- Split View
-
Views
-
Cite
Cite
Valmik K Vyas, Sergei Kuchin, Marian Carlson, Interaction of the Repressors Nrg1 and Nrg2 With the Snf1 Protein Kinase in Saccharomyces cerevisiae, Genetics, Volume 158, Issue 2, 1 June 2001, Pages 563–572, https://doi.org/10.1093/genetics/158.2.563
- Share Icon Share
Abstract
The Snf1 protein kinase is essential for the transcription of glucose-repressed genes in Saccharomyces cerevisiae. We identified Nrg2 as a protein that interacts with Snf1 in the two-hybrid system. Nrg2 is a C2H2 zinc-finger protein that is homologous to Nrg1, a repressor of the glucose- and Snf1-regulated STA1 (glucoamylase) gene. Snf1 also interacts with Nrg1 in the two-hybrid system and co-immunoprecipitates with both Nrg1 and Nrg2 from cell extracts. A LexA fusion to Nrg2 represses transcription from a promoter containing LexA binding sites, indicating that Nrg2 also functions as a repressor. An Nrg1 fusion to green fluorescent protein is localized to the nucleus, and this localization is not regulated by carbon source. Finally, we show that VP16 fusions to Nrg1 and Nrg2 allow low-level expression of SUC2 in glucose-grown cells, and we present evidence that Nrg1 and Nrg2 contribute to glucose repression of the DOG2 gene. These results suggest that Nrg1 and Nrg2 are direct or indirect targets of the Snf1 kinase and function in glucose repression of a subset of Snf1-regulated genes.
THE Snf1 protein kinase is highly conserved from yeast to plants to mammals (for reviews, see Gancedo 1998; Hardie et al. 1998). In Saccharomyces cerevisiae, Snf1 is a key component of the glucose signaling pathway and is essential for the transcription of many glucose-repressed genes in response to glucose limitation (Celenza and Carlson 1986). This kinase also has roles in metabolic regulation, glycogen accumulation, stress responses, meiosis and sporulation, invasive growth, life span, and aging (Thompson-Jaeger et al. 1991; Hardy et al. 1994; Honigberg and Lee 1998; Ashrafi et al. 2000; Cullen and Sprague 2000). Its mammalian homolog, AMP-activated protein kinase, regulates metabolism and transcription in response to the cellular energy supply, and the plant homologs are thought to be involved in sugar regulation.
The role of the Snf1 kinase in transcriptional control has been characterized in some detail. The adaptation of yeast cells to growth on nonpreferred carbon sources is accompanied by major changes in transcriptional patterns, and Snf1 appears to act at many control points. Thus far, Snf1 has been shown to regulate the expression and function of both transcriptional repressors and activators in response to glucose availability (for review, see Carlson 1999). Recent evidence also implicates Snf1 in direct regulation of the RNA polymerase II holoenzyme (Kuchin et al. 2000).
One of the major mechanisms by which Snf1 regulates transcription is by regulating the function of the transcriptional repressor Mig1. Mig1 is a zinc-finger protein that binds to sites in the promoters of many glucose-repressed genes and recruits the global corepressor Ssn6(Cyc8)-Tup1 (Nehlin and Ronne 1990; Treitel and Carlson 1995; Tzamarias and Struhl 1995). Several lines of evidence indicate that Snf1 phosphorylates Mig1 in response to glucose limitation and thereby regulates its nuclear localization and inhibits its repressor function (DeVit et al. 1997; Ostling and Ronne 1998; Treitel et al. 1998).
Snf1 also effects transcriptional control by regulating transcriptional activators. Snf1 regulates the phosphorylation and function of the Cys6 zinc-cluster activators Sip4 and Cat8, which bind to the carbon source-responsive elements (CSRE) of gluconeogenic genes (Lesage et al. 1996; Rahner et al. 1996, 1999; Randez-Gil et al. 1997; Vincent and Carlson 1998; Rahner et al. 1999). Snf1 has been shown to interact physically with Sip4 (Lesage et al. 1996; Vincent and Carlson 1998).
The two-hybrid system has been useful in detecting interactions of Snf1 with transcriptional activators and repressors. Sip4 was first identified by its two-hybrid interaction with Snf1 (Yang et al. 1992). Mig1 also interacts with Snf1 in two-hybrid assays (Treitel et al. 1998), and in this case interaction was much stronger with a catalytically defective mutant Snf1 protein, Snf1K84R (Celenza and Carlson 1986), which bears a substitution of arginine for the conserved lysine in the ATP-binding site. Several Srb/mediator proteins associated with the RNA polymerase II holoenzyme have also been shown to interact with Snf1 (Kuchin et al. 2000).
