Genetic basis of flocculation phenotype conversion in Saccharomyces cerevisiae

Abstract

Two NewFlo-type flocculent transformants Saccharomyces cerevisiae YTS-S and YTS-L were obtained from a partial yeast genomic library. Even though both of the transformants displayed the same flocculation phenotype, they represented different physiological characteristics during detailed investigation. Analysis of the two transformants YTS-L and YTS-S confirmed the presence of FLONL and FLONS genes, respectively. The 3396-bp ORF of FLONS encoded a protein of 1132 amino acids. Meanwhile, the presence of a 1686-bp ORF encoding a 562-amino acid protein was revealed in FLONL. Both FLONL and FLONS showed high identity to FLO1 gene. Aligned with the intact FLO1 gene, FLONS lost two internal repeated regions, whereas one repeated sequence was inserted into the middle of the FLONL gene. All of the altered regions could be found in the middle repetitive sequence of the FLO1 gene. The results indicate that FLONL and FLONS are both derived forms of the FLO1 gene. Genetic variability triggered by tandem repeats in FLO1 gene is believed to be responsible for the differential phenotypic properties of the yeast strains YTS-S and YTS-L.

Introduction

Yeast flocculation can be defined as an asexual aggregative process of yeast cells into clumps with subsequent fast sedimentation in the medium (Johnson et al., 1988; Stratford 1993). Two flocculation phenotypes can be distinguished by their sugar inhibition pattern: the Flo1 type, which is inhibited only by mannose and not by glucose, maltose, sucrose, or galactose, and the NewFlo type, which is inhibited by mannose, glucose, maltose, and sucrose but not by galactose (Stratford & Assinder, 1991).

Flocculation is an important factor in industrial applications of Saccharomyces cerevisiae (Verstrepen et al., 2003). An understanding of its genetics and physiology is therefore essential. One of the genes associated with the Flo1 flocculation phenotype is FLO1, best-known for its expression and regulation (Teunissen et al., 1993a, b; Bidard et al., 1995; Bester et al., 2006). The intact FLO1 gene has been cloned and its nucleotide sequence has been determined (Watari et al., 1989; He et al., 2002). The sequence reveals that the ORF of the intact gene is composed of 4614 bp, which codes for a protein of 1537 amino acids. A remarkable feature of the putative Flo1 protein is that it contains four families of repeated sequences composed of 18, 2, 3 and 3 repeats and that it has a large number of serines and threonines. Genetic methods such as Southern and Northern blot analysis revealed that genetic alteration of FLO1 gene occurred not only in Escherichia coli but also in brewing yeast. Two truncated FLO1 genes, FLO1S and FLO1M, were isolated from yeast. FLO1S contains a large ORF of 2586 bp and codes for a protein of 862 amino acids. This shorter deleted form of the FLO1 gene from S. cerevisiae ABXL-1D can induce weak flocculation in a nonflocculent yeast strain (Watari et al., 1991). Sieiro (1998) isolated a 3.1-kb fragment termed FLO1M and demonstrated that it was able to confer a flocculation phenotype in S. cerevisiae IM18-b. Sequence analysis indicates that the deletion occurred in the middle repeat region in both of the short forms of the FLO1 gene. The difference in flocculation efficiency between FLO1S, FLOM, and FLO1 suggests that the number of repeats influences the degree of flocculation (Sato et al., 2001): the longer the repeats, the stronger the flocculation ability. Recently, the fact that the size variation of FLO1 gene creates quantitative alterations in phenotypes (e.g. adhesion, flocculation, biofilm formation) was confirmed by Verstrepen (2005). However, no report about flocculation type conversion generated by the repeat alteration in FLO1 gene has been released until now.

In this study, two yeast transformants exhibiting NewFlo flocculation phenotype were obtained by constructing a partial yeast DNA library. Sequence analysis of FLONL and FLONS, genes carried by the corresponding transformants YTS-L and YTS-S, showed distinct differences. During detailed investigation of physiological characteristics of flocculation, the two transformants displayed variant patterns. Therefore, an explanation for the variation in physiological characteristics was advanced based on the genetic diversity.

