at 30°, as well as to medium lacking histidine and containing 10 mM 3AT, and incubating for 3 days at 24° or 28°. The His and 3AT phenotypes of the parental strains and the gcd14 and gcd15 mutants that we eventually identified are depicted in Figure 1A.

View larger version (64K): In this window In a new window Download PPT slide   Figure 1. —Phenotypic traits of several Gcd- mutants characterized in this work. (A) The 3ATr and His+ phenotypic traits of Gcd- mutants H68 (gcd15-1), H74 (gcd15-2), and the previously described H75 mutant (gcd13-501; HARASHIMA and HINNEBUSCH 1986 ) are shown relative to their isogenic GCD parent H96 in the three top panels. The same phenotypic traits of strains H160 (gcd14-1), H168 (gcd14-2), and H275 (bearing a dominant gcd mutation) are compared to their isogenic parent H117 in the three bottom panels. Isolated colonies from each strain were replica printed to 10-mm 3AT plates and incubated for 3 days at 28°. The SD-His and SD+His 0.3-mm plates were incubated for 1 day at 30°. (B) Slow-growth phenotypes (Slg-) of the Gcd- mutants H68 (gcd15-1), H74 (gcd15-2), and H75 (gcd13-501) are shown relative to the corresponding parental strain H96 (GCD), and that of Hm295 (gcd14-1), Hm296 (gcd14-2), and H275 (gcd) are shown relative to parental strain H117(GCD). Strain H75 is Slg- and thermosensitive at 37° (Tsm-). Cells from all those strains were streaked for single colonies on YPD medium and incubated 3 days at 28° or 37°. The Hm295 and Hm296 strains are not isogenic to the GCD strain H117, but we have verified these results by comparing the growth rates of transformants of these gcd14 mutant strains bearing cloned GCD14 or vector alone (R. Cuesta, A. G. Hinnebusch and M. Tamame, unpublished results).

Most gcd mutations lead to temperature-sensitive growth or unconditional slow growth on nutrient-rich medium (HINNEBUSCH 1992 ). Therefore, we analyzed this trait by replica-plating and by streaking cells for individual colonies on YPD medium and incubating for 3 days at 24°, 28°, and 37°. This way, it was possible to distinguish among mutants defective for growth only at 37° (Tsm^- phenotype) and those that grow poorly at all temperatures (Slg^- phenotype). The Slg^- phenotype of the gcd14 and gcd15 mutants at 28° and 37° is presented in Figure 1B.

On the basis of these tests, we found that 36 of the Gcd^- mutants had strong 3AT^r and His⁺ phenotypes and also Tsm^- or Slg^- phenotypes, and thus were likely to contain gcd mutations. This class of mutants contained H60, H65, H67, H68, H69, H72, H74, H76, and H88, derived from H96 and H164, H165, H166, H168, H169, H170, H171, H172, H173, H174, H177, H234, H235, H237, H239, H241, H242, H246, H259, H261, H262, H263, H264, H268, H269, H274, and H275, derived from H117. Another five mutants (H66, H77, H78, H243, and H244) had strong 3AT^r and His⁺ phenotypes, but only slight Slg^- phenotypes, and eight strains (H80, H85, H86, H160, H175, H240, H266, and H270) were 3-AT^r and His⁺ but had no detectable Slg^- or Tsm^- phenotype. The remaining 16 mutants did not have strong 3AT^r and His⁺ phenotypes and were not analyzed further.

Dominance–recessiveness of Gcd^- mutants:
To determine whether the gcd mutations in the 49 strains selected for further analysis were dominant or recessive, we constructed diploids between each Gcd^- mutant and the parental strain of opposite mating type (H96 or H117). The formation of diploids was confirmed by inducing sporulation and observing the appearance of asci. The diploids were homozygous for gcn2-101 and gcn3-101 and were heterozygous for the gcd mutations; therefore, dominance or recessiveness was assessed by determining the 3AT^r and His⁺ phenotypes of the diploids, as described above. Only strain H275 was found to contain a gcd mutation with dominant 3AT^r and His⁺ phenotypes. Nine other mutants (H69, H76, H163, H239, H240, H241, H244, H261, and H262) appeared to contain gcd alleles that are semidominant for these phenotypes, and the remaining 39 mutants appeared to contain recessive gcd alleles.

Measurements of HIS4 transcription in Gcd^- mutants:
Under amino acid starvation conditions, GCN4 activates transcription of the HIS4 gene (HINNEBUSCH 1988 ). Gcd^- mutants exhibit high levels of HIS4 transcription under all conditions of amino acid availability because they are constitutively derepressed for GCN4 expression. In previous studies, gcd mutations were found to increase expression of a HIS4-lacZ fusion between five and 40 times in cells grown under nonstarvation conditions on minimal medium (HARASHIMA and HINNEBUSCH 1986 ). To provide a quantitative estimate of the Gcd^- phenotype in our mutants, we assayed the expression of the HIS4-lacZ fusion integrated at the ura3-52 locus in each of the strains after exponential growth on minimal medium. Most of the mutants contained levels of ß-galactosidase activity ranging from two–17-fold higher than that observed in the corresponding isogenic parental strain H96 or H117.