Straina . | Genotypeb . |
---|---|
MCY829 | MATα his3 lys2 ura3 |
MCY2916 | MATa snf1Δ10 his3 leu2 lys2 ura3 |
MCY2693 | MATa snf1-K84R his3 leu2 ura3 |
MCY3647 | MATα his3 leu2 lys2 ura3 |
MCY3912 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY3913 | MATα snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4516 | MATa snf1Δ10 nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4521 | MATa nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4523 | MATa nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4525 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY4527 | MATa snf1Δ10 nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4529 | MATa snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4531 | MATα NRG1-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4536 | MATα NRG2-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4548 | MATa snf1Δ10 nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4549 | MATa nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
FY250c | MATα his3 leu2Δ1 trp1Δ63 ura3 |
HF7Cd | MATa ade2 gal4 his3 leu2 lys2 trp1-901 LYS2::GAL1UAS-GAL1TATA-HIS3 URA3::GAL417-mers(x3)-CYC1TATA-lacZ |
CTY10-5de | MATa gal4 gal80 URA3::lexAop-lacZ his3 leu2 ade2 trp1-901 |
Straina . | Genotypeb . |
---|---|
MCY829 | MATα his3 lys2 ura3 |
MCY2916 | MATa snf1Δ10 his3 leu2 lys2 ura3 |
MCY2693 | MATa snf1-K84R his3 leu2 ura3 |
MCY3647 | MATα his3 leu2 lys2 ura3 |
MCY3912 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY3913 | MATα snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4516 | MATa snf1Δ10 nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4521 | MATa nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4523 | MATa nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4525 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY4527 | MATa snf1Δ10 nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4529 | MATa snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4531 | MATα NRG1-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4536 | MATα NRG2-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4548 | MATa snf1Δ10 nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4549 | MATa nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
FY250c | MATα his3 leu2Δ1 trp1Δ63 ura3 |
HF7Cd | MATa ade2 gal4 his3 leu2 lys2 trp1-901 LYS2::GAL1UAS-GAL1TATA-HIS3 URA3::GAL417-mers(x3)-CYC1TATA-lacZ |
CTY10-5de | MATa gal4 gal80 URA3::lexAop-lacZ his3 leu2 ade2 trp1-901 |
MCY strains and FY250 have the S288C genetic background.
Alleles are ura3-52, his3-Δ200, lys2-801, ade2-101, leu2-3,112, and trp1Δ1 except where otherwise noted.
Gift of F. Winston (Harvard Medical School).
Feilotter et al. (1994).
Gift of R. Sternglanz (SUNY, Stonybrook, NY).
Straina . | Genotypeb . |
---|---|
MCY829 | MATα his3 lys2 ura3 |
MCY2916 | MATa snf1Δ10 his3 leu2 lys2 ura3 |
MCY2693 | MATa snf1-K84R his3 leu2 ura3 |
MCY3647 | MATα his3 leu2 lys2 ura3 |
MCY3912 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY3913 | MATα snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4516 | MATa snf1Δ10 nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4521 | MATa nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4523 | MATa nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4525 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY4527 | MATa snf1Δ10 nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4529 | MATa snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4531 | MATα NRG1-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4536 | MATα NRG2-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4548 | MATa snf1Δ10 nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4549 | MATa nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
FY250c | MATα his3 leu2Δ1 trp1Δ63 ura3 |
HF7Cd | MATa ade2 gal4 his3 leu2 lys2 trp1-901 LYS2::GAL1UAS-GAL1TATA-HIS3 URA3::GAL417-mers(x3)-CYC1TATA-lacZ |
CTY10-5de | MATa gal4 gal80 URA3::lexAop-lacZ his3 leu2 ade2 trp1-901 |
Straina . | Genotypeb . |
---|---|
MCY829 | MATα his3 lys2 ura3 |
MCY2916 | MATa snf1Δ10 his3 leu2 lys2 ura3 |
MCY2693 | MATa snf1-K84R his3 leu2 ura3 |
MCY3647 | MATα his3 leu2 lys2 ura3 |
MCY3912 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY3913 | MATα snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4516 | MATa snf1Δ10 nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4521 | MATa nrg1Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4523 | MATa nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4525 | MATa ade2 his3 leu2 lys2 trp1 ura3 |
MCY4527 | MATa snf1Δ10 nrg2Δ::kanMX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4529 | MATa snf1Δ10 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4531 | MATα NRG1-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4536 | MATα NRG2-GFP(S65T)::kanMX6 snf1Δ10 his3 leu2 lys2 ura3 |
MCY4548 | MATa snf1Δ10 nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
MCY4549 | MATa nrg1Δ::kanMX6 nrg2Δ::His3MX6 ade2 his3 leu2 lys2 trp1 ura3 |
FY250c | MATα his3 leu2Δ1 trp1Δ63 ura3 |
HF7Cd | MATa ade2 gal4 his3 leu2 lys2 trp1-901 LYS2::GAL1UAS-GAL1TATA-HIS3 URA3::GAL417-mers(x3)-CYC1TATA-lacZ |
CTY10-5de | MATa gal4 gal80 URA3::lexAop-lacZ his3 leu2 ade2 trp1-901 |
MCY strains and FY250 have the S288C genetic background.
Alleles are ura3-52, his3-Δ200, lys2-801, ade2-101, leu2-3,112, and trp1Δ1 except where otherwise noted.
Gift of F. Winston (Harvard Medical School).
Feilotter et al. (1994).
Gift of R. Sternglanz (SUNY, Stonybrook, NY).
In an effort to identify new downstream targets of the Snf1 kinase, we performed a two-hybrid screen with the catalytically defective Snf1K84R. We recovered Nrg2, a zinc-finger protein that is a close homolog of the DNA-binding repressor protein Nrg1. Nrg1 functions in glucose repression of the STA1 gene, which encodes one of the glucoamylase isozymes responsible for starch degradation in S. cerevisiae var. diastaticus (Park et al. 1999). Furthermore, the Snf1 kinase is known to be required for derepression of STA2 (Kuchin et al. 1993), which is identical to STA1 throughout the promoter and coding region. We therefore examined the relationship of Nrg1 and Nrg2 to the Snf1 kinase. Both proteins interact with Snf1 in the two-hybrid system and both co-immunoprecipitate with Snf1 from cell extracts. We present evidence that Nrg2 functions as a repressor and that the nuclear localization of Nrg1, unlike that of Mig1, is not regulated by carbon source. Finally, we examine the roles of Nrg1 and Nrg2 in repression of Snf1-dependent genes.
