Introduction

Pseudozyma is a relatively new genus comprised of basidiomycetous yeast-like fungi related to the Ustilaginales. It contains mainly species that were originally misclassified on the basis of morphological characters until Begerow et al. (2000) recently associated them according to the homology of their nuclear LSU rDNA sequences. Because of their limited number, their recent discovery and their inconspicuous nature, Pseudozyma spp. have received limited attention in scientific literature. However, it is becoming increasingly apparent that some species do possess unusual features that warrant further exploration. For instance, Pseudozyma flocculosa was described as a biocontrol agent of fungal plant pathogens, such as powdery mildews (Bélanger and Labbé 2002). This fungus draws its antagonistic activity from the production of an unusual antifungal glycolipid, flocculosin, which was discovered after insertional mutagenesis (Cheng et al. 2003). It is interesting to note that it now appears that other Pseudozyma spp. and anamorphs of Ustilago maydis also produce similar glycolipids (Kulakovskaya et al. 2005; Hewald et al. 2005). These glycolipids are being investigated as promising new therapeutic agents against human pathogens (Mimee et al. 2005; Rodrigues et al. 2006). One of the constraints in their study however is that very little is known about the genetic basis of their production and regulation except for two putative genes isolated from U. maydis by Hewald et al. (2005).

In another line of investigation, some Pseudozyma spp. were shown to have great potential for recombinant gene expression (Avis et al. 2005). Their evolved status would confer them some specific advantages for complex protein production.

In all previous studies with Pseudozyma spp., genetic manipulations have relied on exogenous material because no genetic tools (e.g., promoter or terminator sequences) have been developed for the specific purpose to study this fungal genus. As a matter of fact, different heterologous promoters from yeasts and other ascomycetes tested in P. flocculosa, failed to yield protein production. So far, only the HSP70 promoter from the basidiomycete U. maydis was successfully used in Pseudozyma spp. (Cheng et al. 2001). These results corroborated other studies showing that promoters from ascomycetes often fail to drive gene expression in basidiomycetes (Schillberg et al. 2000; Casselton and Fuente Herce 1989; Mooibroek et al. 1990; Gold and Alic 1993; Godio et al. 2004).

Considering the opportunity of (1) understanding the behaviour of the biocontrol agent P. flocculosa in its natural environment; (2) deciphering the key enzymes implicated in the flocculosin and other glycolipids synthesis and (3) developing new expression vectors for the production of recombinant proteins in Pseudozyma, it becomes evident that the use of promoters that are intrinsic to Pseudozyma would further genetic studies of the genus.

For this purpose, our objective was to find and clone the promoter of the actin gene of P. flocculosa. This promoter was selected because of its strength and constitutive capacity demonstrated in other organisms (Matheucci et al. 1995; Cox et al. 1996; McDade and Cox 2001). The actin protein is one of the most conserved within eukaryotic cells and it plays a major role in various cellular processes like motility of components in cells and regulation of cell growth.

In this work, we report the first cloning and analysis of a promoter intrinsic to the Pseudozyma genus, that of the actin gene in P. flocculosa. Efficacy and reliability of the actin promoter were compared to that of the well-known HSP70 promoter from the closely related species U. maydis. Efficacy within species of Pseudozyma was also tested and described.

Materials and methods

Fungal material

P. flocculosa (Traquair, Shaw, and Jarvis) Boekhout and Traquair (CBS 167.88) and Pseudozyma antarctica (Goto, Sugiyama, and Lizuka) Boekhout (CBS 516.83) were maintained on potato dextrose agar (PDA) at 4°C. All culture media ingredients were supplied by Difco (BD Biosciences, Mississauga, Ontario, Canada).

DNA and RNA extraction

Fungal genomic DNA was isolated from 3-day-old cultures of P. flocculosa in potato dextrose broth using the Genomic-tip 20/G kit (Qiagen, Mississauga, Ontario, Canada). Total RNA was isolated from P. flocculosa using a RNeasy® mini kit (Qiagen) following the manufacturer’s instructions. Bacteriophage DNA was isolated as described by Santos (1991) based on phage precipitation with zinc chloride.

Cloning of a partial actin gene

A fragment of the actin gene was amplified from genomic DNA by using degenerate primers described in Tarkka et al. (2000) and corresponding to conserved amino acids 80–85 and 209–214 (Act-dege-F, 5′-GAYATGGARAAGATYTGG-3′ and Act-dege-R, 5′-TTYTCCTTGATGTCRCG-3′). Amplifications were performed under the following conditions: initial denaturation step at 95°C for 2 min followed by 30 cycles of 95°C for 45 s; 50°C for 1 min and 72°C for 45 s. The amplified fragment was cloned into the pGEM®-T Easy Vector (Promega, Madison, WI) before sequencing.