We selected for further analysis the 18 strains with ß-galactosidase activities at least five times greater than that measured in their isogenic parental strains, and their phenotypes are summarized in Table 3. Addition of 3AT to the growth medium produced a three- to four-fold increase in HIS4-lacZ expression in the GCN GCD strain F35. Under the same conditions, we observed little or no increase in HIS4-lacZ expression in the Gcd^- mutants, indicating that HIS4 transcription is constitutively derepressed in these strains (data not shown).

Meiotic segregation of the Gcd^- phenotype and complementation analysis with cloned genes:
We wished to determine whether the Gcd^- phenotypes of the 18 selected mutants resulted from mutations in single genes. To do so, we formed diploids with the parental strain of opposite mating type (H96 or H117), induced the diploids to sporulate, and performed tetrad analysis on the resulting spores. For 16 of the 18 mutants, the 3AT^r, His⁺, and Slg^- phenotypes segregated 2⁺:2^- in all tetrads, indicating the presence of a single gcd mutation. The crosses involving H169 and H244 did not exhibit 2⁺:2^- segregation for the 3AT^r phenotype, and these mutants were not analyzed further.

Complementation tests were performed with the remaining 15 mutants bearing recessive gcd mutations by transforming each strain with low copy number plasmids (Table 2) containing the following genes, which have been previously implicated in GCN4 translational control: GCN1, GCN2, GCN3, GCD1, GCD2, GCD5, GCD6, GCD7, GCD10, GCD11, SUI2, and SUI3. The resulting transformants were scored for their 3AT^r and His⁺ (Gcd^-) phenotypes as described above, and the results are summarized in Table 4. The Gcd^- phenotypes of strains H72, H88, and H242 were fully complemented by plasmid-borne GCD6, which encodes the subunit of eIF2B (BUSHMAN et al. 1993A ). To provide additional evidence that these three mutants contain a recessive gcd6 mutation, we carried out a complementation analysis with gcd6-1 strains. Two kinds of diploids were constructed by crossing H72, H88, and H242 to a gcn2 gcd6-1 strain (H1916 or H1917) or to the parental gcn2 gcn3 GCD6 strain of opposite mating type (H96 or H117). The diploids constructed with the gcd6-1 strain all displayed a Gcd^- phenotype, whereas the diploids constructed with the GCD6 strain all had a Gcn^- phenotype. Considering that all three mutants were complemented by wild-type GCD6, but failed to complement a gcd6-1 strain, we concluded that each contains a recessive gcd6 mutation.

Interestingly, the gcd6 mutations in H72, H88, and H242 were partially suppressed by complementing the gcn3-101 mutation in each strain with plasmid-borne GCN3 (Table 4), which encodes the nonessential subunit of eIF2B (CIGAN et al. 1991 , CIGAN et al. 1993 ). Similar observations have been made previously for certain gcd1 and gcd2 mutations that affect the and subunits of eIF2B, respectively, which have a Gcd^- phenotype in a gcn3 background but are fully or partially suppressed by the presence of wild-type GCN3 (HARASHIMA et al. 1987 ; PADDON and HINNEBUSCH 1989 ). To explain these interactions, it was suggested that such gcd mutations lead to substantial reductions in eIF2B function only when GCN3 is absent from the eIF2B complex. Similarly, the gcd6 mutations in strains H72, H88, and H242 may alter eIF2B structure or function in a way that can be corrected by having GCN3 present in the eIF2B complex.

The Gcd^- phenotype of H239 was fully complemented by plasmid-borne GCN3 and was partially complemented by plasmid-borne GCD2 or GCD7 (Table 4). Loss-of-function GCN3 mutations, such as gcn3 or gcn3-101, render the eIF2B complex insensitive to inhibition by phosphorylated eIF2 and prevent derepression of GCN4 (Gcn^- phenotype). Because H239 has a Gcd^- phenotype, we suspected that it contains a gcn3^c mutation that leads to a reduction in eIF2B function with attendant derepression of GCN4 (HANNIG et al. 1990 ). This possibility was supported by the fact that crossing H239 with the gcn3^c-R104K strain H1489 produced diploids with a Gcd^- phenotype, whereas the diploids formed by crossing H239 with parental gcn3-101 strain H96 had a Gcn^- phenotype. The fact that a second copy of GCD7 or GCD2 partially suppressed the Gcd^- phenotype of H239 could be explained by proposing that the presumptive gcn3^c mutation in H239 destabilizes eIF2B and causes dissociation of the GCD2 and GCD7 subunits from a fraction of the eIF2B complexes in the mutant cells. Increasing the dosage of GCD2 or GCD7 could then partially offset the effects of the gcn3^c mutation and increase the formation of intact eIF2B complexes. In accordance with this explanation, it was shown recently that GCD2, GCD7, and GCN3 interact with one another and comprise a regulatory domain in eIF2B that mediates the inhibition of guanine nucleotide exchange activity by phosphorylated eIF2 (YANG and HINNEBUSCH 1996 ).

Among the remaining 11 Gcd^- mutants, four were complemented by plasmid-borne GCD11 and three were complemented by plasmid-borne GCD10, suggesting that they contain recessive mutations in these two genes. Interestingly, none of the cloned genes were able to complement the gcd mutations in strains H68, H74, H160, and H168, suggesting that they contain mutations in novel GCD genes.