MATERIALS AND METHODS
Strains and genetic methods: The S. cerevisiae strains used are listed in Table 1. To construct strain MCY4531, we first generated a 2.5-kb DNA fragment using the polymerase chain reaction (PCR) with the template pFA6a-GFP(S65T)-kanMX6 (Longtine et al. 1998) and two primers, one including the 15 codons preceding the termination codon of NRG1 and the other including 45 base pairs following the stop codon. This fragment was used to transform strain MCY2916, with selection on rich medium containing 200 μg/ml G418 (Life Technologies, Gaithersburg, MD), and its integration at the NRG1 locus was verified by PCR. Strain MCY4536 was created in the same way, except that the targeting sequence was for NRG2. To construct the nrg1Δ::kanMX6, nrg2Δ::kanMX6, and nrg2Δ:: His3MX6 mutations, we first amplified kanMX6 or His3MX6 (Longtine et al. 1998) with oligonucleotides containing 45 nucleotides flanking the open reading frame of NRG1 or NRG2. PCR products were used to transform MCY3912 or MCY3913 with selection for the marker, and disruption was verfied by PCR amplification of genomic DNA.
Standard methods for yeast genetic analysis and transformation were used (Rose et al. 1990). Cells were grown in synthetic complete (SC) medium lacking appropriate supplements to maintain selection for plasmids.
Plasmids: Plasmids used in this study are listed in Table 2. pRJ215 contains the BamHI fragment from pRJ80 (Ludin et al. 1998) in the BamHI site of pEG202 (Golemis et al. 1997). pSK106 was created by cloning the BamHI fragment from pRJ215 into the BamHI site of pGBT-9 (Bartel et al. 1993). The coding regions of NRG1 and NRG2 were amplified by PCR from genomic DNA using Vent polymerase (New England Biolabs, Beverly, MA) with oligonucleotides K47 and K48 (Nrg2) and K49 and K50 (Nrg1), each containing a BamHI site. BamHI-digested PCR product was cloned into the BamHI site of plasmids pACTII (Legrain et al. 1994), pVP16 (Vojtek et al. 1993), pWS93 (Song and Carlson 1998), pSH2-1 (Hanes and Brent 1989), and pEG202.
Oligonucleotides: Oligonucleotides used as primers in PCR were the following: K47, GCGCGGATCCTAATGTCCATAGG TTACAAAGAC; K48, GCGCGGATCCTCAACTGCTAGCCT CCCTCC; K49, GCGCGGATCCTAATGTTTTACCCATATAA CTATAG; K50, GCGCGGATCCGTCAATTATTGTCCCTTT TTC (BamHI sites are underlined).
Name . | Vector . | Expressed protein . | Source or reference . |
---|---|---|---|
pSK106 | pGBT-9 | GBD-Snf1K84R | This study |
pRJ190 | pSH2-1 | LexA87-Snf1KD | Jiang and Carlson (1996) |
pRJ192 | pSH2-1 | LexA87-Snf1RD | R. Jiang and M. Carlson, unpublished data |
pV37 | pSH2-1 | LexA87-Nrg2 | This study |
pV38 | pSH2-1 | LexA87-Nrg1 | This study |
pV39 | pACTII | GAD-Nrg2 | This study |
pV40 | pACTII | GAD-Nrg1 | This study |
pV45 | pVP16 | VP16-Nrg2 | This study |
pV46 | pVP16 | VP16-Nrg1 | This study |
pV35 | pWS93 | HA3-Nrg2 | This study |
pV36 | pWS93 | HA3-Nrg1 | This study |
pSK117 | pSK37 | Snf1 | Treitel et al. (1998) |
pSK118 | pSK37 | Snf1K84R | Treitel et al. (1998) |
pIT469 | pEG202 | LexA-Snf1 | Kuchin et al. (2000) |
pRJ215 | pEG202 | LexA-Snf1K84R | R. Jiang and M. Carlson, unpublished data |
Name . | Vector . | Expressed protein . | Source or reference . |
---|---|---|---|
pSK106 | pGBT-9 | GBD-Snf1K84R | This study |
pRJ190 | pSH2-1 | LexA87-Snf1KD | Jiang and Carlson (1996) |
pRJ192 | pSH2-1 | LexA87-Snf1RD | R. Jiang and M. Carlson, unpublished data |
pV37 | pSH2-1 | LexA87-Nrg2 | This study |
pV38 | pSH2-1 | LexA87-Nrg1 | This study |
pV39 | pACTII | GAD-Nrg2 | This study |
pV40 | pACTII | GAD-Nrg1 | This study |
pV45 | pVP16 | VP16-Nrg2 | This study |
pV46 | pVP16 | VP16-Nrg1 | This study |
pV35 | pWS93 | HA3-Nrg2 | This study |
pV36 | pWS93 | HA3-Nrg1 | This study |
pSK117 | pSK37 | Snf1 | Treitel et al. (1998) |
pSK118 | pSK37 | Snf1K84R | Treitel et al. (1998) |
pIT469 | pEG202 | LexA-Snf1 | Kuchin et al. (2000) |
pRJ215 | pEG202 | LexA-Snf1K84R | R. Jiang and M. Carlson, unpublished data |
Name . | Vector . | Expressed protein . | Source or reference . |
---|---|---|---|
pSK106 | pGBT-9 | GBD-Snf1K84R | This study |
pRJ190 | pSH2-1 | LexA87-Snf1KD | Jiang and Carlson (1996) |
pRJ192 | pSH2-1 | LexA87-Snf1RD | R. Jiang and M. Carlson, unpublished data |
pV37 | pSH2-1 | LexA87-Nrg2 | This study |
pV38 | pSH2-1 | LexA87-Nrg1 | This study |
pV39 | pACTII | GAD-Nrg2 | This study |
pV40 | pACTII | GAD-Nrg1 | This study |
pV45 | pVP16 | VP16-Nrg2 | This study |