Materials and methods

Strains, plasmids and media

FLONL and FLONS genes conferring NewFlo-type flocculation ability were obtained by constructing a partial yeast genomic library. Saccharomyces cerevisiae YN79, a commercial brewing yeast strain with a strong NewFlo phenotype flocculation strength, was employed as a DNA donor. The nonflocculent strain of S. cerevisiae YS58 (MATαflo1 ura3-52 leu2-3, 112 his4-519 trp1-789) was used as a receptor to express FLON genes (Teunissen et al., 1993a, b). The typical Flo1-type flocculent strain of S. cerevisiae YS59 (MATαFLO1 ura3-52 leu2-3, his4-519) was used as a control strain during physiological tests (Bayly et al., 2005). Escherichia coli DH5α (SupE44, hsdR17, recA1, ndA1, gyrA96, thi-1 relA1, ΔLacU169 (ϕ80LacZΔM15)) and the centromeric shuttle plasmid YCp50 (URA3) was used for recombinant plasmid construction. YFp-L and YFp-S were the recombinant plasmids carrying the intact FLONL and FLONS genes, respectively. Plasmid pEA-1 was constructed by insertion of the ADH1 terminator into the shuttle vector YEp352 (Jiang et al., 2004). Yeast strains were grown in YPD medium (2% glucose, 2% bactopeptone, 1% yeast extract) and E. coli in Luria–Bertani (LB) medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl), except for the special description. Maltose medium was used to investigate the onset of yeast flocculation during the fermentation procedure.

Measurement of flocculation ability

Cells were grown in 5 mL of liquid YPD medium for 48 h at 20°C with shaking. They were then harvested and washed twice with 0.1 M EDTA and sterile water. The flocculation assay was performed according to the method described by Smit (1992). Cells were resuspended in 3 mL of flocculation buffer (50 mM sodium acetate, 0.1% CaCl₂, pH 4.5) in a 1.0 cm cuvette, with or without 0.1% CaCl₂, to a final concentration equivalent to an A_{600 nm} of 2.0. After 5 min of vigorous agitation, the OD of the yeast suspensions in the presence or absence of Ca²⁺ was measured. Flocculation ability was determined by the following equation: C=1–B/A, where A is the A_{600 nm} without Ca²⁺, B is the A_{600 nm} with Ca²⁺, and C is the flocculation ability. Negative values were regarded as being equal to zero.

Physiological characterization of flocculent transformants

To detect sugar inhibition on flocculation, yeast cells were resuspended in flocculation buffers containing different final sugar concentrations. Fermentable sugars (galactose, glucose, sucrose and maltose) and non-fermentable sugar (mannose) were used in this experiment. Yeast cells were resuspended in flocculation buffer with different pH values and Ca²⁺ concentrations to detect the influence of pH and Ca²⁺ on flocculation. Where necessary, succinate buffer was used in place of citrate to avoid precipitation. Control experiments showed no difference in flocculation in these two buffers.

Saccharomyces cerevisiae YTF-L and YTF-S were inoculated from liquid culture into 250 mL conical flasks containing 100 mL maltose medium. Growth, flocculation ability, and the reducing sugar concentration in the culture were measured at regular intervals. Growth was determined by measuring changes in OD at a wavelength of 600 nm using a spectrophotometer. After 20 h of growth under normal cultivation procedures (28°C, 200 r.p.m.), observable flocculation of yeast strains occurred and was measured. The reducing sugar concentration in the maltose medium was also determined at the same time by the dinitrosalicylic acid method (Bailey, 1988).

All the tests were repeated three times under the same conditions. The average value of three measurements was calculated and is reported in this article.

Construction of a yeast genomic library

A partial genomic library for S. cerevisiae YN79 was constructed based on the plasmid YCp50, containing a centromere sequence. Saccharomyces cerevisiae genomic DNA was prepared and partially digested with BamHI and EcoRI, and fractionated by agarose gel electrophoresis to obtain fragments of <10 kb. Ligated plasmids were then introduced into E. coli DH5α (Sambrook & Russell, 2001). Transformed E. coli cells were screened for clones capable of growing in LB medium containing ampicillin (50 μg mL⁻¹); the plasmids were isolated and transformed into nonflocculent strain S. cerevisiae YS58 by the lithium acetate procedure (Gietz & Woods, 2002). Yeast transformants grown on SC medium with leucine, histidine, and tryptophan were selected out to detect the flocculation ability.

