OUP user menu

Repression vs. activation of MOX, FMD, MPP1 and MAL1 promoters by sugars in Hansenula polymorpha: the outcome depends on cell's ability to phosphorylate sugar

Sandra Suppi, Tiina Michelson, Katrin Viigand, Tiina Alamäe
DOI: http://dx.doi.org/10.1111/1567-1364.12023 219-232 First published online: 1 March 2013


A high-throughput approach was used to assess the effect of mono- and disaccharides on MOX, FMD, MPP1 and MAL1 promoters in Hansenula polymorpha. Site-specifically designed strains deficient for (1) hexokinase, (2) hexokinase and glucokinase, (3) maltose permease or (4) maltase were used as hosts for reporter plasmids in which ß-glucuronidase (Gus) expression was controlled by these promoters. The reporter strains were grown on agar plates containing varied carbon sources and Gus activity was measured in permeabilized cells on microtitre plates. We report that monosaccharides (glucose, fructose) repress studied promoters only if phosphorylated in the cell. Glucose-6-phosphate was proposed as a sugar repression signalling metabolite for H. polymorpha. Intriguingly, glucose and fructose strongly activated expression from these promoters in strains lacking both hexokinase and glucokinase, indicating that unphosphorylated monosaccharides have promoter-derepressing effect. We also show that maltose and sucrose must be internalized and split into monosaccharides to exert repression on MOX promoter. We demonstrate that at yeast growth on glucose-containing agar medium, glucose-limitation is rapidly created that promotes derepression of methanol-specific promoters and that derepression is specifically enhanced in hexokinase-negative strain. We recommend double kinase-negative and hexokinase-negative mutants as hosts for heterologous protein production from MOX and FMD promoters

  • cell permeabilization
  • glucokinase
  • glucose repression
  • glucose signalling
  • hexokinase
  • methylotrophic yeasts


Yeasts live in a sugar-rich environment and prefer sugars over other carbon sources. When a high amount of glucose is available, utilization of alternative carbon sources such as alcohols and organic acids is prevented through transcriptional down-regulation of the respective genes. This mechanism is called glucose repression. To trigger the repression, cells must first detect a sugar. In Saccharomyces cerevisiae, two transporter-like transmembrane proteins, SNF3 and RGT2, with long C-terminal cytosolic extensions sense respectively low and high external glucose concentrations. Binding of glucose to the sensor initiates a signalling pathway that controls expression of glucose transporter genes (reviewed in Gancedo, 2008). Glucose reaching the cell provokes repression of target promoters by a mechanism in which MIG1 protein and one of hexokinase isoforms, HXK2, have specific roles in S. cerevisiae. At glucose abundance, HXK2 and MIG1 proteins enter the nucleus, where MIG1 directs corepressors TUP1 and CYC8 to target promoters. If glucose is depleted, SNF1 protein kinase will inactivate MIG1 by phosphorylation whereby it will be exported from the nucleus, allowing transcriptional activation of the promoters (Ahuatzi et al., 2004; 2007 ,Barnett & Entian, 2005; Gancedo, 2008; Pelaez et al., 2010). At high glucose, HXK2 stabilizes the repressor complex by interfering with MIG1 phosphorylation by SNF1 protein kinase (Ahuatzi et al., 2007). How the presence of glucose is sensed inside the S. cerevisiae cell is still an open question.

Glucose also affects gene expression in methylotrophic yeasts. At growth of these yeasts on methanol, promoters of methanol-specific genes are extremely strongly induced. This is why respective promoters are widely used for heterologous protein production in methylotrophic yeasts (Gellissen, 2000; Hartner & Glieder, 2006). In the presence of abundant glucose, methanol-specific functions of cells are inhibited and methanol is not used (Eggeling & Sahm, 1980; Sibirny et al., 1988; Hartner & Glieder, 2006; Yurimoto & Sakai, 2009; Yurimoto et al., 2011). However, in the case of Hansenula polymorpha, some methanol-specific promoters are also remarkably derepressed at glucose-limited growth (Mayer et al., 1999), which allows toxic and flammable methanol to be excluded from the protein production scheme. Glucose repression mutants of H. polymorpha have also been used to produce heterologous proteins from methanol-specific promoters (Krasovska et al., 2007).

Data on sensing, transport and metabolism of sugars in H. polymorpha are accumulating. Stasyk et al. (2004, 2008) have described two transmembrane sensor proteins GCR1 and HXS1 in this yeast. Whereas the GCR1 has no cytosolic extension (Stasyk et al., 2004), the HXS1 is a typical sensor protein with a long C-terminal signalling domain (Stasyk et al., 2008). One hexose transporter gene, HXT1, which encodes a permease for glucose and fructose, has also been cloned from H. polymorpha (Stasyk et al., 2008). We have characterized transport of glucose (Karp & Alamäe, 1998), maltose and sucrose (Viigand & Alamäe, 2007) in H. polymorpha, and cloned the genes for glucokinase (Laht et al., 2002), hexokinase (Karp et al., 2003), maltase (Liiv et al., 2001) and maltose (α-glucoside) permease (Viigand & Alamäe, 2007). Data in the literature suggest that sugar repression in H. polymorpha cannot be described on the basis of the baker's yeast model. The first evidence of that came from enzymatic assay of hexokinase-negative mutants of H. polymorpha LR9 (Kramarenko et al., 2000). These mutants retained glucose repression, meaning that hexokinase protein has no specific regulatory role in glucose repression. Thereafter, Oliveira et al. (2003) showed that the TUP1 homologue of H. polymorpha had no role in glucose repression of alcohol oxidase, dihydroxyacetone synthase and catalase. In 2007, Stasyk et al. reported that disruption of MIG1 and MIG2 genes of H. polymorpha also had only a minor effect on glucose repression of alcohol oxidase.