Five μg of total RNA was used to amplify the 5′-end of the actin cDNA using a commercial 5′-RACE System (Invitrogen, San Diego, CA). For this purpose, a gene-specific primer (Act1 GW N1, 5′-CTCGAAGCAGATCTGCGTCATCTTCTC-3′) was designed from the partial P. flocculosa sequence.

Cloning of the actin promoter

Fifty μg of P. flocculosa genomic DNA was used to construct a genomic library using the Lambda FIX® II/Xho I Partial Fill-in Vector Kit (Stratagene, La Jolla, CA) following the manufacturer’s protocol. The 5′-RACE fragment was radioactively labelled with 32P-dCTP by random priming using the Rediprime II kit (Amersham BioSciences, Arlington Heights, IL) following the manufacturer’s instructions and used as a probe. The genomic library was screened by plaque lift using Hybond N+ membrane and hybridization carried out as recommended by the manufacturer (Amersham Pharmacia Biotech, Arlington Heights, IL, USA). A positive clone was selected and its DNA was extracted and digested with BglII and sub-cloned into Psp72 (Promega, Madison, WI) linearized with BglII. The sub-cloned fragments were then sequenced.

Primer extension

To map the transcription start site, non-radioactive automated primer extension (NAPE) analysis was used as described by Altermann et al. (1999). Extension reactions were carried out with 9 pmol primer annealed to 10 μg total RNA isolated from P. flocculosa. The primer used in the NAPE experiment (ACT-primer-ext, 5′-CCTCCATCTTGGCTG-AGTGCTTCTG-3′) carried a fluorescent label (IRD700) at the 5′-end. Reference sequencing reactions were initiated using the same primer on a BglII genomic sub-clone using a primer-labelled cycle sequencing kit (LI-COR, Lincoln, NE). Products of the primer extension and sequencing reactions were separated on a DNA sequencer (model 4200L IR2 system, LI-COR).

Southern hybridization

Total DNA (5 μg) was digested separately with BamHI and XhoI and separated on 0.8% (w/v) agarose and transferred to Hybond N+ membrane (GE Healthcare, Piscataway, NJ). The 5′-RACE fragment was radioactively labelled with 32P-dCTP and used as a probe. Membranes were hybridized overnight at 68°C in 2X SSC, 0.25% dry milk, 0.5% SDS and washed at 68°C in 2X SSC, 0.1% SDS.

Plasmid construction

Flanking restriction sites were added to the sGFP(S65T) gene (Avis et al. 2005) using extended PCR primers (RP.gGFP.for 5′-CCCATCGATTGCCAGATCTCGACG-GCGCCGATGGTGAG-3′ and RP.gGFP.rev 5′-CGAGAGGCCTGCTCGAGTACTAG-TGTTTGACAGCCTAGGATCGG-3′). In this fashion, ClaI, BglII and NarI sites were added upstream and AvrII, SpeI and StuI sites downstream of the GFP coding region. The resulting GFP amplicon was digested with ClaI and StuI and inserted into pCPF (Avis et al. 2005) linearized with ClaI and EcoRV to produce pC2PF.GFP. This plasmid contains a GFP expression cassette under the control of the U. maydis HSP70 promoter and terminator. The actin promoter was amplified with KPNI-ACT (5′-GGGGTACCGAATCACTTTGCCCGCCTTG-3′) and CLAI-ACT (5′-CCATCGATG-TTGGCTGAGTGCTTCTGAG-3′) (Fig. 1a) resulting in a 1,149-bp amplicon with ClaI and KpnI restriction sites (underlined) and cloned into pGEM-T Easy (Promega, Madison, WI) to produce pGEM-PAct. This plasmid was then linearized with NcoI (made blunt via T4 DNA polymerase treatment) and ClaI, both located immediately downstream of the actin promoter. A fragment containing the GFP coding region and the HSP70 terminator (excised from pC2PF.GFP by NarI and SmaI digestion) was then cloned downstream of the actin promoter to produce plasmid pCAct.GFP. The latter contains the gfp gene under the control of the actin promoter and HSP70 terminator. Because plasmid pCAct.GFP harbors no selectable marker for fungal transformation, it was used in co-transformation experiments with plasmid pSceI-Hyg (provided by Dr. J. Kronstad, University of British Columbia, Canada). pSceI-Hyg contains a hygromycin B selection cassette comprised of the U. maydis HSP70 promoter, the Escherichia coli hygromycin phosphotransferase gene, and the U. maydis HSP70 terminator.