Plasmid construction

The 2093-bp sequence of the FLONL gene was amplified from YFp-L by PCR, with the forward and reverse primers 5′-TACGAATTCGACTACTGCCTACATATTT-3′ and 5′-TACGGATCCTCAGTGGATAGTGAGGT-3′, which contained a BamHI and EcoRI site, respectively. The PCR product contained the intact ORF and the upstream sequence of the FLONL gene. The resulting plasmid pEAF-L was obtained by inserting the PCR product into pEA-1. Recombinant yeast strain YTSC-L was constructed by introducing pEAF-L into the nonflocculent strain S. cerevisiae YS58.

For construction of the expression plasmids, the ORFs of FLONL and FLONS were amplified by PCR with the primers FLONLS-F: 5′-TACGAGCTCATGACAATGCCTCATCGCTA-3′; FLONS-R: 5′-TACGGTACCTTAAATAATTGCCAGCCAA-3′; and FLONL-R: 5′-TACGGTACCTCAGTGGATAGTAGAGG-3′. The forward and reverse primers contained SacI and KpnI sites, respectively. The PCR product was inserted into pEA-1 containing the ADH1 terminator. The fragments containing the ORFs of the FLON genes and ADH1 terminator were cut out using SacI and BamHI from the correctly ligated plasmids, then inserted into the plasmid YEcup with the PGK1 promoter (Jiang et al., 2004), resulting in the recombinant expression plasmids pEA-fol and pEA-fos.

Sequence analysis

DNA sequence analysis was performed with blast software from the National Center for Biotechnology Information (NCBI). Multiple protein sequence alignments were done with clustal w from the EMBL.

Southern and Northern blot analyses

Southern blot hybridization was performed using the DIG Labelling Kit from Roche Biochemical Products (USA), using the method described in the DIG Application Manual. The 740-bp (1–739) NcoI–BamHI 5′ flanking sequences of the FLONL and FLONS genes were used as probes. Total yeast genomic DNA digested with restriction endonucleases and separated by electrophoresis (5 μg lane⁻¹) were denatured and then transferred to nitrocellulose membranes. Probes were labelled with digoxigenin. Hybridizations were carried out under standard high stringency conditions (Sambrook & Russell, 2001).

Total RNA was isolated from 5 mL yeast cultures grown to a stationary phase by means of the glass bead disruption method (Ausubel et al., 1994). RNA was separated on 1% agarose gels containing 0.7% formaldehyde and thereafter transferred and cross-linked to the nylon membranes. Probes to detect the mRNA of the FLON genes and the actin-encoding ACT1 gene were generated through PCR with the primer FLON-F probe: 5′-ATGACAATGCCTCATCGCTA-3′; FLON-R probe: 5′-GGTGACGGTGGTCATTT-3′; ACT-F probe: 5′-GACGCTCCTCGTGCTGTCTT-3′; and ACT1-R probe: 5′-GGAAGATGGAGCCAAAGCGG-3′. They were then labeled using the PCR DIG probe synthesis kit (Roche Diagnostics). The labelled PCR products correspond to the nucleotides +1 to +898 and +73 to +972 of FLON ORFs and ACT1 ORFs. After hybridization, the probe-target hybrids were visualized as described in the digoxigenin application manual (Roche Diagnostics).

Results

Screening for genes conferring NewFlo-type flocculation ability in yeast

A partial genomic library of the strain S. cerevisiae YN79 was prepared in the centromeric shuttle plasmid YCp50 to obtain the gene conferring NewFlo flocculation phenotype in yeast. The resulting recombinant plasmids were introduced into the nonflocculent strain S. cerevisiae YS58. Among 2000 colonies grown on SC medium with leucine, histidine and tryptophan, two flocculent transformants, YTS-S and YTS-L, exhibited decreased flocculation ability in the flocculation buffer containing mannose or glucose, suggesting the presence of FLO genes conferring NewFlo-type flocculation ability. The BamHI–EcoRI insert of 3.8 and 4.2 kb, termed FLONS and FLONL, respectively, were detected in the recovered recombinant plasmids from the transformants YTS-S and YTS-L.