In the current work we proceed with the study of sugar repression in H. polymorpha. We will analyze expression from four sugar-repressed promoters in H. polymorpha strain 201 applying a simple Gus reporter assay on a set of mutants. Aside of well-known MOX and FMD promoters of H. polymorpha, we included in the assay two less studied, but biotechnologically perspective, novel promoters, those of peroxisomal activator gene MPP1 (Leao-Helder et al., 2003) and the maltase gene MAL1 (Liiv et al., 2001). Novel and intriguing data on promoter regulation in H. polymorpha will be presented.

Materials and methods

Strains, plasmids and oligonucleotide primers

Escherichia coli strain DH5α [supE44 ΔlacU169 (ø80 lacZΔM15) recA1 endA1 hsdR17 thi-1 gyrA96 relA1] (Invitrogen) was used for DNA manipulation procedures.

Hansenula polymorpha strain 201 (HXK1 GLK1 ura3-1 leu2-2 met4-220; Lahtchev et al., 2002) isogenic to CBS 4732 was used as a wild type. In this strain, maltase-disruption mutant mal1 and maltose permease disruption mutant mal2 were constructed by us earlier (Alamäe et al., 2003; Viigand & Alamäe, 2007). The plasmids used in this work and their construction are described in Table 1. The primers used in this work are presented in Supporting Information, Table S1.

View this table:


pRS426HXK1Contains a 2569-bp SpeI-SalI fragment from pJ1 with HXK1 gene of Hansenula polymorpha (Karp et al., 2003) cloned between the same sites of pRS426 polylinkerThis work
pRS426HXK1::URA3The H. polymorpha URA3 gene was amplified from pHpURA3 (from Dr J. Siverio, Spain) using primers HpURA3FwNheI and HpURA3RevNheI, the product was cleaved with NheI and inserted into XmaJI (AvrII) site of HXK1 gene in pRS426HXK1This work
pYT3HXK1Contains a 2569-bp SpeI-SalI fragment from pJ1 with H. polymorpha hexokinase gene HXK1 between the XbaI and SalI restriction sites of pYT3 (Tan et al., 1995)Karp et al. (2003)
pYT3GLK1Contains the H. polymorpha glucokinase gene GLK1 in the XbaI site of the pYT3 (Tan et al., 1995)Laht et al. (2002)
pX4-HNBESXPlasmid with H. polymorpha MOX promoter and AMO terminator, kanR and ScLEU2From Dr. Kiel (Gröningen)
pHIPX8Plasmid with H. polymorpha TEF2 promoter and AMO terminator, kanR and ScLEU2From Dr Kiel (Gröningen)
pHIPMALpromA derivative of pHIPX8 in which the TEF2 promoter is replaced by H. polymorpha MAL1 promoter.Visnapuu et al. (2008)
pHIPMALpromGusGus reporter under the control of H. polymorpha MAL1 promoter. Promoterless gusA gene of E. coli was excised from pIB-GusA (Sears et al., 1998) on a BamHI-PstI fragment and inserted between the same sites in pHIPMALpromThis work
pX4GusAGus reporter under the control of H. polymorpha MOX promoter. Promoterless gusA gene of Escherichia coli was excised from pGUS102 (Marits et al., 2002) on an EcoRI fragment and inserted into the EcoRI site of pX4-HNBESXThis work
pHIPFMDpromGusGus reporter under the control of H. polymorpha FMD promoter. The promoter region (650 bp) of the FMD gene was PCR-amplified from genomic DNA of H. polymorpha 201 using primers FMDpromFwNotI and FMDpromRevBamHI. The obtained fragment, cut with and NotI and BamHI, was cloned between the same sites of pHIPMALpromGus to replace the MAL1 promoterThis work
pHIPMPP1promGusGus reporter under the control of H. polymorpha MPP1 promoter. The promoter region (776 bp) of MPP1 gene was PCR-amplified from genomic DNA of H. polymorpha 201 using primers MPP1promFwNotI and MPP1promRevBglII. The obtained fragment, cut with NotI and BglII, was cloned between the NotI and BamHI sites of pHIPMALpromGus to replace the MAL1 promoterThis work

Disruption of HXK1 gene in H. polymorpha 201, obtaining the hexokinase-negative strain hxk1

HXK1 gene was inactivated in H. polymorpha by homologous recombination with a disruption fragment. For that, a linear 3773-bp fragment was PCR-amplified from pRS426HXK1::URA3 using primers HK13 and HK21, purified and electroporated to wild-type cells. Uracil-positive clones that grew on glucose, glycerol and methanol, but not on sorbitol or fructose, were obtained. Genomic DNA of three clones was isolated and disruption of genomic HXK1 locus was verified by PCR. Extracts of glucose-grown cells of tested isolates did not phosphorylate fructose and displayed no hexokinase band in native polyacrylamide gel electrophoresis (PAGE). Complementation of H. polymorpha hxk1 strains with pYT3HXK1 restored fructose and sorbitol growth to the mutants and hexokinase band was again revealed at PAGE of extracts.

Deletion of the GLK1 gene in hexokinase-negative H. polymorpha, obtaining the double kinase-negative strain hxk1 glk1

We intended to use the URA3 marker once again to disrupt the GLK1 gene in hxk1 H. polymorpha. For that, a PCR-amplified DNA fragment containing the H. polymorpha URA3 gene bordering with GLK1 sequences was electroporated to H. polymorpha hxk1 strain and the cells were plated out onto agar medium containing 2% glycerol and 200 mg L−1 of 2-deoxy-d-glucose (2DG). Double kinase-negative H. polymorpha strains can grow on glycerol or methanol in the presence of 2DG (Kramarenko et al., 2000) and we hoped to promote GLK1 disruption using selective force of 2DG. In this selection, we isolated a strain with a 595-bp deletion in the GLK1 open reading frame as was verified by genomic DNA sequencing. The obtained strain designated as hxk1glk1 did not grow on sugars and retained its leu2 selection marker for introduction of reporter plasmids. Extract of glycerol-grown hxk1glk1 strain had no detectable glucose and fructose phosphorylating activity and native electrophoresis of the extract revealed no bands of hexokinase and glucokinase. As expected, complementation of hxk1glk1 strain with pYT3HXK1 restored its growth on glucose and fructose, and complementation with pYT3GLK1 restored growth on glucose. PAGE of cell extracts of complemented strains agreed with growth phenotype. The hxk1 and hxk1glk1 strains and the wild type (HXK1 GLK1) were used in this study for the assay of sugar-related regulatory events in H. polymorpha 201.