Fig. 1
figure 1

Structure of partial actin gene and nucleotide sequence of the actin promoter from P. flocculosa. aSolid and open boxes indicate the non-coding region. The region amplified by degenerated primers is delimited by the double arrow. b The transcription initiation sites are indicated by asterisks. The open box and underlined sequences correspond to the putative GC box and CT-rich region. The numbers indicate the positions relative to the site of initiation of translation of the actin gene. The arrows indicate the position of the primers (KPNI-ACT and CLAI-ACT) used to clone the promoter

Full size image

pSPF.GFP (Avis et al. 2005) contains two cassettes in the same orientation, the hygromycin selection cassette from pSceI-Hyg and a green fluorescent protein (GFP) expression cassette consisting of the gfp gene under the control of the U. maydis HSP70 promoter and terminator sequences.

Fungal transformation

Protoplasts of P. flocculosa were prepared according to Cheng and Bélanger (2000) by using 10% glucanex (Sigma, Mississauga, Ontario, Canada) as the lysing enzyme. For each co-transformation, 10 μg of pSceI-Hyg DNA were used with 10 μg of pCAct.GFP or pC2PF.GFP. pSceI-Hyg and pC2PF.GFP were linearized with XhoI and pCAct.GFP with XmnI. After digestion, the plasmids were purified by ethanol precipitation before being used to transform protoplasts. Transformants were selected on YMPDA medium amended with 0.8 M sucrose as osmotic stabilizer and 100 μg ml−1 of hygromycin B as the selective agent (Cheng et al. 2003).

P. antartica was transformed by electroporation as described by Fortier (2006). Plasmids used for this transformation were prepared as described previously for the transformation of P. flocculosa. Transformants were selected on YMPDA medium amended with 300 μg ml−1 of hygromycin B.

Screening of transformants

Transformants were tested by PCR to confirm the integration of the gfp gene. Gene-specific primers (GFPfwd 5′-GGGTGGTGCCCATCCTGGTCG-3′ and GFPrev 5′-TGAGTGATCCCGGCGGCGGTC-3′) were used to specifically amplify the gfp gene. PCR was conducted under the following conditions: 94°C for 5 min, 30 cycles of 94°C, 20 s; 65°C, 20 s; 68°C, 40 s and a final extension (68°C, 5 min). Ten positive transformants were randomly selected and maintained on PDA and 100 μg ml−1 hygromycin for each co-transformation.

Liquid culture of transformants

For GFP expression analysis, pre-cultures were prepared by growing the strains for 3 days at 25°C in enriched YMPD (containing 3 g l−1 yeast extract, 3 g l−1 malt extract, 7.5 g l−1 peptone water and 15 g l−1 dextrose) in 50-ml conical plastic tubes (Sarstedt, Montréal, QC, Canada) with agitation (250 rpm). Then, 0.5 ml of the pre-culture was used to inoculate 10 ml of enriched YMPD medium in 50-ml tubes at 25°C with agitation (250 rpm). Samples were taken at 31 h (mid-exponential phase): 1 ml for Northern blot analysis and 0.75 ml for Western blot analysis.

Western blot analysis of transformants

For Western blot analysis, cells were harvested by centrifugation for 5 min at 5,000×g and washed with 1 volume of cold 50 mM sodium phosphate. Protein extracts were prepared by adding 1 volume of loading buffer (50 mM Tris pH 6.8, 2.3% (w/v) SDS, 10% (v/v) glycerol, 5% (v/v) β-mercaptoethanol) to cell pellets and by heating at 68°C for 10 min. Two volumes of 2 mm acid-washed glass beads were added to the sample and the mixture was submitted to 6 cycles of 1 min boiling and 1 min vortexing. The crude extract was centrifuged and proteins in the supernatant were quantified by silver dot staining (Draber 1991). Protein extracts were separated on a 12% polyacrylamide gel by SDS-PAGE and transferred to a nitrocellulose membrane (Hybond, Amersham BioSciences, Piscataway, NJ). GFP was detected with an anti-GFP antibody (Roche, Molecular Biochemicals, Mannheim, Germany) at a dilution of 1:2,000. After incubation with the chemiluminescent HRP-substrate, the membrane was exposed to a BioMax MS film (Eastman-Kodak, Rochester, NY) and analyzed by densitometry measurements (Scion image 4.0.3) and standardized with a commercial GFP control.