Southern blot hybridization for detecting FLONL and FLONS genes in the genome of the donor strain S. cerevisiae YN79 was performed using the fragments corresponding to the 5′ flanking sequences in the FLON genes as probes. The presence of FLONS gene in the genome of S. cerevisiae YN79 was confirmed by detecting a single hybridization band of 3.8 kb (FLONS) with the FLONS probe. However, the 4.2-kb band (FLONL) was not observed in the Southern blot analysis of S. cerevisiae YN79 DNA, using the FLONL probe. A 5.1-kb fragment signal (FLO1 gene) appeared in the chromosome of the typical Flo1-type strain S. cerevisiae YS59 with both of the FLON probes, suggesting a high identity between FLON genes and FLO1 gene. Moreover, the signal appeared as a single band when FLONS hybridized with the probe of FLONL, revealing a significant similarity between FLONL and FLONS (Fig. 1).

To explain the confusing result of the Southern blot, PCR was performed using the 5′- and 3′- nucleotide sequence of FLO1 gene as primers: FloF: 5′-ATGCCTCATCGCTATATGTTTTTGG-3′; FloR: 5′- TTAAATAATTGCCAGCAATAAGGA-3′. The obtained PCR result showed multiple bands shorter than the intact FLO1 gene when a chromosome of the donor strain S. cerevisiae YN79 was used as a template. The allele specificity of the 3.8-kb reaction product confirmed the presence of FLONS in the genome of S. cerevisiae YN79. However, the absence of the PCR fragment of the exact same size as FLONL corroborated the result of the Southern blot analysis: the FLONL gene did not exist in the chromosome of the strain S. cerevisiae YN79 (Fig. 2). The multiple-banded PCR profiles revealed that the FLO1 gene existed as various deleted forms in the genome of S. cerevisiae YN79.

Physiological characteristics of yeast flocculation

An obvious difference in flocculation levels was observed in yeast strains YTS-L and YTS-S during screening of transformants. Hence, further research on their phenotypic characteristics (sugar inhibition, pH and Ca²⁺ effects, the onset of flocculation) was carried out in detail.

Inhibition of sugars on flocculation

Cells of YTS-S and YTS-L were exposed to different sugars. Flocculation of transformants YTS-S and YTS-L was inhibited by glucose, sucrose, maltose, and mannose, demonstrating the typical NewFlo phenotype. For each type of sugar, the concentration required for complete flocculation inhibition of YTS-L was higher than that required for YTS-S (Table 1), suggesting the affinity of FloNS protein for all of the sugars was stronger than that of FloNL protein. The sugar concentration had a dramatic effect on the flocculation level of the yeast cells. Furthermore, flocculation of the strains YTS-S and YTS-L was inhibited more easily by glucose than by mannose (Fig. 3). Interestingly, galactose, as a noninhibitor of both Flo1 and NewFlo-type flocculation, strongly inhibited the flocculation of YTS-S, while displaying a weaker inhibition on flocculent cells of YTS-L. The results suggest that the lectin-like proteins involved in sugar affinity were different in the yeast strains YTS-S and YTS-L.

Effects of Ca²⁺ and pH on flocculation

The transformants containing FLON genes revealed Ca²⁺-dependent flocculation ability. Two different behaviours were observed when Ca²⁺ concentration was more than 20 mM in the buffer (Fig. 4a): Ca²⁺ ion weakly inhibited flocculation of strain YTS-L, whereas there was almost no influence on YTS-S. Different flocculating responses among yeast strains to the Ca²⁺ concentration can be attributed to the difference in flocculation lectins.

Both YTS-S and YTS-L flocculated remarkably well in a wider pH range: 2.5–6 (Fig. 4b). YTS-S was more sensitive to low pH values (<2.0) in buffer than was YTS-L. Strangely enough, the flocculation ability of YTS-L again increased when pH values were above 5. It has been suggested that the conformational change of flocculins may occur when the electrostatic charge of surface proteins changes (Minoura et al., 2004). The results here suggest a pH-sensitive conformational change in FloN proteins.

Onset of flocculation

The onset of flocculation also differed between the yeast strains. Strains YTS-S and YTS-L became flocculent upon entering the stationary phase of growth (about 28 h) before the wort was completely attenuated, defined as NewFlo-type (Fig. 5). In addition, the flocculation ability of YTS-S increased during the entire test duration, whereas the flocculent strength of YTS-L experienced a stagnant phase between 56 and 72 h, and then increased abruptly.