Genetic identification of the GCD14 and GCD15 genes:
To confirm that strains H68, H74, H160, and H168 contain mutations in novel GCD genes and to determine the number of new GCD genes involved, we crossed these strains (or strains of opposite mating type containing the same gcd mutations) to one another and to strains containing known gcd1, gcd2, gcd6, gcd7, gcd10, gcd11, or gcd13 alleles, and we conducted tetrad analysis of the meiotic products (see MATERIALS AND METHODS). All the diploids produced in these crosses were homozygous for his1-29, gcn2, and gcn3 mutations; therefore, the GCD ascospores derived from these diploids will be His^- and 3AT^s, whereas the gcd spores will be AT^r and His⁺. Crosses between strains bearing allelic gcd mutations should produce tetrads in which all four spores are AT^r His⁺ (parental ditypes). Crosses involving nonallelic gcd mutations should produce tetratypes (three ascospores AT^r:1AT^s), nonparental ditypes (2 AT^r:2AT^s), and parental ditypes (4AT^s:0AT^s). The results of our analysis showed that the gcd mutations in H68, H74, H160, and H168 recombined with gcd1, gcd2, gcd6, gcd7, gcd10, gcd11, or gcd13 mutations, confirming that they are not alleles of these known GCD genes. The mutations in H160 and H168 did not recombine with one another, but they recombined with those present in H68 and H74; accordingly, they were designated gcd14-1 (H160) and gcd14-2 (H168). The mutations in H68 and H74 did not recombine with one another, and they were designated gcd15-1 (H68) and gcd15-2 (H74).

The gcd13-501 gcd15 combination is synthetically lethal:
From the genetic analysis just described, we obtained evidence that gcd13 gcd15 double mutants are inviable. The Slg^- phenotype of gcd13-501 is more drastic than that of gcd15-1 and gcd15-2 at 28°, and gcd13-501 is temperature sensitive for growth at 37° (Tsm^-). Therefore, we can distinguish gcd13-501 and gcd15 ascospores from one another. In the crosses between the gcd13-501 and gcd15 strains summarized in Table 5, we found that all of the four-spored tetrads were parental ditypes containing two 3AT^r gcd13-501 GCD15 spores and two 3AT^r GCD13 gcd15 spores. The most frequent tetrads, expected to be tetratypes, contained a 3AT^r gcd13-501 GCD15 spore, a 3AT^r GCD13 gcd15 spore, and a 3AT^s GCD13 GCD15 spore; thus, the gcd13-501 gcd15 spore was always inviable. Finally, two-spored tetrads containing two 3AT^s GCD13 GCD15 spores were obtained, again indicating that the doubly mutant ascospores were inviable.

A gcn4 mutation is epistatic to gcd14 and gcd15:
The fact that gcd14 and gcd15 alleles were isolated as suppressors of the 3AT^s phenotype of gcn2-101 gcn3-101 double mutants indicates that mutations in GCD14 or GCD15 can restore high-level transcription of genes under GCN4 control in strains lacking translational activators of GCN4 expression. The simplest explanation for this finding is that the gcd14 and gcd15 alleles, like other gcd mutations, lead to constitutive derepression of GCN4 expression. Alternatively, it was possible that they increase transcription of HIS3 and HIS4 by a mechanism that is independent of GCN4. In an effort to rule out the latter possibility, we determined the effect of inactivating GCN4 on the 3AT^r phenotype of gcd14 and gcd15 mutants. This was accomplished by analyzing the crosses summarized in Table 6 between gcn2-101 gcn3-101 gcd14 or gcn2-101 gcn3-101 gcd15 strains and a gcn2-101 gcn3-101 gcn4-103 strain. If the Gcd^- phenotype of the gcd14 and gcd15 alleles is independent of GCN4, then a 2AT^r:2AT^s segregation in the tetrads should be observed. If the gcn4 mutation is epistatic to the gcd mutations, then 1AT^r:3AT^s and 0AT^r:4AT^s tetrads should be produced in addition to 2AT^r:2AT^s tetrads, and the 1AT^r:3AT^s tetrads should be the predominant class. The results in Table 6 indicate that both gcd 14 and gcd15 mutations are hypostatic to gcn4-103.

To confirm that the gcn4 mutation masks the Gcd^- phenotype of the gcd14 and gcd15 mutations, we transformed the gcn2-101 gcn3-101 gcn4-103 gcd14-1 (or gcd15-2) ascospores obtained from these crosses with plasmid p164 containing GCN4 (Table 2). In all cases, introduction of wild-type GCN4 restored the 3AT^r and His⁺ phenotypes of the gcd14 and the gcd15 mutations, whereas gcn2-101 gcn3-101 gcn4-103 GCD14 (or GCD15) spores transformed with p164 retained their Gcn^- phenotype.

