OUP user menu

Mitochondrial dysfunction enhances Gal4-dependent transcription

Branka Jeličić, Ana Traven, Vedrana Filić, Mary Sopta
DOI: http://dx.doi.org/10.1016/j.femsle.2005.09.033 207-213 First published online: 1 December 2005


Mitochondrial dysfunction has been shown to elicit broad effects on nuclear gene expression. We show here that transcription dependent on the prototypical acidic activator Gal4 is responsive to mitochondrial dysfunction. In cells with no mitochondrial DNA, Gal4-dependent gene expression is elevated. A minimal Gal4 activator containing the DNA binding and activation domain is sufficient for this response. Transcription dependent on a fusion of Gal4 to a heterologous DNA binding domain is similarly elevated in a mitochondrial mutant. Analysis of different Gal4-dependent promoters and gel mobility shift assays suggest that the effect of mitochondrial dysfunction on Gal4 activity is related to increased DNA binding to the cognate Gal4 element. Given that fermentation is the only means to obtain energy in respiratory deficient cells, it is possible that higher Gal4 activity in cells with dysfunctional mitochondria works to promote more efficient fermentation of galactose.

  • Yeast
  • Mitochondrial dysfunction
  • Gal4
  • Transcription

1 Introduction

Regulation of GAL gene expression in Saccharomyces cerevisiae is a paradigm for regulation of transcription by metabolic signals, reviewed in [1,2]. The protein products of the GAL genes are required for efficient galactose utilization and they are repressed by glucose and strongly activated by galactose. Transcription of the GAL genes is under the control of Gal4, a prototypical acidic activator, reviewed in [1,2]. When galactose is absent, the activity of Gal4 is repressed by Gal80, which binds to the activation domain of Gal4 thereby preventing its interactions with other components of the transcription machinery [3,4]. The galactose signal is transduced to the inactive Gal4–Gal80 complex by the inducer protein Gal3, which, in the presence of galactose, binds to Gal80 and relieves its inhibition of Gal4 [59]. In addition to being negatively regulated by Gal80, Gal4 is positively regulated by phosphorylation [1013]. More recently it has been shown that regulation of Gal4 protein stability and transcriptional activation also occurs via the ubiquitin degradation system [14].

Mitochondria are the central metabolic organelle in the cell and their functional state is directly involved in regulation of transcription factors and gene expression. Microarray analysis of whole genome expression in wild type versus mitochondrial mutants has shown a variety of coordinate changes in response to mitochondrial dysfunction [15,16]. The best studied transcriptional response to loss of mitochondrial function is the retrograde signaling pathway, which is activated by mitochondrial dysfunction and positively regulates the transcription factors Rtg1 and Rtg3, whose target genes are required for the metabolic adaptation of the cell to respiratory deficiency [15,1719]. Additionally, mitochondrial dysfunction has been shown to change transcription of the rRNA genes and genes encoding enzymes of the TCA cycle [20,21], convert the corepressor Tup1/Ssn6 to a coactivator of transcription [22] and increase the activity of Pdr3, a yeast transcription factor involved in pleiotropic drug resistance [23,24].

In this report, we were interested in analyzing more directly the influence of changes in mitochondrial status on transcription factor activity. To that end, we chose to examine the yeast Gal4 activator, since transcriptional mechanisms of Gal4 action are well studied and it is a system that is responsive to changes in metabolic requirements.