Sequence analysis of FLONL and FLONS

DNA sequences of FLONL and FLONS were determined and submitted to GenBank (access numbers: EF182714 for FLONS and EF417874 for FLONL). The full length of FLONS and FLONL was 3843 and 4293 bp, respectively. The length of the ORF of FLONS was calculated to be 3396 bp, encoding a 1132-amino acid residue structural protein. However, FLONL gene presented an ORF of 1686 bp, encoding a 562-amino acid protein, significantly shorter than the FloNS and Flo1 proteins.

Amino acid residues 1–302, 303–615, and 616–1132 in the FloNS protein had complete identity with amino acid residues 1–302, 527–842, and 1023–1537 in the Flo1 protein. Two internal regions of 225 amino acids (303–527) and 180 amino acids (799–979) in the Flo1 protein were not presented in the FloNS protein (Fig. 6). In the missing regions, three threonine-rich repeat units were found: repeat unit A of 90 bp, repeat unit B of 45 bp, and repeat unit C of 45 bp (Fig. 7). Furthermore, an additional repeated insert from nucleic acid residues 2186 to 2834 (650 bp) was found in FLONL when compared with the FLO1 gene. The 1–2185- and 2835–4293-bp regions in FLONL had 100% identity to the FLONS gene. One repeat unit of 135 bp in the insert region was revealed by the tandem repeat finder software available online (http://tandem.bu.edu/). The nucleic acid sequence of this repeat unit was completely identical with that of repeat unit A in the missing region of FLONS.

FloNL protein possesses the intact N-terminal of the Flo1 protein, but has lost most of the C-terminal sequence. To confirm that the truncated protein FloNL confers flocculation ability in yeast, the recombinant strain YTSC-L was constructed by introducing the recombinant plasmid pEAF-L with the coding sequence of FLONL into the nonflocculent strain YS58. The high flocculation level of 78.4% in the yeast strain YTSC-L strongly confirmed that the truncated coding sequence of FLONL absolutely governed flocculation phenotype in yeast.

Expression of FLONL and FLONS

The upstream sequences of FLONL and FLONS were different: 157 bp for FLONS and 407 bp for FLONL. To detect the influence of upstream sequence size on the expression of FLON genes, the recombinant plasmids pEA-fol and pEA-fos, containing the ORFs of FLONL and FLONS under the regulation of the same set of promoter and terminator, were constructed. The expression pattern of each ORF, regulated by the PGK1 promoter in the nonflocculent strain S. cerevisiae YS58, was displayed by Northern blot analysis (Fig. 8). No obvious difference was found in the mRNA transcript level of the two genes. This result indicates that the longer upstream sequence of FLONL did not contribute to the higher flocculation level of YTS-L. In addition, the mRNA transcript of the FLONL ORF was much shorter than that of the FLONS ORF, reinforcing the previous notion that the FloNL protein was much shorter than the FloNS protein. The flocculation level of the transformants containing the expression plasmids pEA-fos and pEA-fol was 76.4% and 75.8%, respectively.

Discussion

Flocculation is one of the most important characteristics of industrial yeast strains. The FLO1 gene product is thought to play a major role in flocculation. However, the exact function of this protein in the cell–cell interaction is still unclear. In this paper, we showed for the first time with two derivates of the FLO1 gene that repeated deletions in the FLO1 gene caused the flocculation phenotype conversion from Flo1 to NewFlo. Moreover, genetic analyses were performed on these two FLO1 derivates conferring NewFlo-type flocculation ability. We further attempted to explain the mechanisms for flocculation existing in different yeast strains and environmental conditions based on the analysis.

The deleted forms of FLO1 reported by Teunissien et al. (1993a) and Watari (1989), Watari & Keranen (1994) revealed that the deletion occurred at slightly different positions. Teunissien et al. speculated that recombination between repeats caused the deletion in the FLO1 gene during propagation in E. coli. It also has been reported that the Lg-FLO1 gene originated from a recombination between YHR211 and YAL065 generated by chromosomal duplication and translocation in brewing yeast (Kobayashi et al., 1999). In this study, sequence alignment showed that either the deletion or the insertion was found in the repeat section of the FLO1 gene. From the Southern blot analysis results, it was believed that FLONS was located in the S. cerevisiae YN79 genome, whereas FLONL seemed to be present during manipulation in E. coli. Both of them were considered to be derived forms of the FLO1 gene. Therefore, deletion events were most likely caused by recombinatorial rearrangements in the middle repeated region in the FLO1 gene. The deleted and insert repeat region mainly contained one repeat unit, TFTSTSTEMTTVTGTNGVPTDETVIVIRTP (Fig. 7), suggesting that this repeat unit is the hot spot of recombination in the Flo1 protein.