gcd14 and gcd15 mutations lead to constitutive derepression of GCN4 translation:
To obtain direct evidence that GCD14 and GCD15 repress GCN4 expression under nonstarvation conditions, we measured ß-galactosidase expression from a GCN4-lacZ fusion born on plasmid p180 in gcd14 and gcd15 strains. This fusion construct contains the wild-type GCN4 mRNA leader containing all four uORFs and thus exhibits wild-type translational control of GCN4 expression (MUELLER and HINNEBUSCH 1986 ). As shown in Figure 2, low-level expression of the GCN4-lacZ fusion on p180 was observed in the GCD14 and GCD15 strains under nonstarvation (repressing, R) and histidine starvation (derepressing, DR) conditions because the gcn2 allele in these strains impairs derepression of GCN4 translation (HINNEBUSCH 1985 ). The gcd14-2 and gcd15-1 mutations led to GCN4-lacZ expression at levels roughly fivefold (gcd14-2) or 17–20-fold (gcd15-1) greater than that observed in the correspondingGCD strains. These results demonstrate that GCD14 and GCD15 are required for repression of GCN4 expression under nonstarvation conditions.

View larger version (20K): In this window In a new window Download PPT slide   Figure 2. —The gcd14-2 and gcd15-1 mutations lead to constitutive derepression of GCN4-lacZ translation independently of the positive regulator GCN2. GCN4-lacZ fusions were introduced into yeast strains on low copy number plasmids p180, p226, and p227. The relevant genotypes of the strains are shown on the left. The GCD14 and gcd14-2 strains (Hm316G and Hm316g, respectively) are isogenic, and the GCD15 and gcd15-1 strains (Hm348 and Hm347, respectively) are congenic. The four uORFs in the leader sequence of p180 are shown as open boxes () and point mutations that remove the AUG codons of uORFs 1–3 (p226) or 1–4 (p227) are shown as Xs. ß-Galactosidase activity was measured in cells grown to midlogarithmic phase under nonstarvation, repressing (R) conditions, or derepressing (DR) conditions of histidine starvation induced by 3AT.

Efficient repression of GCN4 translation requires the presence of uORFs 3 or 4, and derepression in response to amino acid starvation also requires uORF1. Consequently, GCN4-lacZ expression is very high and unregulated from a construct lacking all four uORFs, and it is low and unregulated from a construct containing uORF4 alone (MUELLER and HINNEBUSCH 1986 ). If GCD14 and GCD15 repress GCN4 expression at the translational level via the uORFs, then elimination of all four uORFs should abolish the derepressing effects of gcd14 and gcd15 mutations on GCN4-lacZ expression. The same result is expected upon removal of uORFs 1–3, leaving only uORF4 intact. In contrast, if GCD14 and GCD15 downregulate the synthesis or stability of GCN4 mRNA rather than repress its translation, then mutations in these genes should increase GCN4-lacZ expression independently of the uORFs. As shown in Figure 2, the gcd14 and gcd15 mutations had little or no effect on the expression of GCN4-lacZ fusions lacking all four uORFs (p227) or containing uORF4 alone (p226). These results indicate that GCD14 and GCD15 regulate GCN4 expression at the translational level via the uORFs. In accordance with this last conclusion, we found that the steady-state amounts of GCN4 mRNA are roughly equivalent in gcd14 and gcd15 mutants compared to the corresponding wild-type GCD strains under both nutritional conditions (Figure 3A). This result is in sharp contrast to the strong derepressing effects of these mutations on the levels of HIS3 mRNA, a target of GCN4 transcriptional activation (Figure 3B).

View larger version (35K): In this window In a new window Download PPT slide   Figure 3. —gcd14 and gcd15 mutations lead to increased transcription of HIS3 without alterations in GCN4-mRNA abundance. (A) Isogenic GCD14 and gcd14-2 strains (Hm316G and Hm316) and congenic GCD15 and gcd15-1 strains (Hm348 and Hm347) were grown under repressing (R) nonstarvation conditions or derepressing (DR) conditions of histidine starvation induced by 3AT. Total RNA was extracted and analyzed by RNA blot hybridization with radiolabeled probes for GCN4 and PYK1 mRNAs. The latter was examined to control for the amounts of RNA loaded in each lane. (B) RNAs were obtained from cells grown under the same conditions as described in A, from strains H96, H68, H74, H117, H160, H168, and the control wild-type strain F35 (from left to right). Their relevant genotypes are indicated at the top of each pair of lanes. RNAs were analyzed with radiolabeled probes for HIS3 and PYK1 mRNAs.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

GCN4 is a transcriptional activator of amino acid and purine biosynthetic genes, and its synthesis is regulated by amino acid availability through an intricate translational control mechanism (HINNEBUSCH 1994B , HINNEBUSCH 1996 ). Phosphorylation of eIF2 by GCN2, with attendant reduction in the level of the eIF2.GTP.Met-tRNA_i^Met ternary complex, leads to derepression of GCN4 translation in amino acid–starved cells. The GCN4 protein that is produced induces transcription of amino acid and purine biosynthetic genes, alleviating the nutrient starvation that triggers eIF2 phosphorylation in yeast. GCN4-specific translational regulation has proven to be exquisitely sensitive to alterations in the amounts or activities of translation initiation factors eIF2 and eIF2B. GCD genes, which encode several of the subunits of these factors, were first identified by recessive mutations that constitutively derepress GCN4 translation. It was predicted that the same genetic selection that yielded mutations in eIF2 and eIF2B subunits would uncover novel genes whose products have an essential function in translation initiation (HINNEBUSCH 1996 ). Given the large degree of sequence and functional conservation between the yeast and mammalian translational apparatus (MERRICK and HERSHEY 1996 ), the identification of novel yeast factors should be highly relevant to mammalian systems.