2 Materials and methods

2.1 Yeast strains and plasmids

The yeast strains used for assaying β and α-galactosidase activity and electrophoretic mobility shifts are W3031B (MATαade2-1 trp1 leu2-3 112 his3-11 15 ura3-1), GGY1gal4Δgal80 tyr1 ade leu2 his3 ura3) and NLY2 (Δgal4Δgal80 ura3 his3 leu2 trp1 lys2). Mutants that lack mitochondrial DNA (rho0) were induced by treatment with ethidium bromide, as described [25] and respiratory deficiency was confirmed by their inability to use glycerol as the sole carbon source. Reporters SV15 [26] and RJR227 [4] were integrated into the URA3 locus of GGY1. The reporter pSH18-34 [27] was transformed into the strain W3031B and its respective rho0 mutant. Plasmids expressing full length Gal4 (amino acids 1–881), or its deletion derivative containing the DNA binding domain (amino acids 1–100) fused to the activation domain (amino acids 840–881) are described in [4]. The plasmids are HIS3, CEN/ARS and Gal4 and its derivatives are expressed from the natural GAL4 promoter. Lex–Gal4 [27] was expressed from LEU2, CEN/ARS plasmids that contain LexA residues 1–202 fused to Gal4 (amino acids 74–881). For EMSA experiments a plasmid expressing short Gal4 under the control of the natural Gal4 promoter was used [4].

2.2 β-Galactosidase assays

For assaying β-galactosidase activities, yeast strains were transformed and grown in selective 2% galactose, 2% sucrose media to logarithmic phase. Specific β-galactosidase activities were determined from yeast cultures by the method of [28]. The assays were performed in 0.5 ml of reaction buffer and β-galactosidase units were calculated as described [28]. The assay with Lex–Gal4 in the rho0 mutant was performed in triplicate from two independent transformants. The standard deviation was <20%, except for the full length Gal4 in rho+ cells in Fig. 2(a), which was 22%. The activities are expressed as percentages of Gal4 activity in the wild type (rho+) strain, which is set to 100%.

Figure 2

Activation of a heterologous Gal4 fusion construct is elevated in a rho0 background. (a) The activity of Lex–Gal4 was assayed on reporter pSH18-34 transformed into rho+ or rho0 W3031B. The SE was ≤20%. (b) Western blot showing levels of Lex–Gal4 protein in the wild type and rho0 background.

2.3 Protein extracts and α-galactosidase assays

Cells from 10 ml of selective 2% galactose, 2% sucrose media were harvested at OD600= 1.0–1.4 and resuspended in 250 μl of breaking buffer (20 mM Tris–Cl, pH 8, 0.1 mM DTT, 10% glycerol). After addition of an equal volume of glass beads and 12.5 μl of 40 mM phenylmethanesulfonyl fluoride (PMSF), samples were vigorously vortexed (6 × 30 s). Extracts were cleared by centrifugation (15 min at 16,000 g, 4 °C) and 5–20 μl of supernatants were used for α-galactosidase assays performed as described in [29]. Units were calculated as: α-galactosidase units = 1000 × OD400/(0.0182 ×μg of proteins × time (min)). All assays were performed in triplicate from at least four independent colonies and the standard deviation was between 2% and 18%. The activity of the Gal4 construct in the wild type (rho+) strain is set to 100% and the other activities are expressed as percentages compared to the 100% activity in rho+ cells.

2.4 Northern analysis

Poly A+ mRNA was isolated from yeast cells (NLY2 harbouring a plasmid expressing full length Gal4) using the Qiagen Oligotex kit according to the manufacturer's protocol. 500 ng poly A+ mRNA was analyzed by Northern blotting as described [30]. A PCR fragment of the Gal4 gene was used as probe, generated with the following primers, 5′-CATGGCATCATTGAAACAGC-3′ and 5′-CAGGCAAAATATGGGGTGAC-3′. Northern blot analysis of the ACT1 transcript (a PCR fragment used as a probe was generated with following primers: 5′-TTTCAACGTTCCAGCCTTCT-3′ and 5′-TTGGTCAATACCGGCAGATT-3′) served as a loading control. Probes were labeled with [α-32P] dCTP using the Random primed DNA labeling kit (Roche). Unincorporated nucleotides were removed using Nick columns (Amersham Biosciences).