The recombinant yeast strain YTSC-L bearing the coding sequence of FLONL was constructed. Cells of YTSC-L strongly flocculated, suggesting the short ORF of FLONL definitely conferred flocculation ability, even though the FloNL protein lost a majority of the C-terminal of the Flo1 protein. Resequencing of the FLONL gene excluded the possibility that the short ORF was due to a mere sequencing error. A potential GPI-modification site was found in the C-terminal (amino acid 1110) of the FloNS protein by GPI-som and big-PI predictor software available online (http://mendel.imp.ac.at; http://genomics.unibe.ch), whereas none was detected in the FloNL protein, indicating that the FloNL protein could not be anchored to the plasma membrane in the same way as the Flo1 and FloNS proteins. More experiments are necessary to establish whether the FloNS and FloNL proteins rely on a different anchoring mechanism which leads to stabilization of the protein in the cell wall.

All type of sugars such as mannose, glucose, and galactose used in this study inhibited flocculation of YTS-L and YTS-S greatly, indicating the NewFlo-type flocculation in yeast. Moreover, galactose, as a non-inhibitor for both Flo1- or NewFlo-type flocculation, strongly inhibited flocculation of YTS-S, whereas it exhibited a slightly weaker inhibition on YTS-L. This result suggests that the missing region in the C-terminal of the FloNS protein is involved in galactose recognition. The internal deletion feature of FLONL and FLONS (amino acid residues 303–527 and amino acid residues 799–979) compared with FLO1 demonstrated that the deleted repeats in the FLO1 gene induced the phenotype conversion from Flo-type to NewFlo-type flocculation. Verstrepen et al. inserted the URA3 marker into different hot spots among 18 intragenic repeats, NSTFTSTSTELTTVTGTNGVRTDETIIVIRTPTTATTAITTTEPW, producing variant flocculation levels in yeast (Verstrepen et al., 2005). In their study, the changed adhesion strength among yeast cells was believed to generate the variability of flocculation ability. In this case, however, the higher binding affinities of FloN proteins for sugars triggered the flocculation phenotype conversion from Flo1-type to NewFlo-type. It is believed that the internal repeat deletion in the Flo1 protein functionally influenced the sugar-binding strength. Though further research is expected to clarify the protein–sugar interaction in the Flo1 protein, we strongly suspect that deletions in the repeat regions activated a latent, high-affinity conformational state of FloN proteins for both the C-2 hydroxyl group of glucose and the C-4 hydroxyl group of galactose. This active conformation, referred to as ‘open state’, exhibited increased affinities for all of the sugars detected in this experiment. The higher affinities for sugars of the FloNL protein implies that the C-terminal in the Flo1 protein can also influence sugar binding strength, despite the indirect demonstration that the N-terminal region of the Flo1 protein contains the sugar recognition domain. The missing region of the C-terminal in the FloNL protein contains a large number of repeat units. Therefore, different affinities for sugars in the two proteins show a quantitative relationship to the repeat number in the Flo1 protein: the more repeats, the greater the sugar binding strength.

In summary, flocculation type conversion caused by repeated deletions in FLO1 was revealed in this study. This is the first report of a change-of-function mutation occurring in the FLO1 gene. Repeats in the middle of the Flo1 protein functionally affected the flocculation phenotype conversion in yeast. This research provides good material to study the relationship between the genetic variation and functional diversities of the Flo1 protein. Because the characteristics and functions of very few cell wall proteins have been identified, the FloN proteins could represent valuable models for deciphering the mechanisms of mannoprotein incorporation in the yeast cell wall. Moreover, the yeast strains displaying NewFlo-type flocculation ability are predicted to have great potential for industrial applications in the future.

Acknowledgements

This research was supported by the National Science Foundation of China (grant #30370025).

References