In this work, we characterized 65 Gcd^- mutants isolated previously as suppressors of the 3AT sensitivity (Gcn^- phenotype) of gcn2 gcn3 double mutants (S. HARASHIMA and A. G. HINNEBUSCH, unpublished results). Forty-eight of the 65 mutants displayed a strong Gcd^- phenotype, as judged by growth on medium containing 3AT. Among these, 40 also had a slow-growth phenotype on rich medium at all temperatures; 11 were temperature-sensitive at 37°, but only three (H171, H234, and H235) were both Slg^- and Tsm^- at the same time. Only one of the mutants (H275) had a dominant Gcd^- phenotype. Dominant mutations have been isolated in GCD11 (HANNIG et al. 1993 ); however, we have found that H275 does not contain a mutation in GCD11 (data not shown). The recessive nature of the majority of the gcd mutations indicates that they lack a gene product or function required for repression of genes subjected to GCN4 control under nonstarvation conditions. The Slg^- and Tsm^- phenotypes of Gcd^- mutants bearing these traits are also recessive, suggesting that the gene products defective in these mutants are required for an essential function.

By assaying HIS4-lacZ expression, we identified 17 recessive Gcd^- mutants in which HIS4 transcription was derepressed under nonstarvation conditions by a factor of five or more. Of these, 15 showed monogenic segregation of the Gcd^- phenotype, and these were subjected to complementation tests using previously cloned GCN and GCD genes. Eleven of the 15 mutants were complemented by GCN3, GCD6, GCD10, or GCD11. Mutants H68, H74, H160, and H168 were not complemented by any of the known GCN or GCD genes, and they appear to contain mutations in two novel genes that we called GCD14 (H160 and H168) and GCD15 (H68 and H74). The gcd14 and gcd15 mutations are hypostatic to gcn4-103 (Table 6), consistent with a role for GCD14 and GCD15 in repressing GCN4 expression. Direct evidence for this conclusion came from the fact that expression of a GCN4-lacZ construct containing all four uORFs was constitutively derepressed in gcd14 and gcd15 strains. In contrast, expression of GCN4-lacZ constructs containing uORF4 alone or no uORFs was relatively insensitive to the gcd14 and gcd15 mutations. These results indicate that GCD14 and GCD15 repress GCN4 expression at the translational level, and that they are required for the inhibitory effects of uORFs 3 and 4 on initiation at GCN4 under nonstarvation conditions. In the context of our model for GCN4 translation control, ribosomes scanning downstream from uORF1 would fail to reinitiate at uORFs 3 and 4 in gcd14 and gcd15 mutants and would reinitiate at GCN4 instead, even when there is no phosphorylation of eIF2 by GCN2. Presumably, this occurs because the mutations reduce either the level of eIF2.GTP.Met-tRNA_i^Met ternary complexes or the rate at which ternary complexes rebind to 40S ribosomes scanning downstream from uORF1. Either defect would account for the reduced frequency of reinitiation at uORFs 3 and 4 with concommitant increase in reinitiation downstream at GCN4, which occurs in gcd14 and gcd15 mutants.

The gcd15-1 mutation appears to have a greater quantitative effect than does the gcd14-2 mutation in reducing reinitiation at uORFs 3 and 4, thereby stimulating reinitiation at GCN4. In addition, it leads to an approximately twofold increase in GCN4 translation in the presence of uORF4 alone and a twofold decrease in GCN4 translation in the absence of all four uORFs (Figure 2). These latter effects could indicate that conventional initiation events at the 5'-most initiation codons of these GCN4-lacZ alleles occur less efficiently in gcd15 mutants vs. wild-type GCD15 strains. The much larger quantitative effects of the gcd15 mutations seen when all four uORFs are present vs. uORF4 alone would imply that the efficiency of reinitiation at uORF4 after translation of uORF1 is reduced much more than are conventional initiation events at uORF4 when it is the 5'-most start site in the mRNA. We previously explained this differential effect by postulating that reinitiation has kinetic constraints that do not apply to conventional initiation at the first AUG codon on an mRNA (HINNEBUSCH 1996 ). It is believed that 40S ribosomes normally bind the ternary complex before interacting with mRNA (MERRICK and HERSHEY 1996 ). Thus, a reduction in ternary complex levels in gcd15 mutants may decrease the pool of 40S preinitiation complexes that are able to bind to the 5' ends of mRNAs, but it should not cause ribosomes already bound to the mRNA to bypass an AUG codon encountered while scanning the leader (leaky scanning). In contrast, a reduction in the rate of ternary complex binding to ribosomes scanning downstream from uORF1 should lead to considerable leaky scanning at uORFs 2–4 (with attendant stimulation of GCN4 translation) because ribosomes must rebind the ternary complex in the time it takes to scan from uORF1 to the start sites of these uORFs.