2.5 Western blot analysis

Protein extracts were prepared as described previously [31]. Western blots were performed as described previously [30]. 100 μg of protein extract preparation was analyzed. Proteins were resolved on a 7.5% SDS–polyacrylamide gel and transferred to a nitrocellulose membrane. The nitrocellulose membrane was incubated with a 1:10,000 dilution of anti-lexA monoclonal antibody (Santa Cruz Biotechnology), followed by a 1:5000 dilution of goat anti-mouse horseradish peroxidase-conjugated secondary antibody (Novagen). Signals were detected by enhanced chemiluminescence as per the manufacturer's protocol (Pierce).

2.6 Electrophoretic mobility shift assays

Protein extracts for EMSA, from GGY1 transformed with short Gal4 (1–100 + 840–881), were made in the same way as for α-galactosidase assays except for the difference in the breaking buffer (100 mM Tris-Cl, pH 8.0, 1 mM DTT and 20% glycerol). A double-stranded 32P-labeled oligonucleotide (5′-TCCGGAGGACTGTCCTCCGGT-3′) containing the consensus Gal4 site [4] was used as a probe. The same unlabeled oligo was used as a control cold competitor. A mutant oligo (5′-TCCAGAGGACTGTCCTCTGG-3′) from [32] was also used as a control competitor.

Gel electrophoretic mobility shift assays (EMSA) reactions were set up in 20 μl with 18 μg of protein extract, 20 fmol of an end-labelled probe in a 2× buffer containing 20 mM HEPES, pH 7.5, 300 mM NaCl, 1 mM MgCl2, 0.1 mM ZnCl2, 10% glycerol and 200 μg/ml bovine serum albumin. Binding reactions were incubated for 10–20 min at 4 °C and resolved at 180V for 2 h on a native pre-run 4% polyacrylamide gel (29:1) containing 0.5× TBE and 1% glycerol. The gel was dried and subjected to autoradiography.

3 Results

To address whether the functional state of mitochondria influences transcriptional activation by Gal4, the reporter plasmid SV15 was integrated into the URA3 locus of wild type (rho+) or the rho0 mitochondrial mutant of GGY1. As shown in Fig. 1(a), the activity of Gal4 was approximately 2-fold higher in the mitochondrial mutant compared to the wild-type strain.

Figure 1

The DNA binding and activation domains of Gal4 are sufficient for its regulation by mitochondrial signals. (a) Plasmids expressing full length Gal4 (residues 1–881) or Gal4 containing the DNA binding domain (amino acids 1–100) fused to the activation domain (amino acids 840–881) were transformed into the rho+ or rho0 of GGY1. Reporter SV15 was integrated into the URA locus and β-galactosidase assays were performed. (b) The positive effect of dysfunctional mitochondria on transcriptional activation by Gal4 using the endogenous MEL1 gene as the reporter. Plasmids expressing Gal4 (1–100 + 840–881) or Gal4 (1–881) were transformed into rho+or rho0 of GGY1 and α-galactosidase assay was performed. (c) Level of Gal4 transcript is unchanged in rho0 versus wild-type cells. 500 ng of poly A+ mRNA from rho+ and rho0 of NLY2 was hybridized with PCR generated probes for Gal4 and Act1.

The DNA binding and activation domains of Gal4 are sufficient for regulation of this activator by a number of signals [33]. We, therefore, analyzed whether they were also sufficient for regulation of Gal4 by mitochondrial dysfunction. As shown in Fig. 1a, the expression of lacZ in cells transformed with a plasmid expressing the DNA binding domain of Gal4 fused to the minimal activation domain (amino acids 1–100 + 840–881) was elevated in the rho0 mitochondrial mutant as compared to respiratory proficient cells (rho+). The construct expressing just the DNA binding domain (amino acids 1–100) showed no activity in either strain (data not shown).