Cultivation of yeasts and bacteria

Yeasts were grown for reporter assay on 0.67% Yeast Nitrogen Base (YNB) medium without amino acids (Difco) to which 2% agar and adequate auxotrophic supplements were added. Carbon sources were used at concentrations shown in the text. All sugars were autoclaved separately in distilled water and then added to the medium. The cells pregrown on 2% glycerol were streaked onto three sectors of agar plate (see insert of Fig. 1) containing various carbon sources. Standard-size Petri dishes with 20 mL of the agar medium were used. One Petri dish was used for each transformant and at least three different transformants of the strain were grown and analyzed in each reporter assay. Incubation temperature was 37 °C and the growth time was either 2 or 3 days. If the cells were grown in liquid culture, batch cultivation in flasks on orbital shaker was used. In tests of yeast growth on YNB agar medium with 1% methanol + 2DG, the latter was added at concentrations 200, 400 or 1000 mg L−1. Escherichia coli was grown at 37 °C in Luria–Bertani medium containing kanamycin (0.1 mg mL−1) when required.


Gus expression from MOX, FMD and MPP1 promoters in wild-type Hansenula polymorpha grown on agar plates containing 2% glucose (glc), 2% glycerol (gly) or 1% methanol (met) as carbon sources. Growth time of transformants (reporter strains) was 2 days in the case of glucose and glycerol and 3 days in case of methanol. Average Gus activities in Miller units and standard deviations for at least three transformants are indicated. The insert shows the agar plate with three seeded sectors of a reporter strain.

DNA manipulations, sequencing, PCR and transformation

DNA manipulations were carried out using standard methods (Sambrook et al., 1989). Plasmid DNA was purified with AxyPrep™ kit (Axygen Biosciences). DNA fragments for cloning and gene disruption were purified using UltraClean 15 kit from MoBio. Yeast genomic DNA was isolated as in Liiv et al. (2001). Restriction endonuclease digestions and DNA ligations were performed according to the manufacturer's (Fermentas) recommendations. In PCR-based cloning of promoter regions, the Pfu DNA polymerase (Fermentas) was used. DNA was sequenced using an ABI PrismTM 377 DNA sequencer (Perkin Elmer) and DYEnamic™ ET terminator cycle sequencing kit (Amersham). Hansenula polymorpha was electrotransformed according to Faber et al. (1994) using a BioRad Xcell GenePulser.

Preparation of yeast suspensions and assay of β-glucuronidase (Gus) activity in permeabilized cells on microtitre plates

The cells grown on agar plates were harvested using a plastic inoculation loop, washed twice in K-phosphate buffer (50 mM, pH 7.0) by centrifugation, resuspended in the same buffer to optical density (OD)600 nm ∼10 and the cell suspension was kept on ice during the experiment. For Gus activity assay, 20 μL of this suspension were pipetted in triplicate into wells of a 96-well transparent flat-bottom microtitre plate (655101, Greiner bio-one, Germany) and 80 μL of the buffer (50 mM K-phosphate, 0.01 mM EDTA, pH 7.0) containing 0.1% cetyl trimethylammonium bromide (CTAB) as a permeabilizing agent (Alamäe & Järviste, 1995) was added to each well. The plate was gently rocked for 20 min for cell permeabilization and then the β-glucuronidase reaction was initiated by adding 10 μL of 10 mM p-nitrophenyl-β-d-galacturonide (p-NPG). At the appearance of pale yellow colour, the reaction was stopped by adding 50 μL of 1 M Na2CO3. Reference samples consisted of permeabilized cells to which 50 μL of 1 M Na2CO3 and 10 μL of p-NPG were added simultaneously. Absorbance of the samples was measured at 405 nm, using a Tecan Sunrise™ microplate reader (Tecan Group Ltd., Switzerland) and Magellan™ data analysis software (Tecan). Both the cell permeabilization and enzyme assay were performed at room temperature.

Gus activity was calculated in Miller units according to the formula: Embedded Image where ΔOD405 min−1 refers to the absorbance of reaction mixture at 405 nm from which the absorbance of reference sample was subtracted, divided by the reaction duration in minutes. OD600 is the absorbance of the prepared yeast suspension at 600 nm, measured after suitable dilution in a standard 1-cm pathway-length cuvette and Vsusp is the amount of suspension (mL) used in the reaction.

Enzyme assay in cell extracts

Preparation of cell extracts, determination of protein concentration and hexokinase activity are described in Karp et al. (2003). Maltase and β-glucuronidase assays are described in Viigand et al. (2005). For every assay, at least three parallel samples from at least three transformants were prepared and analyzed. Enzyme activity was expressed in either nanomoles or micromoles of substrate converted per min (mU and U, respectively) per mg of protein in the reaction mixture.

Determination of glucose concentration in the medium and inside the cells

To assess residual glucose concentration in the agar medium, the cells were scraped off, the agar plate surface was washed once with distilled water and small discs were excised from it using sterile 1-mL pipette tips (Putrinš et al., 2011; see also Fig. 4c). The discs were cut from two regions: adjacent to yeast growth area and from underneath of it. The disks were melted at 100 °C and cooled to 65 °C. Glucose content in melted agar was determined with a Glucose Liquicolor kit (Human GmbH, Germany) according to the instructions of the producer. The Glucose Liquicolor assay was also used to determine glucose concentration in supernatants of liquid cultures and inside the cells. The extracts for intracellular glucose assay were prepared according to Miseta et al. (2003). In calculations, 1 mg of dry weight was taken to correspond to 2 μL of cell water (Guijarro & Lagunas, 1984). Cell dry weight was measured as in Karp & Alamäe (1998). Presence of glucose in agar plates was visualized by overlay of agar surface with 5 mL of 50 mM Tris-buffer (pH 7.5) containing 0.1% agarose, 100 U mL−1 of glucose oxidase and 60 μg mL−1 of both N-methylphenazonium ethanesulphonate and nitroblue tetrazolium.