The idea that GCD14 and GCD15 are required for the formation of the eIF2.GTP.Met-tRNA_i^Met ternary complex or its binding to ribosomes is in accordance with the slow-growth and temperature-sensitive phenotypes of the gcd14 and gcd15 mutants on nutrient-rich medium. In addition, the gcd13 and gcd15 mutations showed synthetic lethality, and we found recently that gcd13 mutants exhibit aberrant polysome profiles indicative of a general defect in translation initiation (R. CUESTA, unpublished results). Thus, it seems likely that GCD13 and GCD15 are required for production of the same component or for execution of the same step of the initiation pathway. For example, they could be subunits of a novel initiation factor required for formation or utilization of the ternary complex. Alternatively, they could function in controlling the expression of eIF2, tRNA_i^Met, eIF2B, or eIF3, indirectly influencing the levels of the ternary complex and its binding to ribosomes.

We recently achieved the cloning of GCD14 and GCD15, allowing us to confirm their identities as novel GCD genes (R. CUESTA, O. CALVO, J. ANDERSON, M. T. GARCÍA-BARRIO, A. G. HINNEBUSCH and M. TAMAME, unpublished results). The association of GCD14 and GCD15 with other components of the yeast translational machinery, as well as the cloning of GCD13, are both underway. Many of the 65 Gcd^- mutants described in this report remain to be fully characterized, suggesting that the field of gcd mutations may not yet be saturated, and that novel GCD genes remain to be identified in the future.

	ACKNOWLEDGMENTS

We wish to thank FRANCISCO ANTEQUERA and CARLOS R. VÁZQUEZ DE ALDANA for helpful discussions regarding this work, and to OLGA CALVO, ROSA ESTEBAN, and NOELIA GUTIÉRREZ for thoughtful reading of the manuscript. This work was supported by grant PB94-1103 from the Spanish Dirección General de Investigación Científica y Técnica (DGICYT) awarded to M.T., and by Collaborative Research Grant 920605 from NATO awarded to A.G.H. and M.T. R.C. acknowledges support from fellowship granted by the Spanish Ministerio de Educación y Ciencia through the "Consejo Superior de Investigaciones Científicas" (CSIC).

Manuscript received June 18, 1997; Accepted for publication December 5, 1997.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ABASTADO, J. P., P. F. MILLER, B. M. JACKSON, and A. G. HINNEBUSCH, 1991a  Suppression of ribosomal reinitiation at upstream open reading frames in amino acid-starved cells forms the basis for GCN4 translational control. Mol. Cell. Biol. 11:486-496[Abstract/Free Full Text].

ABASTADO, J. P., P. F. MILLER, and A. G. HINNEBUSCH, 1991b  A quantitative model for translational control of the GCN4 gene of Saccharomyces cerevisiae.. New Biol. 3:511-524[Medline].

BOEKE, J. D., J. TRUEHEART, G. NATSOULIS, and G. R. FINK, 1987  5-Fluoroorotic acid as a selective agent in yeast molecular genetics. Methods Enzymol. 154:164-175[Medline].

BUSHMAN, J. L., A. I. ASURU, R. L. MATTS, and A. G. HINNEBUSCH, 1993a  Evidence that GCD6 and GCD7, translational regulators of GCN4, are subunits of the guanine nucleotide exchange factor for eIF-2 in Saccharomyces cerevisiae.. Mol. Cell. Biol. 13:1920-1932[Abstract/Free Full Text].

BUSHMAN, J. L., M. FOIANI, A. M. CIGAN, C. J. PADDON, and A. G. HINNEBUSCH, 1993b  Guanine nucleotide exchange factor for eukaryotic translation initiation factor 2 in Saccharomyces cerevisiae : interactions between the essential subunits GCD2, GCD6, and GCD7 and the regulatory subunit GCN3. Mol. Cell. Biol. 13:4618-4631[Abstract/Free Full Text].

CIGAN, A. M., E. K. PABICH, L. FENG, and T. F. DONAHUE, 1989  Yeast translation initiation suppressor sui2 encodes the subunit of eukaryotic initiation factor 2 and shares sequence identity with the human subunit. Proc. Natl. Acad. Sci. USA 86:2784-2788[Abstract/Free Full Text].

CIGAN, A. M., M. FOIANI, E. M. HANNIG, and A. G. HINNEBUSCH, 1991  Complex formation by positive and negative translational regulators of GCN4.. Mol. Cell. Biol. 11:3217-3228[Abstract/Free Full Text].

CIGAN, A. M., J. L. BUSHMAN, T. R. BOAL, and A. G. HINNEBUSCH, 1993  A protein complex of translational regulators of GCN4 is the guanine nucleotide exchange factor for eIF-2 in yeast. Proc. Natl. Acad. Sci. USA 90:5350-5354[Abstract/Free Full Text].

DEVER, T. E., L. FENG, R. C. WEK, A. M. CIGAN, and T. D. DONAHUE et al., 1992  Phosphorylation of initiation factor 2 by protein kinase GCN2 mediates gene-specific translational control of GCN4 in yeast. Cell 68:585-596[Medline].