We next tested whether this effect on Gal4 activity could be seen using the endogenous MEL1 gene as a reporter, where a single weak Gal4 binding site in the promoter determines Gal4-dependent transcription. For those experiments α-galactosidase activity was assayed in strain GGY1 transformed with plasmids expressing different versions of Gal4. As shown in Fig. 1b, both full length Gal4 (1–881) or the version containing just the DNA binding and activation domains (1–100 + 840–881) showed higher activity in rho0 compared to rho+ cells (1.7- and 2.7-fold, respectively). The control construct with the DNA binding domain alone (residues 1–100) again exhibited no activity in either strain (data not shown). This effect was observed when Gal4 was expressed either from its own natural promoter or expressed from the heterologous ACT1 promoter (data not shown).

To determine whether this upregulation was a function of increased transcription from the Gal4 promoter itself we performed Northern blot analysis examining the level of Gal4 transcript. Fig. 1c shows that the level of Gal4 transcript is unchanged in rho0 versus wild-type cells. We also tested for the ability of mitochondrial dysfunction to affect Gal4-dependent transcription from a reporter with multiple high affinity sites (reporter RJR227 with 5X 17 bp consensus Gal4 binding sites). No difference in transcriptional activity between rho+ and rho0 strains was observed for either the full length or short version of Gal4 (Table 1). In contrast, both Gal4 constructs showed higher activity in the mitochondrial mutant when just one strong consensus binding site was present in the UAS (reporter SV15, Fig. 1a) or when transcription of the MEL1 gene was assayed, which has a single weak Gal4 binding site (Fig. 1b). Thus, the presence of five high affinity consensus binding sites abrogates the effect of mitochondrial dysfunction on Gal4 transcription.

Figure Table 1

Transcriptional activation by Gal4 on a 5X Gal4 binding sites reporter

To test whether the Gal4 response to mitochondrial dysfunction is mediated via the activation domain or the DNA binding domain, we used a lex-Gal4 fusion construct to assay transcriptional activity. As seen in Fig. 2a the heterologous Gal4 fusion construct was responsive to mitochondrial dysfunction indicating that the response is specifically mediated through the Gal4 activation domain. Western blot analysis showed that the protein levels were equivalent in the wild type and rho0 backgrounds (Fig. 2b).

Since experiments with reporter constructs with different UAS structure indicated that mitochondrial dysfunction could regulate Gal4 activity by influencing occupancy of its binding sites, we sought to determine if Gal4 DNA binding activity increases in respiratory deficient cells. We used gel electrophoretic mobility shift assays with a 32P-labeled consensus Gal4 DNA binding site as a probe to examine the amount of DNA binding activity in wild type and rho0 strain backgrounds. Short Gal4 protein was expressed from its own natural promoter. As shown in Fig. 3, we observed that Gal4 binding activity was increased in the rho0 strain relative to the rho+ strain. A non-specific higher band was observed while the lower band represents Gal4 specific binding. This lower band was competed with cold wild-type competitor, but not competed when we used a mutant binding site (lanes 1, 2, 5 and 6, Fig. 3). Anti-Gal4 specific antibodies induced a supershift in the band corresponding to Gal4 binding to its cognate site (lane 7, Fig. 3) and this band was absent in a strain background devoid of Gal4 expression (lane 8, Fig. 3). This experiment suggests that mitochondrial dysfunction results in higher levels of Gal4 DNA binding activity which ultimately leads to higher Gal4-dependent transcription.

Figure 3

Mobility shift assay with short Gal4 (100 + 840–881). In each lane 18 μg of protein extract from GGY1 transformed with short Gal4 (1–100 + 840-881) was incubated with a 32P-labeled Gal4 consensus binding site and Gal4–DNA complexes were resolved on a 4% polyacrilamide gel. Lanes are as follows: 1 – rho+ cells + cold mutant competitor, 2 – rho+ cells + cold wild-type competitor, 3 – rho+ cells, 4 – rho0 cells, 5 – rho0 cells + cold wild-type competitor, 6 – rho0 cells + cold mutant competitor, 7 – rho0 cells + Gal4 antibody, 8 – negative control (strain lacking Gal4). The arrows indicate the position of the Gal4–DNA complex.