Results and discussion

Our toolbox

In earlier experiments on sugar repression in H. polymorpha, we used the LR9 strain and its mutants that were derived by chemical mutagenesis. Enzymatic analysis of these mutants suggested that hexokinase protein has no specific role in glucose repression in this yeast and referred to some metabolite as sugar repression triggerer (Kramarenko et al., 2000; Karp et al., 2003). In the current study, we used targeted mutagenesis to design stable hxk1 and hxk1glk1 mutants of H. polymorpha 201. These two mutants and earlier constructed mal1 (Alamäe et al., 2003) and mal2 (Viigand & Alamäe, 2007) mutants of strain 201were used as hosts for reporter plasmids. In these plasmids, β-glucuronidase gene of E. coli was cloned under control of MOX, FMD, MPP1 and MAL1 promoters of H. polymorpha. The transformants (reporter strains) were grown on agar medium of varied carbon source composition and β-glucuronidase (Gus) activity of permeabilized cells was quantified on microtitre plates. Notably, we have earlier used microtitre plate-based assay on permeabilized cells for semiquantitative measurement of alcohol oxidase and maltase activities (Alamäe & Liiv, 1998; Kramarenko et al., 2000) and for quantitative assay of levansucrase activity (Alamäe et al., 2012). Due to simple cultivation mode and high-throughput enzymatic assay we could address a large number of strains, promoters and carbon sources. As will be presented further, agar plate cultivation creates nutrient limitation conditions that are crucial for promoter derepression.

MPP1 promoter is regulated by carbon sources similarly to MOX and FMD promoters

The MPP1 gene was described as a regulator of peroxisomal protein levels of H. polymorpha. It encodes a putative transcriptional activator of 684 amino acids with N-terminal zinc-finger domain. In mpp1 mutants grown on glycerol + methanol, the amount of alcohol oxidase and several peroxins was strongly reduced compared with wild type, and dihydroxyacetone synthase was absent. A MPP1-GFP fusion protein was absent from H. polymorpha cells grown on glucose, but was present and detected in the nucleus in methanol-grown cells (Leao-Helder et al., 2003). As the MPP1 gene was most induced after the transfer of H. polymorpha from glucose to methanol (van Zutphen et al., 2010), its promoter was expected to have biotechnological potential in regulated overexpression of proteins in H. polymorpha.

First, we compared regulation of MOX, FMD and MPP1 promoters by carbon sources in wild-type H. polymorpha. Figure 1 shows that Gus expression from all three promoters was highest at methanol growth with strongest activation recorded for the MOX promoter (∼700 Miller units) and about twice less for FMD and MPP1 promoters. These results contradict the transcriptomic assay data by van Zutphen et al. (2010) according to which the MPP1 showed the highest methanol induction (394-fold), FMD the second best induction (347-fold) and the MOX only a 17-fold induction. Our data show that at growth on 2% glycerol or 2% glucose, the MOX and FMD promoters had low expression, and the MPP1 promoter-driven Gus activity was below detection (Fig. 1). All three promoters were significantly derepressed at growth of cells on 0.1% of glucose. The highest glucose derepression (∼470 Miller units) was recorded for the MOX promoter. Notably, methanol-induced and glucose-derepressed Gus expression from FMD promoter were quite the same.

In methylotrophic yeasts, aside from MPP1 protein, three methanol-specific zinc finger activators have been described: MXR1 in Pichia pastoris (Lin-Cereghino et al., 2006) and TRM1 and TRM2 in Candida boidinii (Sasano et al., 2008, 2010). TRM1 was considered a master transcriptional regulator of methanol-specific gene activation (Sasano et al., 2008). TRM2, which is homologous to MXR1 of P. pastoris and ADR1 of S. cerevisiae, is essential for TRM1-dependent activation. As described in Sasano et al. (2008), the MUT3 protein of H. polymorpha that has high sequence similarity to TRM1 of C. boidinii may have a role in activation of methanol-specific genes in H. polymorpha. The mut3 strain of H. polymorpha could not grow on methanol and had reduced MOX promoter activity (Vallini et al., 2000). Unfortunately, the MUT3 protein itself has not been studied.

Hexokinase-negative H. polymorpha is insensitive to fructose repression and has enhanced derepression of methanol-specific promoters at low glucose growth

Glucose and fructose repression of MOX, FMD and MPP1 promoters was assayed in the wild type and hexokinase-negative strains of H. polymorpha 201. In both strains, MOX, FMD and MPP1 promoters were not induced by methanol in the presence of 2% glucose. In hxk1 H. polymorpha 201, glucose but not fructose repressed methanol-specific promoters, which agrees with our earlier data on H. polymorpha LR9 (Kramarenko et al., 2000; Karp et al., 2003). Figure 2 illustrates the data on MOX and FMD regulation. According to the literature, the MOX and FMD genes are significantly derepressed when cells are growing at glucose limitation (Egli et al., 1980; Kensy et al., 2009). We presumed that promoter derepression may be higher for hxk1 H. polymorpha. Figure 3 shows that this is indeed the case. In both strains, the studied promoters were equally repressed at growth of cells on 2% glucose and equally slightly derepressed at growth on 0.5% glucose. At growth on 0.1% glucose, the hxk1 strain exhibited much higher derepression than the wild type. When comparing the promoters, the highest derepression was recorded for MOX and the lowest for MPP1 promoter. According to our results, the MOX promoter was more strongly activated at methanol growth and more highly derepressed at glucose limitation compared with the FMD promoter.


Gus expression from MOX and FMD promoters in wild-type Hansenula polymorpha (WT) and hexokinase-negative mutant (hxk1) grown for 3 days on agar plates containing either 2% glucose (glc) or 2% fructose (fru) and 1% methanol (met) as carbon sources. Average Gus activities in Miller units and standard deviations of at least three transformants are indicated.