pV46 | pVP16 | VP16-Nrg1 | This study |
pV35 | pWS93 | HA3-Nrg2 | This study |
pV36 | pWS93 | HA3-Nrg1 | This study |
pSK117 | pSK37 | Snf1 | Treitel et al. (1998) |
pSK118 | pSK37 | Snf1K84R | Treitel et al. (1998) |
pIT469 | pEG202 | LexA-Snf1 | Kuchin et al. (2000) |
pRJ215 | pEG202 | LexA-Snf1K84R | R. Jiang and M. Carlson, unpublished data |
Name . | Vector . | Expressed protein . | Source or reference . |
---|---|---|---|
pSK106 | pGBT-9 | GBD-Snf1K84R | This study |
pRJ190 | pSH2-1 | LexA87-Snf1KD | Jiang and Carlson (1996) |
pRJ192 | pSH2-1 | LexA87-Snf1RD | R. Jiang and M. Carlson, unpublished data |
pV37 | pSH2-1 | LexA87-Nrg2 | This study |
pV38 | pSH2-1 | LexA87-Nrg1 | This study |
pV39 | pACTII | GAD-Nrg2 | This study |
pV40 | pACTII | GAD-Nrg1 | This study |
pV45 | pVP16 | VP16-Nrg2 | This study |
pV46 | pVP16 | VP16-Nrg1 | This study |
pV35 | pWS93 | HA3-Nrg2 | This study |
pV36 | pWS93 | HA3-Nrg1 | This study |
pSK117 | pSK37 | Snf1 | Treitel et al. (1998) |
pSK118 | pSK37 | Snf1K84R | Treitel et al. (1998) |
pIT469 | pEG202 | LexA-Snf1 | Kuchin et al. (2000) |
pRJ215 | pEG202 | LexA-Snf1K84R | R. Jiang and M. Carlson, unpublished data |
Two-hybrid screen: A two-hybrid screen (Fields and Song 1989) for proteins that interact with a Gal4 DNA-binding domain (GBD) fusion to Snf1K84R was carried out in strain HF7C, which contains a chromosomally located GAL1-HIS3 reporter. The strain was transformed with pSK106, which expresses GBD-Snf1K84R, and with a library (gift of S. Elledge, Baylor College of Medicine) of S. cerevisiae cDNAs fused to the Gal4 activating domain (GAD). Transformants were selected on SC-His + 2% glucose plates for a His+ phenotype. Five plasmids conferred a His+ phenotype and gave blue color in CTY10-5d and were subjected to sequence analysis. Besides NRG2, the recovered sequences were SNF4, GPM1, POR1, and TDH1. The others were not analyzed further.
Two-hybrid assays with LexA fusion proteins were carried out in strain CTY10-5d or in strain FY250 transformed with the pSH18-34 reporter, containing LexA binding sites 5′ to a GAL1-lacZ reporter (Estojak et al. 1995; Golemis et al. 1997).
Invertase and β-galactosidase assays: Invertase activity was assayed in whole cells as previously described (Jiang and Carlson 1996). β-Galactosidase activity was assayed in permeabilized cells (Guarente 1983) and expressed in Miller units (Miller 1972). A filter assay in which β-galactosidase activity confers blue color was also used as previously described (Jiang and Carlson 1996).
Co-immunoprecipitation assays: Preparation of protein extracts and immunoprecipitation procedures were essentially as described previously (Celenza et al. 1989). The extraction buffer was 50 mm HEPES (pH 7.5), 150 mm NaCl, 0.5% Triton X-100, 1 mm dithiothreitol, 10% glycerol, and contained 2 mm phenylmethylsulfonyl fluoride and complete protease inhibitor cocktail (Roche Molecular Biochemical). Proteins were immunoprecipitated with anti(α)-hemagglutinin (HA) monoclonal antibody (Roche Molecular Biochemical) or α-LexA monoclonal antibody (CLONTECH, Palo Alto, CA) in the same buffer, except that it contained 0.25% Triton X-100 for immunoprecipitation with α-HA and 50 mm NaCl and 0.1% Triton X-100 for immunoprecipitation with α-LexA.
Immunoblot analysis: Proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and analyzed by immunoblotting using polyclonal α-Snf1 (Celenza and Carlson 1986), monoclonal α-HA (Roche Molecular Biochemical), or polyclonal α-LexA (Invitrogen, San Diego). Antibodies were detected by chemiluminescence with ECL or ECL Plus reagents (Amersham, Arlington Heights, IL).
Imaging of green flourescent protein fluorescence: Cells were grown to midlog phase in synthetic media containing 5% glucose or 2% glycerol plus 2% ethanol, harvested by centrifugation, and resuspended in nonfluorescent media [0.9 g/liter KH2PO4, 0.23 g/liter K2HPO4, 0.5 g/liter MgSO4, 3.5 g/liter (NH4)2SO4] containing the appropriate carbon source. Nuclei were stained by addition of 0.8 μg/ml of 4′,6-diamidino-2-phenylindole (DAPI) for 5 min. Fluorescence of green fluorescent protein (GFP) fusion proteins was visualized in unfixed cells by using a Nikon Eclipse E800 fluorescent microscope. Images were captured by using a digital camera (Hamamatsu Orca-100, Hamamatsu, Japan) and Openlab software (Improvision) and were converted to Adobe Photoshop 2.5.1 files for processing.