DEVER, T. E., J. J. CHEN, G. N. BARBER, A. M. CIGAN, and L. FENG et al., 1993  Mammalian eukaryotic initiation factor 2 kinases functionally substitute for GCN2 in the GCN4 translational control mechanism of yeast. Proc. Natl. Acad. Sci. USA 90:4616-4620[Abstract/Free Full Text].

DEVER, T. E., W. YANG, S. ASTROM, A. S. BYSTROM, and A. G. HINNEBUSCH, 1995  Modulation of tRNA_i^Met, eIF-2 and eIF-2B expression shows that GCN4 translation is inversely coupled to the level of eIF-2.GTP.Met-tRNA_i^Met ternary complexes. Mol. Cell. Biol. 15:6351-6363[Abstract].

DONAHUE, T. F. and A. M. CIGAN, 1988  Genetic selection for mutations that reduce or abolish ribosomal recognition of the HIS4 translational initiatior region. Mol. Cell. Biol. 8:2955-2963[Abstract/Free Full Text].

FOIANI, M., A. M. CIGAN, C. J. PADDON, S. HARASHIMA, and A. G. HINNEBUSCH, 1991  GCD2, a translational repressor of the GCN4 gene, has a general function in the initiation of protein synthesis in Saccharomyces cerevisiae.. Mol. Cell. Biol. 11:3203-3216[Abstract/Free Full Text].

GARCÍA-BARRIO, M. T., T. NARANDA, C. R. VAZQUEZ DE ALDANA, R. CUESTA, and A. G. HINNEBUSCH et al., 1995  GCD10, a translational repressor of GCN4, is the RNA-binding subunit of eukaryotic translation initiation factor-3. Genes Dev. 9:1781-1796[Abstract/Free Full Text].

GRANT, C. M. and A. G. HINNEBUSCH, 1994  Effect of sequence context at stop codons on efficiency of reinitiation in GCN4 translational control. Mol. Cell. Biol. 14:606-618[Abstract/Free Full Text].

GRANT, C. M., P. F. MILLER, and A. G. HINNEBUSCH, 1995  Sequences 5' of the first upstream open reading frame in GCN4 mRNA are required for efficient translational reinitiation. Nucleic Acids Res. 23:3980-3888[Abstract/Free Full Text].

GUTHRIE, C., and G. R. FINK, 1991 Guide to yeast genetics and molecular biology. Methods Enzymol. 194.

HANNIG, E. M. and A. G. HINNEBUSCH, 1988  Molecular analysis of GCN3, a translational activator of GCN4: Evidence for posttranslational control of GCN3 regulatory function. Mol. Cell. Biol. 8:4808-4820[Abstract/Free Full Text].

HANNIG, E. M., N. P. WILLIAMS, R. C. WEK, and A. G. HINNEBUSCH, 1990  The translational activator GCN3 functions downstream from GCN1 and GCN2 in the regulatory pathway that couples GCN4 expression to amino acid availability in Saccharomyces cerevisiae.. Genetics 126:549-562[Abstract].

HANNIG, E. M., A. M. CIGAN, B. A. FREEMAN, and T. G. KINZY, 1993  GCD11, a negative regulator of GCN4 expression, encodes the gamma subunit of eIF-2 in Saccharomyces cerevisiae.. Mol. Cell. Biol. 13:506-520[Abstract/Free Full Text].

HARASHIMA, S. and A. G. HINNEBUSCH, 1986  Multiple GCD genes required for repression of GCN4, a transcriptional activator of the amino acid biosynthetic genes in Saccharomyces cerevisiae. Mol. Cell. Biol. 6:3990-3998[Abstract/Free Full Text].

HARASHIMA, S., E. M. HANNIG, and A. G. HINNEBUSCH, 1987  Interactions between positive and negative regulators of GCN4 controlling gene expression and entry into the yeast cell cycle. Genetics 117:409-419[Abstract/Free Full Text].

HERSHEY, J. W. B., 1989  Protein phosphorylation controls translation rates. J. Biol. Chem. 264:20823-20826[Free Full Text].

HILL, D. E. and K. STRUHL, 1988  Molecular characterization of GCD1, a yeast gene required for general control of amino acid biosynthesis and cell-cycle initiation. Nucleic Acids Res. 16:9253-9265[Abstract/Free Full Text].

HINNEBUSCH, A. G. and G. R. FINK, 1983  Positive regulation in the general amino acid control of Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 80:5374-5378[Abstract/Free Full Text].

HINNEBUSCH, A. G., 1985  A hierarchy of trans-acting factors modulates translation of an activator of amino acid biosynthetic genes in Saccharomyces cerevisiae.. Mol. Cell. Biol. 5:2349-2360[Abstract/Free Full Text].

HINNEBUSCH, A. G., 1988  Mechanisms of gene regulation in the general control of amino acid biosynthesis in Saccharomyces cerevisiae.. Microbiol. Rev. 52:248-273[Free Full Text].

HINNEBUSCH, A. G., 1992 General and pathway-specific regulatory mechanisms controlling the synthesis of amino acid biosynthetic enzymes in Saccharomyces cerevisiae, pp. 319–414 in The Molecular and Cellular Biology of the Yeast Saccharomyces. Gene Expression, edited by E. W. JONES, J. R. PRINGLE and J. R. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

HINNEBUSCH, A. G., 1994a  The eIF-2 kinases: Regulators of protein synthesis in starvation and stress. Cell. Biol. 5:417-426.