Gus expression from MOX, FMD and MPP1 promoters in wild-type Hansenula polymorpha (WT) and hexokinase-negative mutant (hxk1) grown on agar plates containing varied concentrations of glucose (glc). The growth time was 2 days. Average Gus activities in Miller units of at least three transformants and standard deviations are presented.

We next asked whether addition of methanol to low glucose (0.1%) medium would further enhance expression from these promoters. It turned out that in the wild type, addition of methanol almost doubled the expression from all three promoters, whereas in the hxk1 strain, no further increase of reporter expression was detected (data not shown). Significant derepression of methanol-specific promoters in media lacking methanol is a specific feature of H. polymorpha (Hartner & Glieder, 2006) that can be implemented in H. polymorpha-based heterologous protein production systems. So, a very high expression level (13.5 mg L−1) was achieved in production of secreted phytase in H. polymorpha from FMD promoter by shifting glycerol-pregrown cells to glucose limitation (Mayer et al., 1991). Our results show that hexokinase-negative mutant of H. polymorpha should be superior to the wild type in a similar type of expression system using, for example, the MOX or FMD promoter.

Glucose limitation is created at growth of yeast on glucose-containing agar plates

Figure 3 shows derepression of methanol-specific promoters when H. polymorpha is grown on low glucose agar medium. In contrast to Eggeling & Sahm (1980), we could not detect derepression of methanol-specific enzymes at liquid batch cultivation of H. polymorpha on either 0.1% or 0.2% of glucose, even in late stationary growth phase (our unpublished data). As derepression was disclosed at growth on agar plates, we suggested that this cultivation mode created glucose limitation conditions. Figure 4 indicates that this is indeed the case. Panel D of Fig. 4 demonstrates rapid exhaustion of glucose from the agar medium under the cells, becoming very low (∼9 mM) already on second day of growth, whereas in the periphery of the agar plate it still remains high at that time. Thus, yeasts growing on glucose-containing agar plate certainly face glucose limitation. Glucose limitation has been created by cultivation of H. polymorpha in a chemostat run at a low dilution rate (Egli et al., 1980). Recently, a highly interesting application to create glucose limitation at batch cultivation was described for H. polymorpha. In that application, liquid cultures were supplied with silicon discs from which glucose was released slowly (Kensy et al., 2009; Scheidle et al., 2010). Our current approach is certainly much simpler than the above-mentioned methods.


Glucose limitation is created at growth of yeast on glucose-containing agar medium. Wild-type Hansenula polymorpha was grown on small (diameter 4 cm) Petri plates with YNB agar containing 2% (111 mM) glucose as shown in (a). On days 2, 4, 6 and 8, cells were scraped off the agar, the surface of the agar plate was gently washed with distilled water, and discs were cut off from the medium for determination of glucose concentration. Two discs were excised from under the cells (the area encircled by broken line) and two from the adjacent area as indicated in (c). The time course of glucose consumption is shown in (d). On day 4, the presence of glucose in the medium was developed in a chromogenic assay (b).

Glucose-6P as potential signal metabolite in triggering sugar repression

Our experimental data predict that glucose-6P is a mediator of sugar repression in H. polymorpha. This prediction is supported by following facts: (1) phosphorylation of a sugar by hexokinase or glucokinase is required for repression (Kramarenko et al., 2000; see also Fig. 2); (2) a glucose analogue 2DG mimics the effect of glucose as a repressor despite it not being metabolized beyond the phosphorylation step (see further). According to our hypothesis, Glc6P also signals for fructose repression, because after phosphorylation of fructose in hexokinase reaction it can be isomerized to Glc6P. According to our regulation model, H. polymorpha is capable of sensing intracellular Glc6P concentration and responds accordingly by transcriptional down-regulation of target genes, for example those of methanol and disaccharide catabolism (see Fig. 5).


Proposed position of glucose-6-phosphate (Glc6P) in sugar repression signalling pathway in Hansenula polymorpha.

To find out whether Glc6P production rate may depend on sugar concentration in the medium as well as on the hexose kinase pattern of the strain, we measured glucose phosphorylation in cell extracts. The wild-type H. polymorpha 201 and hxk1 strains were grown in liquid culture on high (2%) and low (0.2%) glucose. Table 2 shows that the glucose phosphorylating activity of cells depends on glucose concentration in the medium being much lower in the case of low glucose. Also, the glucose-phosphorylating activity of the hxk1 strain was about twice as low as that of the wild type. We consider that low glucose-phosphorylating activity means a reduced amount of Glc6P in the cell, allowing promoter derepression. Hexokinase-negative mutant does not phosphorylate fructose and is therefore not capable of producing Glc6P. According to our hypothesis, this feature explains the lack of fructose repression in hkx1 H. polymorpha (Figs 2 and 5). The growth pattern of strains on methanol + 2DG medium agrees with our model. The wild-type, hxk1 and hxk1glk1 strains were streaked onto YNB agar containing 1% methanol at different concentrations (200, 400 and 1000 mg L−1) of 2DG. Only hxk1glk1 strain was able to grow.

View this table:

Glucose phosphorylation in cell extracts of wild-type and hexokinase-negative strains of Hansenula polymorpha grown at different glucose concentrations

StrainCarbon source in the mediumGlucose phosphorylation, U mg−1
Wild type (WT)2% glucose1.60 ± 0.12
Wild type (WT)0.2% glucose0.47 ± 0.01
Hexokinase-negative (hxk1)2% glucose1.05 ± 0.11
Hexokinase-negative (hxk1)0.2% glucose0.21 ± 0.02
  • Cells for preparing the extracts were grown in liquid batch culture and harvested for enzymatic assay in the mid-exponential growth phase. Mean values ± standard deviation are presented for two separate cultures of the wild type and three different hxk1 strains.