Microarray analysis: Strain MCY3912 carrying pV46 or pVP16 was grown in SC-Leu + 5% glucose. Total yeast RNA was extracted with hot phenol, and poly(A)+ RNA was purified by oligo(dT) chromatography (QIAGEN, Chatsworth, CA). Fluorescently labeled cDNA was prepared, and expression of 5805 yeast open reading frames (ORFs) was analyzed using DNA microarrays as described (DeRisi et al. 1997). Microarrays were produced by the Columbia University Microarray Project.
RESULTS
Identification of Nrg2 in a two-hybrid screen for interaction with Snf1: We carried out a two-hybrid screen for proteins that interact with the catalytically defective Snf1 protein kinase, Snf1K84R. GBD-Snf1K84R was used as a bait to screen a library of cDNAs fused to GAD. We recovered six clones that were His+ in combination with GBD-Snf1K84R (see materials and methods). Five clones also caused blue color in combination with both LexA-Snf1K84R and LexA-Snf1 in strain CTY10-5d.
Sequence analysis showed that one of these clones (pKRIP6) encodes an in-frame fusion of GAD at a position five nucleotides preceding codon 1 of the NRG2 gene (YBR066C). NRG2 encodes a protein of 220 amino acids with a predicted molecular mass of 25 kD and two zinc fingers at the C terminus. The zinc fingers are homologous to those of the transcriptional activators Msn2 and Msn4, which bind stress response elements (Martinez-Pastor et al. 1996), and to those of the repressors Mig1 and Mig2. However, the closest homolog to Nrg2 is Nrg1, which is 44% similar overall and 84% identical in the zinc-finger region (Figure 1).
NRG1 was identified as a multicopy inhibitor of the glucose-repressible STA1 promoter in S. cerevisiae var. diastaticus (Park et al. 1999). In the selection scheme, the STA1 promoter was used to drive expression of TPK2, encoding a catalytic subunit of cyclic AMP-dependent protein kinase, which is toxic to cells at high levels. Multicopy NRG1 restored healthy growth and repressed glucoamylase expression. Nrg1 bound to sites in the STA1 upstream region, and a LexA fusion to Nrg1 repressed transcription of a reporter with LexA sites, dependent on the corepressor Ssn6-Tup1 (Park et al. 1999). Mutation of NRG1 relieved glucose repression of STA1. Release from glucose repression of glucoamylase genes requires the Snf1 protein kinase (Kuchin et al. 1993), and thus these studies suggest a functional connection between Nrg1 and Snf1. We therefore included Nrg1 in subsequent experiments to assess its interaction with Snf1.
The Snf1 kinase comprises a catalytic domain (residues 1-392, designated Snf1KD) and a regulatory domain (residues 392-633, designated Snf1RD), which binds to the kinase domain and inhibits its activity (Jiang and Carlson 1996). We assayed the two-hybrid interaction of GAD-Nrg1 and GAD-Nrg2 with LexA DNA-binding domain (LexA87) fusions to Snf1KD and Snf1RD. Interaction was monitored by assaying β-galactosidase activity in cells grown under glucose-repressing conditions (5% glucose) and after a 3-hr shift to 0.05% glucose. Both GAD-Nrg1 and GAD-Nrg2 interacted with LexA87-Snf1KD but not with LexA87-Snf1RD (Table 3).
Co-immunoprecipitation of the Snf1 kinase with Nrg1 and Nrg2: To confirm that Snf1 interacts with Nrg1 and Nrg2, we tested for co-immunoprecipitation of the kinase with triple HA epitope-tagged proteins. Protein extracts were prepared from snf1Δ cells expressing HA-Nrg1 or HA-Nrg2 and Snf1, Snf1K84R, or no Snf1 protein. Immunoblot analysis showed that levels of both HA-Nrg1 and HA-Nrg2 were severely reduced in cells expressing no Snf1 protein, but levels were normal in cells expressing Snf1K84R and thus defective only for Snf1 catalytic activity (Figure 2A; input panel); these results are consistent with physical interactions in vivo. Proteins were immunoprecipitated with α-HA antibodies, and the precipitates were analyzed by SDS-PAGE and immunoblot analysis with α-Snf1 antibodies. Snf1 and Snf1K84R co-immunoprecipitated with both HA-Nrg1 and HA-Nrg2, but coprecipitation with HA-Nrg2 was more efficient (Figure 2A; CoIP panel). In control experiments, Snf1 did not coprecipitate with the triple HA tag expressed from the vector.
We also tested for co-immunoprecipitation of HA-Nrg1 and HA-Nrg2 with LexA-Snf1 (Figure 2B). Extracts were prepared from wild-type cells expressing LexA-Snf1 or LexA-Snf1K84R and HA-Nrg1 or HA-Nrg2, and LexA proteins were immunoprecipitated with α-LexA antibodies. Immunoblot analysis with α-HA antibody showed that HA-Nrg1 and HA-Nrg2 co-immunoprecipitated with both LexA-Snf1 and LexA-Snf1K84R, but not with LexA alone. Again, HA-Nrg1 was less efficiently recovered than HA-Nrg2.