HINNEBUSCH, A. G., 1994b  Translational control of GCN4: An in vivo barometer of initiation-factor activity. Trends Biochem. Sci. 19:409-414[Medline].

HINNEBUSCH, A. G., 1996 Translational control of GCN4: Gene-specific regulation by phosphorylation of eIF-2, pp. 199–244 in Translational Control, edited by J. W. B. HERSHEY, M. B. MATHEWS and N. SONENBERG. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

HINNEBUSCH, A. G., and S. W. LIEBMAN, 1991 Protein synthesis and translational control in Saccharomyces cerevisiae, pp. 627–735 in The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome Dynamics, Protein Synthesis, and Energetics, Vol. I, edited by J. R. BROACH, J. R. PRINGLE and E. W. JONES. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

ITO, H., Y. FUKADA, K. MURATA, and A. KIMURA, 1983  Transformation of intact yeast cells treated with alkali cations. J. Bacteriol. 153:163-168[Abstract/Free Full Text].

JACKSON, R. J., 1991 Binding of met-RNA, pp. 193–223 in Translation in Eukaryotes, edited by H. TRACHSEL. CRC Press, Boca Raton, FL.

JOHNSTON, M. and R. W. DAVIS, 1984  Sequences that regulate the divergent GAL1-GAL10 promoter in Saccharomyces cerevisiae.. Mol. Cell. Biol. 4:1440-1448[Abstract/Free Full Text].

LANKER, S., J. L. BUSHMAN, A. G. HINNEBUSCH, H. TRACHSEL, and P. P. MÜELLER, 1992  Autoregulation of the yeast lysyl-tRNA-synthetase gene GCD5/KRS1 by translational and transcriptional control mechanisms. Cell 70:647-657[Medline].

LUCCHINI, G., A. G. HINNEBUSCH, C. CHEN, and G. R. FINK, 1984  Positive regulatory interactions of the HIS4 gene of Saccharomyces cerevisiae.. Mol. Cell. Biol. 4:1326-1333[Abstract/Free Full Text].

MARTON, M. J., D. CROUCH, and A. G. HINNEBUSCH, 1993  GCN1, a translational activator of GCN4 in S. cerevisiae, is required for phosphorylation of eukaryotic translation initiation 2 by protein kinase GCN2. Mol. Cell. Biol. 13:3541-3556[Abstract/Free Full Text].

MERRICK, W. C., 1992  Mechanism and regulation of eukaryotic protein synthesis. Microbiol. Rev. 56:291-315[Abstract/Free Full Text].

MERRICK, W. C., and J. W. B. HERSHEY, 1996 The pathway and mechanism of eukaryotic protein synthesis, pp. 31–69 in Translational Control, edited by J. W. B. HERSHEY, M. B. MATHEWS and N. SONENBERG. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

MOEHLE, C. M. and A. G. HINNEBUSCH, 1991  Association of RAP1 binding sites with stringent control of ribosomal protein gene transcription in Saccharomyces cerevisiae.. Mol. Cell. Biol. 11:2723-2735[Abstract/Free Full Text].

MÜELLER, P. P. and A. G. HINNEBUSCH, 1986  Multiple upstream AUG codons mediate translational control of GCN4.. Cell 45:201-207[Medline].

PADDON, C. J. and A. G. HINNEBUSCH, 1989  gcd12 mutations are gcn3-dependent alleles of GCD2, a negative regulator of GCN4 in the general amino acid control of Saccharomyces cerevisiae.. Genetics 122:543-550[Abstract/Free Full Text].

ROLFES, R. J. and A. G. HINNEBUSCH, 1993  Translation of the yeast transcriptional activator GCN4 is stimulated by purine limitation: implications for the activation of the protein kinase GCN2. Mol. Cell. Biol. 13:5099-5111[Abstract/Free Full Text].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SIKORSKI, R. S. and P. HIETER, 1989  A system of shuttle vectors and yeast host strains designated for efficient manipulation of DNA in Saccharomyces cerevisiae.. Genetics 122:19-27[Abstract/Free Full Text].

TZAMARIAS, D., I. ROUSSOU, and G. THIREOS, 1989  Coupling of GCN4 mRNA translational activation with decreased rates of polypeptide chain initiation. Cell 57:947-954[Medline].

WEK, R. C., M. RAMIREZ, B. M. JACKSON, and A. G. HINNEBUSCH, 1990  Identification of positive-acting domains in GCN2 protein kinase required for translational activation of GCN4 expression. Mol. Cell. Biol. 10:2820-2831[Abstract/Free Full Text].

WILLIAMS, N. P., A. G. HINNEBUSCH, and T. F. DONAHUE, 1989  Mutations in the structural genes for eukaryotic initiation factors 2 and 2ß of Saccharomyces cerevisiae disrupt translational control of GCN4 mRNA. Proc. Natl. Acad. Sci. USA 86:7515-7519[Abstract/Free Full Text].

WOLFNER, M., D. YEP, F. MESSENGUY, and G. R. FINK, 1975  Integration of amino acid biosynthesis into