2DG is the substrate for the H. polymorpha glucose transport system (Karp & Alamäe, 1998) and for hexokinase and glucokinase as well (Laht et al., 2002; Karp et al., 2003). 2DG does not provide growth, and it cannot be isomerized and converted to Fru6P (Klein & Stitt, 1998). We interpret the repressive effect of 2DG on methanol growth as follows. 2DG is transported to H. polymorpha cells and phosphorylated by hexokinase and glucokinase producing 2DG-6P, which mimics the effect of repression mediator Glc6P. Therefore methanol-specific promoters will be repressed making growth on methanol impossible. Double kinase-negative mutants do not phosphorylate 2DG, and they can therefore induce methanol-specific functions and grow on methanol in the presence of 2DG.

Disaccharides repress expression of methanol-specific promoters only if transported into the cell, hydrolyzed to monosaccharides and phosphorylated

Disaccharides, maltose and sucrose, repress methanol-specific promoters. Having in hand disruption mutants of maltase and maltose permease genes, we asked two questions: (1) should a disaccharide be transported into the cell to exert repression and (2) should a disaccharide be hydrolyzed inside the cell to exert repression? To get the answer, reporter analysis of MOX promoter was carried out on strains cultivated on methanol + a disaccharide. Figure 6 shows that the MOX promoter is repressed by maltose and sucrose in the wild type. If either permease or maltase gene is inactivated, the repressing ability of a disaccharide is abolished and the MOX promoter can be induced by methanol. In mutants grown on maltose + methanol, the MOX promoter was more highly induced than at sucrose + methanol growth. We do not know the reason for that. However, measurement of alcohol oxidase activity of same cells disclosed the same trend (data not shown).


Gus expression from the MOX promoter in wild-type Hansenula polymorpha (WT), maltose permease-negative mutant (mal2) and maltase-negative mutant (mal1) harbouring the reporter plasmid with the MOX promoter. Cells were grown on agar media containing 2% sucrose + 1% methanol and 2% maltose + 1% methanol for 3 days and Gus activity was measured in cells. Average activities in Miller units and standard deviation are shown for three transformants of each strain.

Next, we inspected the MOX promoter repression at sucrose growth of H. polymorpha strains. As hxk1 H. polymorpha cannot metabolize fructose moiety of sucrose, this disaccharide was predicted to cause less repression on MOX promoter in this strain. For reporter assay, the strains were grown on (1) 1% sucrose and (2) 0.5% glucose + 0.5% fructose. At growth on both media, Gus activity of the wild type was very low, below 9 Miller units. As expected, the hxk1 strain had much higher Gus activity (up to 78 Miller units) at growth on both media. We conclude from these results that to repress a methanol-specific promoter, maltose and sucrose should be transported into the cell, split by maltase to monosaccharidic products and then phosphorylated by hexokinase and/or glucokinase.

In double kinase-negative H. polymorpha, glucose activates the promoters that are considered glucose-repressible

We have earlier recorded a peculiar feature of double kinase-negative mutants of H. polymorpha LR9 – the cells grown on methanol + glucose had much higher alcohol oxidase activity than those grown on methanol alone (Kramarenko et al., 2000; Laht et al., 2002). Here, we reinvestigated this matter in the hxk1glk1 strain of H. polymorpha 201.

Figure 7 shows a typical glucose repression phenotype for the wild type: 2% glucose in the medium prevents induction of MOX, FMD and MPP1 promoters by methanol, whereas 0.1% glucose allows induction. The MPP1 promoter was more sensitive to glucose than the MOX and FMD promoters – it had less methanol induction in the presence of glucose. Most intriguingly, in the hxk1glk1 strain, both tested concentrations (2% and 0.1%) of glucose clearly enhanced expression from all three promoters – it was always higher than in respective methanol-grown cells. The highest expression level (almost 2900 Miller units) was recorded for the MOX promoter at growth of hxk1glk1 strain on 1% methanol + 0.1% glucose. This value is more than four times higher than the respective activity in methanol-grown wild type. So, glucose clearly activates methanol-specific promoters in hxk1glk1 H. polymorpha.


Glucose acts as inducer of methanol-specific promoters in double kinase-negative Hansenula polymorpha. Wild-type H. polymorpha (WT) and double kinase-negative mutant (hxk1glk1) harbouring Gus reporter plasmids for MOX, FMD and MPP1 promoters were grown on agar plates containing 1% methanol (met), and on the mixture of methanol and either 2% or 0.1% of glucose (glc) for 3 days and then assayed for β-glucuronidase activity. Average Gus activities in Miller units of at least three transformants and standard deviations are presented.

We had shown earlier that the glycerol-grown double kinase-negative mutant of H. polymorpha LR9 has a much higher alcohol oxidase activity than the respective wild type control (our unpublished data). Here, we inspected hxk1glk1 H. polymorpha 201 from the same aspect. The wild type and hxk1glk1 strains of H. polymorpha 201 carrying the MOX promoter reporter plasmid were grown in liquid medium with 2% glycerol and assayed for reporter activity in cell extracts. Gus activity of the wild type was 300 mU mg−1, whereas that of the hxk1glk1 strain was over twice as high, 793 mU mg−1. Given that glucose activated the MOX promoter in the hxk1glk1 strain (Fig. 7), we suspected that glycerol-grown hxk1glk1 H. polymorpha may contain glucose. Determination of intracellular glucose in wild-type, hxk1 and hxk1glk1 strains grown on agar plates with 2% glycerol revealed that hxk1glk1 cells contained 13 mM glucose in cell water, whereas no free glucose was detected in wild-type and hxk1 cells. In accordance with our data, Miseta et al. (2003) recorded a significant amount of intracellular glucose in triple kinase-negative mutant of S. cerevisiae grown on lactate.

At growth on non-sugar substrates such as glycerol, methanol, lactate and ethanol, gluconeogenesis reactions are obligatory to supply the cell with building blocks for cell wall, glycolipids, storage sugars (glycogen) and stress protectants (trehalose) (Argüelles, 2000; Türkel, 2006). If sugar-containing structures and molecules are recycled and storage sugars are hydrolyzed, unphosphorylated glucose should accumulate in the cell lacking glucose-phosphorylating enzymes.