LexA fusion . | GAD fusion . | 5% glucose . | Shift . |
---|---|---|---|
LexA87-Snf1KD | GAD-Nrg1 | 2.6 | 11 |
GAD-Nrg2 | 12 | 13 | |
GAD | 0.7 | 3.1 | |
LexA87-Snf1RD | GAD-Nrg1 | 0.5 | 0.3 |
GAD-Nrg2 | 0.6 | 0.3 | |
GAD | 0.2 | 0.1 |
LexA fusion . | GAD fusion . | 5% glucose . | Shift . |
---|---|---|---|
LexA87-Snf1KD | GAD-Nrg1 | 2.6 | 11 |
GAD-Nrg2 | 12 | 13 | |
GAD | 0.7 | 3.1 | |
LexA87-Snf1RD | GAD-Nrg1 | 0.5 | 0.3 |
GAD-Nrg2 | 0.6 | 0.3 | |
GAD | 0.2 | 0.1 |
Transformants of strain FY250 expressed the indicated fusion proteins from plasmids listed in Table 2 and carried the lacZ reporter plasmid pSH18-34. Transformants were grown in selective SC + 5% glucose medium (5% glucose) and shifted to SC + 0.05% glucose for 3 hr (Shift). Values are the average β-galactosidase activity for 5-15 transformants. Standard errors were <15%. Additional control experiments showed that strains expressing LexA87 in combination with each of the GAD proteins produced no significant β-galactosidase activity.
LexA fusion . | GAD fusion . | 5% glucose . | Shift . |
---|---|---|---|
LexA87-Snf1KD | GAD-Nrg1 | 2.6 | 11 |
GAD-Nrg2 | 12 | 13 | |
GAD | 0.7 | 3.1 | |
LexA87-Snf1RD | GAD-Nrg1 | 0.5 | 0.3 |
GAD-Nrg2 | 0.6 | 0.3 | |
GAD | 0.2 | 0.1 |
LexA fusion . | GAD fusion . | 5% glucose . | Shift . |
---|---|---|---|
LexA87-Snf1KD | GAD-Nrg1 | 2.6 | 11 |
GAD-Nrg2 | 12 | 13 | |
GAD | 0.7 | 3.1 | |
LexA87-Snf1RD | GAD-Nrg1 | 0.5 | 0.3 |
GAD-Nrg2 | 0.6 | 0.3 | |
GAD | 0.2 | 0.1 |
Transformants of strain FY250 expressed the indicated fusion proteins from plasmids listed in Table 2 and carried the lacZ reporter plasmid pSH18-34. Transformants were grown in selective SC + 5% glucose medium (5% glucose) and shifted to SC + 0.05% glucose for 3 hr (Shift). Values are the average β-galactosidase activity for 5-15 transformants. Standard errors were <15%. Additional control experiments showed that strains expressing LexA87 in combination with each of the GAD proteins produced no significant β-galactosidase activity.
To examine the possibility that Snf1 phosphorylates Nrg1 or Nrg2, we performed kinase assays with both of the above sets of immunoprecipitates. No Snf1-dependent phosphorylation of Nrg1 or Nrg2 was detected (data not shown). In addition, we examined HA-Nrg1 and HA-Nrg2 for phosphorylation in vivo by immunoblot analysis. No Snf1-dependent differences in the mobility of either protein were detected when wild-type and snf1 mutant cells were grown in 5% glucose or shifted to 0.05% glucose or shifted to 2% glycerol plus 2% ethanol (data not shown). Subsequent analysis of LexA fusion proteins also revealed no Snf1-dependent differences in mobility (Figure 3C).
Repressor function of Nrg1 and Nrg2: Previous studies showed that LexA87-Nrg1 represses expression of a reporter with LexA binding sites in glucose-grown cells (Park et al. 1999). We therefore tested the ability of LexA87-Nrg2 to function as a repressor in this assay. Wild-type strain MCY829 was transformed with plasmids expressing LexA87-Nrg1, LexA87-Nrg2, or LexA87 alone from the ADH1 promoter, and a CYC1-lacZ reporter with either four or zero LexA binding sites 5′ to the upstream activation sites (UAS; Figure 3A). Reporter gene expression was monitored by assaying β-galactosidase activity. LexA87-Nrg1 and LexA87-Nrg2 both repressed transcription about 15-fold in cells grown in 5% glucose (Figure 3B).
We then examined the regulation of transcriptional repression by carbon source. Repression by both LexA87-Nrg1 and LexA87-Nrg2 was maintained after a shift to low glucose (0.05%) or during steady-state growth in raffinose, sucrose, or galactose (data not shown). In accord with these findings, the STA genes are not derepressed under these growth conditions, but only during growth in glycerol plus ethanol (Pretorius et al. 1986; Kartasheva et al. 1996). No repression of the reporter was observed when cells were grown in 2% glycerol plus 2% ethanol (Figure 3B); however, immunoblot analysis showed that levels of both fusion proteins were very low when cells were grown in glycerol plus ethanol (Figure 3C), which could account for the lack of repression. The low protein levels can probably be attributed to reduced expression from the ADH1 promoter because levels of ADH1 mRNA are 6- to 10-fold lower during growth on ethanol than on glucose (Denis et al. 1983). Moreover, the levels of HA-Nrg1, HA-Nrg2, and LexA87-Sip4, when expressed from this vector, were also very low in glycerol/ethanol-grown cells (data not shown). These data exclude effects specific to the LexA87 tag, and Sip4 is unlikely to be specifically degraded under these conditions as it functions in the activation of gluconeogenic genes. Park et al. (1999) reported that LexA87-Nrg1 does not repress in cells grown in glycerol plus ethanol, but it is likely that the protein was not present because their expression vector was identical to ours.