Microscope observations

Two transformants expressing high fluorescence levels for each construction and each Pseudozyma spp. were selected. These transformants were grown on PDA for 3 days. Samples of cells were mounted in water on glass slides and observed under a microscope (Olympus model BH2-RFCA; Olympus, Tokyo, Japan). For fluorescence microscopy, observations were carried out under blue light (488 nm) using the same microscope equipped with a BH2-RFL-T3 mercury lamp. Images were recorded using a Coolsnap-Pro (MediaCybernetics, Silver Spring, MD) color camera and analyzed with Image Pro Plus software (MediaCybernetics). Observations were made at ×100.

Results

Cloning and sequence analysis of a partial actin gene and promoter

Using degenerate primers based on conserved amino acid motifs, a 404-bp fragment was amplified by PCR from the P. flocculosa genomic DNA (Fig. 1a). The 5′-most region of the P. flocculosa actin coding region was obtained by performing 5′-RACE on first strand cDNA. Subsequently, this 5′-RACE fragment was used to screen a genomic library. After restriction mapping, a ∼7,800 bp BglII sub-clone was sequenced and found to contain the promoter region. After the alignment of the 5′-RACE and genomic DNA sequences, a partial actin gene sequence of 2,612 bp comprising 1,133 bp of the promoter and the genomic sequence coding for the first 214 amino acids (including 4 introns) was obtained (GenBank accession number DQ913895) (Fig. 1a).

The transcription start site was defined by primer extension analysis (Fig. 2). A single initiation site located 109 bp upstream of the start codon was found and it lies within one of two CT-rich regions that precede the ATG. The actin promoter region has a relatively high G+C content (67.4%) and was found to lack a typical TATA or CAAT box but contained three putative GC boxes, respectively 752, 384 and 182 bp upstream of the start codon ATG (Fig. 1b). Two CT-rich regions were observed 100 and 28 bp upstream of the start codon (Figs. 1b and 2).

Fig. 2
figure 2

Mapping of actin transcript from P. flocculosa by NAPE. Two μl of primer extension reaction (1:1, lane 1; 1:4, lane 2; 1:10, lane 3). Lanes A, C, G and T correspond to the reference sequencing reactions. The sequence of the coding strand encompassing the 3′-end of the extension product (asterisk) is shown to the right

Full size image

Copy number of the actin gene

To determine the copy number of the actin gene in P. flocculosa, a Southern hybridization was conducted. Genomic DNA was digested with either BamHI or XhoI and probed using the 5′-RACE fragment. A single band was visible in each digestion (Fig. 3) indicating that the genome of P. flocculosa has only one copy of the actin gene.

Fig. 3
figure 3

Southern blot analysis of the genomic DNA of P. flocculosa. Lane 1: molecular marker, lane 2: BamHI, lane 3: XhoI

Full size image

GFP expression analysis

To estimate the activity of the actin promoter in Pseudozyma, we compared the levels of protein expression of the gfp gene under the control of either the actin or HSP70 promoter in the exponential phase of development (31 h of culture) (Fig. 4). In P. flocculosa, GFP protein expression levels were very weak under the control of the HSP70 promoter compared to those with the actin promoter (Fig. 4a). Indeed, the greatest amount of GFP protein obtained was only 2.1 mg g−1 total protein with the HSP70 promoter as opposed to 27.8 mg g−1 total protein with the actin promoter. In P. antarctica (Fig. 4b), the difference between the actin and HSP70 promoters was less pronounced. GFP protein levels were quite similar regardless of the promoter used. In P. antarctica, we obtained up to 21.9 mg g−1 total protein with the HSP70 promoter and as much as 17.7 mg g−1 total protein with the actin promoter.

Fig. 4
figure 4

GFP expression in ten P. flocculosa (a) or P. antarctica (b) transformants expressing the gfp gene under the control of the P. flocculosa actin promoter or the U. maydis HSP70 promoter. The transformants are listed in decreasing order of efficiency. Expression in a wild-type (WT) culture of each fungus was used as control. Culture samples were taken 31 h after inoculation of 10 ml YMPD liquid cultures. GFP expression was quantified at the protein levels as described in the “Materials and methods” section

Full size image

To assay the strength of the actin promoter in cells grown on solid media, two transformants for each promoter and for each Pseudozyma species were observed under blue light (Fig. 5). A similar pattern of GFP expression was observed on solid culture as was seen previously on liquid culture. GFP fluorescence was visible in P. flocculosa and P. antarctica when it was under the control of the actin promoter. In contrast, transformants with the HSP70 promoter showed fluorescent activity only in P. antarctica.