4 Discussion

In this report, we analyzed the influence of mitochondrial status on the activity of Gal4. We have established that mitochondrial dysfunction positively influences Gal4-dependent transcription and that the DNA binding and minimal activation domains of Gal4 are sufficient for the observed increase in transcriptional activity in mitochondrial mutants. Although the effect of compromised mitochondrial function on Gal4 activity is modest (approximately 2-fold), it is comparable to what has been shown for the Gal4 fusion to Rtg1, a transcription factor known to be involved in the mitochondria to nucleus retrograde signaling pathway [34]. To assess whether the effect on Gal4 might be an indirect effect whereby mitochondrial dysfunction upregulates transcription of the Gal4 gene itself we assayed transcript levels by Northern blot analysis and found that transcription of the Gal4 gene is not upregulated in a rho0 background relative to wild type.

The effect of dysfunctional mitochondria on Gal4 activity appears to depend on the number of binding sites in the UAS: when reporters that have one Gal4 binding site (a strong consensus site, as in SV15 or a weak site as in the UAS of MEL1) were assayed, Gal4-dependent transcription was higher in the mitochondrial mutant relative to wild type. However, this effect was not seen when Gal4 activity was assayed with a reporter containing five strong consensus Gal4 binding sites. We interpret this to mean that mitochondrial dysfunction increases Gal4-dependent transcription by regulating occupancy of the UAS by this activator. It has been shown that the degree to which Gal4 activates transcription depends on the degree of occupancy of its binding sites and one consensus binding site or the weak site from the MEL1 UAS are not completely saturated in vivo [4,35]. Mitochondrial dysfunction might act to increase binding site occupancy by Gal4 resulting in more efficient activation. When multiple strong consensus sites are present (as in RJR227) Gal4 binds cooperatively and saturates the UAS, resulting in maximal activity in rho+ cells that cannot be further increased by respiratory deficiency. This conclusion is supported by our EMSA results, which show that the level of Gal4 DNA binding activity is higher in rho0 cells. Alternatively, increased DNA binding to the Gal4 consensus site as measured by EMSA may reflect an increase in Gal4 protein levels. It remains to be determined whether the signaling of mitochondrial dysfunction to Gal4 involves post-translational events such as phosphorylation or ubiquitination which could potentially regulate protein stability or DNA binding.

We also tested the ability of a fusion of the Gal4 activation domain with the bacterial lexA DNA binding domain for responsiveness to mitochondrial dysfunction. We observed that such a construct is responsive to mitochondrial dysfunction indicating that the signaling pathway acts specifically via the Gal4 activation domain. The Gal4 activation domain mediates several functions in the transcription process including recruitment of cofactors and general initiation factors, regulation of protein stability and activation via ubiquitination, as well as regulation of activation via phosphorylation. It is possible that mitochondrial dysfunction impinges on any one or more of these processes to elicit increased transcriptional activation.

Our results suggest that Gal4 is another transcription factor regulated by the functional state of mitochondria and that the activation domain is the primary target of the signaling pathway. In the case of Gal4 this signal leads to increased protein stability and/or DNA binding which ultimately results in increased Gal4-dependent transcription. Studies of the Rtg1-3 pathway show that mitochondrial dysfunction activates a transcriptional response that mainly acts to reprogram metabolic pathways in order for the cells to maintain essential cellular functions [36]. The reason for enhancing Gal4 activity in cells in which respiration is compromised is not obvious, but it could be to increase fermentation of galactose, since rho0 cells can obtain energy exclusively via non-respiratory pathways.


We thank Drs. M. Ptashne, R. Reece, K. Struhl and F. Winston for strains and plasmids and Dr.W. Schaffner for support. This work was supported by a grant to MS from the Croatian Ministry of Science and Technology and a SCOPES grant from the Swiss National Science Foundation (7KRPJ065578).


  • 1 These authors contributed equally to this work.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
View Abstract