We attempted to obtain evidence for a role of Snf1 in inhibiting repressor function by examining release from repression in wild-type and snf1 mutant cells after a shift from glucose to glycerol plus ethanol. Unfortunately, in wild-type cells the release occurred too slowly. Cells expressing LexA87-Nrg1 or LexA87-Nrg2 and containing a reporter with LexA sites only doubled their β-galactosidase activity during a 3-hr shift (repression ratios decreased to ∼7; data not shown), and protein levels were already lower by 3 hr (Figure 3C). In snf1-K84R mutant cells, LexA87-Nrg1 and LexA87-Nrg2 repressed transcription ∼15-fold during growth on glucose, and repression was not relieved during the 3-hr shift (repression ratios of ∼20; data not shown); however, protein levels remained high in the mutant cells (Figure 3C).
Nuclear localization of Nrg1: One of the mechanisms by which the Snf1 kinase regulates the function of the Mig1 repressor in response to the glucose signal is by regulating its nuclear localization (DeVit et al. 1997; DeVit and Johnston 1999). To address the possibility that the localization of Nrg1 or Nrg2 is similarly regulated, we constructed strains expressing Nrg1 or Nrg2 fused to GFP from the cognate chromosomal promoters. We did not detect significant fluorescence from Nrg2-GFP. Nrg1-GFP was localized in the nucleus when cells were grown in glucose, consistent with its function as a repressor, but also remained in the nucleus when cells were grown in glycerol plus ethanol (Figure 4). Similarly, no export of Nrg1-GFP was observed when cells were shifted from glucose to glycerol for 15 min (data not shown). Immunoblot analysis confirmed that Nrg1-GFP is a stable protein (data not shown). These findings suggest that the ability of Nrg1 to function as a repressor is not regulated by differential subcellular localization.
Nrg1 and Nrg2 fused to the VP16 activation domain activate low-level SUC2 expression in glucose: The repressor Mig1 has broad roles in glucose repression of many genes, whereas thus far, Nrg1 is known to affect only the glucoamylase genes. NRG1 and NRG2 have not been identified genetically in searches for regulators of other glucose-repressed genes; however, it is possible that their contributions to repression are modest. To test the possibility that Nrg1 and Nrg2 regulate SUC2, another glucose-repressed gene that is controlled by the Snf1 pathway, we fused the viral VP16 activation domain to both Nrg proteins. Our rationale was that the overexpressed VP16-Nrg fusion protein would compete with the native protein for binding to its sites. Similar fusions of VP16 to the repressor Mig1 activate expression of SUC2 in glucose-grown cells (Ostling et al. 1996).
We first used a SUC2::HIS3 reporter that exhibits glucose-repressible expression of HIS3 (Tu and Carlson 1994). Transformants of strain MCY3912 (his3) bearing SUC2::HIS3 and expressing VP16-Nrg1 or VP16-Nrg2 were His+ on medium containing 2% glucose, whereas transformants expressing VP16 alone were His- (Figure 5A). All transformants were His+ on medium containing 2% sucrose. We also assayed expression of invertase activity from the chromosomal SUC2 locus in cells grown in 5% glucose. VP16-Nrg1 and VP16-Nrg2 both activated SUC2 expression about 10-fold relative to VP16 alone (Figure 5B). However, invertase activity was still about 30-fold lower than the levels typically detected in derepressed wild-type cells. Although it is possible that overexpression of the fusion proteins leads to aberrant binding, these data suggest that the native Nrg1 and Nrg2 proteins have some role in glucose repression of SUC2, which may involve either direct binding to the SUC2 promoter or an indirect mechanism. Consistent with these results, an independent study has demonstrated that an nrg1Δ mutation causes defects in glucose repression of transcription of the SUC2, GAL1, and GAL10 genes (Zhou and Winston 2001).
Mutations in NRG1 and NRG2 affect glucose repression of the DOG2 gene: To identify other Snf1-dependent genes that are targets of repression, we used DNA microarray analysis to identify genes that are upregulated in glucose-grown cells expressing VP16-Nrg1, as compared to cells expressing VP16 alone (data not shown; see materials and methods). Among the genes that were upregulated, we identified one gene, DOG2, that is known to be regulated by glucose repression and by Snf1 (Randez-Gil et al. 1995; Lutfiyya et al. 1998; Tsujimoto et al. 2000). DOG2 encodes 2-deoxyglucose-6-phosphate phosphatase and confers resistance to 2-deoxyglucose toxicity. DOG2 is regulated by the repressors Mig1 and Mig2 in response to glucose and also by the stress response factors Msn2 and Msn4.
To explore the regulation of DOG2 by Nrg1 and Nrg2, we constructed nrg1Δ and nrg2Δ single and double mutants (see materials and methods) and introduced a plasmid bearing a DOG2 promoter fusion to lacZ, pBM3501 (Lutfiyya et al. 1998). When transformants were grown in 2% glucose, the mutants showed a two- to threefold elevation in β-galactosidase activity (2.1-3.3 units) relative to wild type (1.1 units); values are averages for 5-15 transformants. Although the defect in glucose repression is small, Nrg1 and Nrg2 are predicted to have only a modest role. Mig1 and Mig2 are largely responsible for glucose repression of DOG2-lacZ, and the residual repression in a mig1Δ mig2Δ double mutant is only twofold (Lutfiyya et al. 1998). The nrg1Δ and nrg2Δ mutations did not affect derepression as all strains produced similar activity (15-20 units) after a shift to 0.05% glucose for 3 hr.