Regulation of MAL1 promoter by carbon sources in H. polymorpha

First, we compared regulation of the MAL1 promoter between the wild type and hxk1 strains of H. polymorpha. Table 3 shows that high glycerol and glucose concentrations repressed the MAL1 promoter in both strains. In both strains, 2% glucose in the medium prevented induction from MAL1 promoter by disaccharides, whereas 2% fructose did so only in the wild type. Our results also indicate that hydrolysis products of maltose and sucrose repress the MAL1 promoter and that their phosphorylation is obligatory for the repression. This conclusion is based on following results: (1) lower disaccharide concentrations in the medium provide higher MAL1 promoter expression level, and (2) at growth on disaccharides, the MAL1 promoter was more highly activated in the hxk1 mutant.

View this table:

Regulation of MAL1 promoter by carbon sources in wild-type (WT) and hexokinase-negative (hxk1) strains of Hansenula polymorpha

Carbon sourcesβ-glucuronidase activity (Miller units)
2% glucose9 ± 213 ± 4
2% glycerol6 ± 25 ± 2
2% maltose87 ± 9249 ± 22
0.5% maltose186 ± 14266 ± 33
0.2% maltose216 ± 22397 ± 44
2% sucrose151 ± 18266 ± 30
0.5% sucrose167 ± 10402 ± 47
2% glucose + 2% maltose29 ± 743 ± 8
2% glucose + 2% sucrose40 ± 635 ± 7
2% fructose + 2% maltose18 ± 4370 ± 44
2% fructose + 2% sucrose17 ± 6497 ± 46
  • The strains harbouring Gus reporter plasmid with MAL1 promoter were grown on agar media for 3 days. Mean β-glucuronidase values ± standard deviation for at least three different transformants of each strain are shown.

As glucose activated methanol-specific promoters in hxk1glk1 H. polymorpha, the MAL1 promoter was addressed from the same aspect. Respective reporter strains were grown on glycerol-containing agar plates with addition of different sugars. Figure 8 clearly indicates that glucose, fructose, maltose and sucrose have an inducing effect on MAL1 promoter in hxk1glk1 strain. In the case of all studied carbon source combinations, the Gus activities in hxk1glk1 strain were higher than in maltose- or sucrose-growing wild type. Mesurement of glucose concentration in the medium indicated that although the hxk1glk1 strain cannot grow on sucrose and maltose, transport and hydrolysis of these sugars certainly took place. At least part of the glucose produced from disaccharide metabolism was excreted to the medium, most probably using some glucose transporter. After 2 days of growth of hxk1glk1 strain on 2% glycerol + 2% sucrose, the glucose concentration in the agar medium below the yeast growth area was 35 mM, whereas it was about twice as high (68 mM) in the case of growth on 2% glycerol + 2% maltose. Respective samples for wild-type cells contained no detectable glucose. Jansen et al. (2002) have reported for maltose-growing S. cerevisiae that if the cell cannot manage the metabolism of glucose resulting from maltose hydrolysis in the cell, intracellular glucose concentration rises and excess glucose will be exported by hexose transporters. Our data therefore suggest that accumulation of unphosphorylated hydrolysis products of sucrose and maltose in hxk1glk1 strain may cause the MAL promoter activation shown in Fig. 8.


Glucose acts as inducer of maltase gene promoter in double kinase-negative Hansenula polymorpha. Wild-type H. polymorpha (WT) and double kinase-negative mutant (hxk1 glk1) harbouring Gus reporter plasmid with MAL1 promoter were grown on agar plates containing 2% glycerol (gly) and 2% glycerol supplemented with 2% of glucose (glc), fructose (fru), maltose (mal), or sucrose (suc) for 3 days and then assayed for β-glucuronidase activity. Average Gus activities in Miller units of at least three transformants and standard deviations are presented.

We also carried out an experiment in which maltase induction by glucose was assayed in liquid culture. Wild type, hxk1 and hxk1glk1 strains were grown in liquid YNB medium with 2% glycerol and maltase activity was measured in cell extracts. Table 4 shows that glycerol-grown hxk1glk1 strain has over 20 times higher maltase activity than that of the wild-type and hxk1 strains. Already very high maltase activity was further increased if glucose was added to glycerol-growing hxk1glk1 cells.

View this table:

Maltase activity (mU mg−1) in cell extracts of Hansenula polymorpha strains grown in YNB liquid medium containing 2% glycerol

StrainCarbon sourceMaltase activity
Wild type (WT)2% glycerol24 ± 1
Hexokinase-negative (hxk1)2% glycerol21 ± 3
Double kinase-negative (hxk1glk1)2% glycerol560 ± 58
Double kinase-negative (hxk1glk1)2% glycerol + 0.1% glc*2829 ± 156
  • * Cells grown on 2% glycerol were washed, suspended in fresh medium containing 2% glycerol + 0.1% glucose and further grown for 9 h.