Fig. 5
figure 5

Microscope observations of two P. flocculosa and two P. antarctica transformants expressing GFP under the control of the actin promoter (rows 1 and 2) or the HSP70 promoter (rows 3 and 4). All observations were made under blue light (488 nm) (columns 1 and 3) or visible light (columns 2 and 4). Bar = 100 μm

Full size image

Discussion

In this study, we report the first cloning of a promoter endogenous to the multi-faceted yeast-like fungus P. flocculosa, that of the actin gene. This promoter presents many characteristics typical of housekeeping genes, which are considered to be constitutively expressed in the cell (Blake et al. 1990). The absence of TATA or CAAT sequences and the presence of multiple GC boxes is a feature shared by many promoters of such genes. While relatively little information is available on GC boxes, of which there are three putative examples in the P. flocculosa promoter, they are believed to play a major role in the initiation of transcription (Blake et al. 1990; Hapgood et al. 2001). A final feature of the actin promoter is the location of the site of transcription initiation within one of two long CT-rich repeats upstream of the ATG. Other studies showed that when the transcription initiation site lies in a CT-rich sequence, it often determines the accuracy and frequency of transcription initiation (Smale and Baltimore 1989; Hariharan and Perry 1990; Kolluri et al. 1992; Miyashita et al. 1995; Punt et al. 1990; Chen and Roxby 1997; Machida et al. 1996; Yamazaki et al. 2002; McNeil 1988; Dobson et al. 1982).

Our results confirmed that only one copy of the actin gene was present in the P. flocculosa genome. In lower eukaryotic organisms, the actin gene was generally found in one copy. This was found to be the case in yeasts (Deshler et al. 1989; Losberger and Ernst 1989; Wery et al. 1996) and several other fungi (Dudler 1990; Matheucci et al. 1995; Fidel et al. 1988; Mertins and Gallwitz 1987). Among the basidiomycetes, U. maydis, a closely related member of the Ustilaginales, also contains only one copy of the gene. On the other hand, a small actin gene family was found in two other basidiomycetous fungi, Schizophyllum commune and S. bovine (Tarkka et al. 2000), a feature the author attributed to the more complex nature of these fungi.

In a previous study, among five heterologous promoters, only the HSP70 promoter from U. maydis was found to be functional in P. flocculosa. Promoters from ascomycetes and other lower fungi failed to drive the expression of a selectable marker gene (Cheng et al. 2001). In this work, it is surprising to note that the HSP70 promoter was rather weak in P. flocculosa. In investigating the differences between this work and that of Cheng et al. (2001), we found that the HSP70 promoter was stronger in P. flocculosa when cells were grown in the presence of hygromycin (R. Anguenot, personal communication). The absence of hygromycin during the culture of the transformants would explain the very low level of GFP expression obtained with the HSP70 promoter in P. flocculosa in the present work. No such requirement for hygromycin was observed in P. antarctica, which resulted in high levels of GFP expression with the HSP70 promoter. This difference between P. flocculosa and P. antarctica can possibly be explained by their relatively distant position within the genus Pseudozyma (Begerow et al. 2000). Indeed, based on highly conserved rDNA sequences, P. flocculosa and P. antarctica were classified in two different sub-groups of the Pseudozyma genus. These results highlight the specificity of promoters within closely related genetic groups as reported in many other studies (Godio et al. 2004; Casselton and Fuente Herce 1989; Mooibroek et al. 1990; Schillberg et al. 2000).

On a comparative basis, the use of the actin promoter resulted in a higher level of expression of the GFP protein in P. flocculosa than the HSP70 promoter in the absence of hygromycin. In P. antarctica, the level of GFP expression was similar with both promoters. When the amount of GFP protein expressed intracellularly under the control of the actin promoter was measured, the specific yield was equivalent to that reported in Pichia pastoris (28 mg g−1), a well-known and often-used platform for recombinant protein expression (Pérard 2000). This attests to the strength and utility of this new promoter.

By isolating a promoter endogenous to the genus Pseudozyma, we have expanded the spectrum of applications to further our understanding or exploitation of these poorly known fungi. For instance, fluorescence microscopy results presented here showed that GFP expression under the actin promoter allows easy observations of individual fungal cells. In turn, this opens up an opportunity for analyses of the ecological behavior of P. flocculosa in its natural environment as demonstrated by Neveu et al. (2006).

In conclusion, we have isolated and cloned a new actin promoter from P. flocculosa and compared its efficacy at driving the production of heterologous proteins against the well-known HSP70 promoter from U. maydis. Our results showed that this actin promoter was as strong as HSP70, not dependent on hygromycin for upregulation and transferable to other Pseudozyma spp. Considering the mounting interest in the unusual properties of these yeast-like fungi, this work provides a new versatile genetic tool for the study of the Pseudozyma genus.