Previous studies showed that derepression of DOG2 in response to glucose limitation requires the Snf1 kinase (Tsujimoto et al. 2000). To examine the relationship of Snf1 to Nrg1 and Nrg2, we assayed derepression of DOG2-lacZ in strains mutant for snf1Δ in combination with nrg1Δ or nrg2Δ or both. Transformants were grown in 2% glucose, shifted to 0.05% glucose for 3 hr, and assayed for β-galactosidase activity. Both nrg1Δ and nrg2Δ partially suppressed the snf1Δ mutant defect (Figure 6). Again the effect is small, but only partial suppression would be expected because repression by Mig1 is not relieved in a snf1Δ mutant (DeVit et al. 1997; Ostling and Ronne 1998; Treitel et al. 1998). These findings suggest that one of the roles of Snf1 in regulating the DOG2 promoter is to inhibit repression by Nrg1 and Nrg2.
DISCUSSION
We have identified Nrg2 in a two-hybrid screen for proteins that interact with the Snf1 protein kinase. Nrg2 was of particular interest because its close homolog, Nrg1, functions in glucose repression of the STA1 gene, and release from repression depends on Snf1. We here present evidence that both Nrg1 and Nrg2 interact physically with the Snf1 kinase. Both proteins interact with the catalytic domain of Snf1 in two-hybrid assays and co-immunoprecipitate with Snf1 from cell extracts.
We show that Nrg2, like Nrg1, functions as a transcriptional repressor. LexA87-Nrg2 represses transcription of a reporter containing LexA binding sites in glucose-grown cells, as previously reported for LexA87-Nrg1 (Park et al. 1999). We also examined the regulation of this repression and found that repression was not relieved by a shift from high to low (0.05%) glucose or by steady-state growth in sucrose, raffinose, or galactose. Correspondingly, glucoamylase gene expression is not derepressed by growth in these carbon sources (Pretorius et al. 1986; Kartasheva et al. 1996). An attempt to assess repression during growth in glycerol plus ethanol, a condition allowing derepression of glucoamylase genes, was inconclusive because the fusion proteins were not expressed well.
The finding that release from repression of STA genes occurs only when the carbon source is glycerol/ethanol is at first glance puzzling because the Snf1 kinase is also active during growth on sucrose, raffinose, and galactose. A plausible explanation comes from recent evidence that the Snf1 catalytic subunit and one of the β-subunits of the kinase, Gal83, are enriched in the nucleus when cells are grown on glycerol/ethanol but not when cells are grown on a fermentable carbon source (Vincent et al. 2001). The nuclear localization of Nrg1-GFP is consistent with the idea that nuclear import of Snf1 is required for inhibition of Nrg1 repressor function.
Together, the evidence suggests that Nrg1 and Nrg2 are either direct or indirect targets of Snf1. Previous studies showed that Nrg1 mediates glucose repression of glucoamylase genes (Park et al. 1999) and that Snf1 is required for release from repression (Kuchin et al. 1993). Here we present genetic evidence that Nrg1 and Nrg2 contribute modestly to glucose repression of DOG2 and that one of the roles of the Snf1 kinase in derepression of this gene is to inhibit repression by Nrg1 and Nrg2. We further show that Nrg1 and Nrg2 interact physically with the Snf1 kinase. Although we did not detect Snf1-dependent phosphorylation of Nrg1 or Nrg2, our data do not exclude the possibility that Snf1 phosphorylates these proteins in vivo. Alternatively, other as yet unidentified components may be involved in the regulatory mechanism; for example, Snf1 may phosphorylate another protein that is bound to and affects the function of Nrg1 and Nrg2.
Repression by Nrg1 and Nrg2 may be regulated at multiple steps. However, regulation of Nrg1 function does not appear to involve nuclear export; Nrg1-GFP was localized to the nucleus whether cells were grown in glucose or glycerol plus ethanol. Some control may be exerted at the RNA level. Park et al. (1999) found that NRG1 RNA levels are 6-fold lower in glycerol/ethanol than in glucose, but it has also been reported that NRG1 RNA is induced 2.7-fold during the diauxic shift (DeRisi et al. 1997). NRG2 RNA is neither significantly induced nor repressed during the diauxic shift (DeRisi et al. 1997).
The targets and physiological roles of Nrg1 and Nrg2 are still largely unknown, and it is possible that the repressor function of Nrg1 or Nrg2 is regulated in response to other signals besides carbon source. Expression of the NRG2 gene is clearly regulated by other signals. DNA microarray analysis of genomic expression patterns showed that NRG2 RNA levels are elevated five-fold in response to zinc limitation, and NRG2 has a potential binding site for the zinc-responsive transcription factor Zap1 (Lyons et al. 2000). NRG2 RNA is also induced by alkaline pH, and induction is partially dependent on the transcription factor Rim101, which is required for expression of various alkaline response genes (Lamb et al. 2001). The physiological significance of these findings remains to be clarified.
Acknowlegement
We thank Robert Townley and Olivier Vincent for help with localization studies, Tony Ferrante for assistance with microarray analysis, and Rong Jiang and Mark Johnston for plasmids. The Columbia University Microarray Project is supported by the Columbia Genome and Naomi Berrie Diabetes Centers. This work was supported by grant GM34095 from the National Institutes of Health (NIH) to M.C. V.K.V. also received support from NIH training grant T32GM08224.
Footnotes
Communicating editor: F. Winston
LITERATURE CITED