Growth of yeasts on disaccharides: unanswered questions

It is yet not clear how yeasts detect the presence of disaccharides in the medium to induce disaccharide-specific (MAL) genes. Wang et al. (2002) showed for S. cerevisiae that maltose permease is required for maltase induction by maltose. In that paper, the authors conclude that permease does not act as maltose sensor but rather provides the inducer for MAL genes inside the cell. Quite recently, it was proposed that the transcriptional activator of MAL genes (MAL activator) may function as intracellular sensor for maltose in S. cerevisiae. Ran et al. (2008) reported that the MAL63 activator forms a complex inside the cell with several chaperon proteins, including yeast homologues of Hsp70 and Hsp90. When maltose was added to the growth medium, MAL63 was released from the complex and maltase was induced. The authors suggest that binding of maltose to either MAL63 or its chaperon was responsible for the release. Our data on H. polymorpha also show that α-glucoside permease is required for the induction of maltase by sucrose and maltose (Viigand & Alamäe, 2007). A putative MAL activator gene is present in genomic MAL locus of H. polymorpha (Viigand & Alamäe,2007), but its function has yet to be proved. Results of current study show that hydrolysis products of maltose and sucrose activate maltase expression if they are not phosphorylated in the cell, i.e. in hxk1glk1 H. polymorpha (Table 4, Fig. 8). We consider that this is how initial derepression of MAL genes may take place, and it will probably be followed by up-regulation of these genes by yet unknown specific inducers. We have shown that H. polymorpha grown on non-sugar substrates is prepared to transport and split maltose and sucrose (Viigand et al., 2005; Viigand & Alamäe, 2007). In good accordance with our data, van Zutphen et al. (2010) showed that the maltase and the permease genes were respectively 89- and and 181-fold derepressed after the shift of H. polymorpha from glucose to methanol medium. Maltase has very low affinity for disaccharides (Liiv et al., 2001). Therefore, maltose and sucrose should be concentrated first into the cell by energy-dependent transport (Viigand & Alamäe, 2007), and only then will their efficient hydrolysis proceed. The hydrolysis reaction should produce a significant amount of glucose (and fructose) inside the cell, and because glucose- and fructose- phosphorylating activity of H. polymorpha growing on gluconeogenic carbon sources is low (Parpinello et al., 1998; Kramarenko et al., 2000), at least part of glucose and fructose should stay unphosphorylated and act as inducer for MAL genes.

Induction of glucose-repressed genes by low glucose has been reported for Aspergillus fungi. So, acetate-growing Aspergillus oryzae did not produce α-amylase, whereas addition of small amount of glucose induced its production (Carlsen & Nielsen, 2001). Murakoshi et al. (2012) showed for Aspergillus nidulans that aside of transglucosylation products of maltose (e.g. isomaltose and kojibiose) also glucose acted as physiological inducer of α-amylase. All these sugars promoted nuclear entry of respective activator protein AmyR, whereas for maltose as inducer, maltase activity was required. Although glucose induces α-amylase production in A. nidulans, this activity is masked to a certain extent by CreA-dependent catabolite repression. Accordingly, glucose acted as a much more potent inducer of α-amylase in a creA-defective strain (Murakoshi et al., 2012). Here, a parallel can be drawn with our results showing that glucose reveals its inducing ability in glucose repression-deficient (hxk1glk1) mutants.

Summing up, we consider that growth of H. polymorpha (and probably other yeasts) on disaccharides is complicated, because hydrolysis products of disaccharides that promote initial derepression of MAL genes will later cause repression due to accumulation of their phosphorylated species. Therefore, disaccharide transport and maltase expression should be finely accommodated with glycolytic flux to provide an appropriate expression level of disaccharide-specific genes.

Coregulation of MOX, FMD, MPP1 and MAL1 genes

This study suggests that H. polymorpha senses by an as yet unknown mechanism the presence and amount of glucose-6P in the cell to trigger repression of methanol-specific and MAL genes. We have evidence that these two sets of genes can be coregulated. Namely, in a regulatory mutant L63 of H. polymorpha, a single recessive mutation released maltase, alcohol oxidase and catalase from glucose repression (Alamäe & Liiv, 1998) suggesting a defect or absence of a shared repressor protein. We speculate that the availability of glucose-6P in the cytosol may modulate the activity of this hypothetical repressor, for example, to promote its nuclear entry. So, if glucose-6P is present and sensed in the cytosol, the glucose-repressed promoters will remain down-regulated. In the absence of free glucose-6P, the repressor will stay in the cytosol and allow derepression of the promoters. According to our data, unphosphorylated glucose and fructose also stimulate derepression of methanol-specific and MAL genes. We speculate that they may participate in maintaining cytosolic location of the hypothetical common repressor protein.

In addition to derepression, specific induction is also contributing to gene expression. It would make sense if specific activator genes were induced first in response to a new carbon source. After 2 h of transfer of glucose-grown H. polymorpha to methanol, the MPP1 gene is extremely highly (394 times) induced (van Zutphen et al., 2010). So, the MPP1 protein may indeed act as the main transcriptional activator of methanol-specific genes. The mechanism of induction is yet not known, but certain inducer metabolites may also have a role in it. To draw a parallel, acetaldehyde was shown to be a physiological inducer of ethanol-specific genes in A. nidulans (Flipphi et al., 1994). Regarding maltase expression, we have shown here that unphosphorylated glucose and fructose activate MAL1 promoter expression in H. polymorpha.

Concluding remarks

Analysis of reporter gene expression from MOX, FMD, MPP1 and MAL1 promoters in wild-type, hexokinase-negative (hxk1) and double kinase-negative (hxk1glk1) strains of H. polymorpha suggested that most probably the rate of sugar flux in the cell is sensed to initiate repression of these promoters. We propose glucose-6P as candidate for repression signalling metabolite. Intriguingly, our data on hxk1glk1 H. polymorpha show that unphosphorylated sugars in the cells promote derepression of MOX, FMD, MPP1 and MAL1 promoters. We consider that under certain conditions, unphosphorylated sugars may accumulate also in wild-type cells, causing initial derepression of methanol- and disaccharide-specific promoters that will be followed by growth substrate-specific induction. New data on regulation of putative peroxisomal activator protein gene MPP1 were presented. We showed that with regard to sugar repression, the MPP1 promoter was regulated similarly to MOX and FMD promoters. Yet, its expression was slightly more sensitive to glucose and its methanol-induced strength was lower compared with MOX promoter. We also show that cultivation of yeasts on agar plates created nutrient-limitation allowing assessment of promoter derepression. According to our results, hexokinase-negative and double kinase-negative strains of H. polymorpha are considered perspective hosts for foreign protein production from sugar-repressed promoters.

Supporting Information

Table S1. Primers used in this work.


This work was supported by grants 7528 and 9072 from the Estonian Science Foundation and by targeted financing grant SF0180088s08.


  • Editor: Monique Bolotin-Fukuhara


View Abstract