OUP user menu

Characterization of a novel tyrosine permease of lager brewing yeast shared by Saccharomyces cerevisiae strain RM11-1a

Fumihiko Omura, Haruyo Hatanaka, Yoshihiro Nakao
DOI: http://dx.doi.org/10.1111/j.1567-1364.2007.00310.x 1350-1361 First published online: 1 December 2007


In Saccharomyces cerevisiae yeast, the uptake of aromatic amino acids is mediated by the relatively specific permeases Tat1p, Tat2p, Bap2p, and Bap3p, as well as by two other permeases with broader specificities: Gap1p and Agp1p. Here, a novel permease gene TAT3 (Tyrosine Amino acid Transporter) identified in the S. cerevisiae-type subset genome of the lager brewing yeast strain Weihenstephan Nr.34 (34/70) is reported. The TAT3 sequence was also found in the genome of S. cerevisiae strain RM11-1a, but not in S. cerevisiae strain S288C. Tat3p showed a significant similarity to Penicillium chrysogenum ArlP permease, which has transport activity for aromatic amino acids and leucine. When overexpressed in ssy1Δgap1Δ mutant cells, Tat3p exhibited a tyrosine transport activity with an apparent Km of 160 μM. TAT3 transcription in lager brewing yeast was subjected to nitrogen catabolite repression in a manner similar to that of GAP1. Furthermore, the subcellular localization of Tat3p–green fluorescent protein (GFP) fusion protein was dependent on the quality of the nitrogen source, indicating a post-translational control of Tat3p function.

  • amino acid permease
  • tyrosine transport
  • aromatic amino acid
  • lager yeast
  • Saccharomyces pastorianus
  • Saccharomyces cerevisiae


Yeast cells take up amino acids from the surroundings both for protein synthesis and for use as sources of nitrogen. The amino acid transport across the plasma membrane is ensured by a number of permeases (André, 1995; Sophianopoulou & Diallinas, 1995). The Saccharomyces cerevisiae genome encodes 24 such members, including some lacking any documented functions (Nelissen et al., 1997; Regenberg et al., 1999). These S. cerevisiae permeases belong to the amino acid permease (AAP) family (André, 1995), which is in turn a member of the transporter superfamily for amino acids, polyamines, and cholines (APC) shared by fungi, plant, animal, and bacteria (Nelissen et al., 1997; Wipf et al., 2002). Most of the permeases are specific for one or several related l-amino acids, whereas the general AAPs Gap1p (Jauniaux & Grenson, 1990) and Agp1p (Schreve et al., 1998; Iraqui et al., 1999) have broader substrate specificity spectra. They exhibit different properties with regard to substrate affinity, transport capacity, and regulation for functional expression, allowing yeasts to take up amino acids under different physiological conditions (Sophianopoulou & Diallinas, 1995; Regenberg et al., 1999; Magasanik & Kaiser, 2002). Two methionine permeases Mup1p and Mup3p are distantly related to the other AAPs (André, 1995; Isnard et al., 1996; Nelissen et al., 1997; Wipf et al., 2002), and the whole methionine transport system is regulated both transcriptionally and posttranscriptionally by different ubiquitin-dependent mechanisms (Menant et al., 2006).

Within the AAP family members, Gap1p (Jauniaux & Grenson, 1990), the proline permease Put4p (Jauniaux et al., 1987), and the acidic AAP Dip5p (Regenberg et al., 1998) are subjected to nitrogen catabolite repression (NCR), by which their transcription is repressed when preferred nitrogen sources such as ammonia and glutamine are available. The expression of some low-capacity, relatively specific, and high-affinity permeases is transcriptionally induced by sensing of an extracellular amino acids via the Ssy1p-Ptr3p-Ssy5p sensor system (Didion et al., 1998; Iraqui et al., 1999; Forsberg et al., 2001; Kodama et al., 2002; Wu et al., 2006), exemplified by the branched-chain AAPs Bap2p and Bap3p (Didion et al., 1996; de Boer et al., 2000), the tyrosine and tryptophan permease Tat1p (Schmidt et al., 1994; Bajmoczi et al., 1998), and the histidine permease Hip1p (Crabeel & Grenson, 1970; Tanaka & Fink, 1985).

With respect to posttranslational regulation, Gap1p activity is downregulated by endocytosis and vacuolar degradation when a preferred nitrogen source is available. The downregulation is preceded by dephosphorylation and Npi1p/Rsp5p-dependent ubiquitination of the permease (Hein et al., 1995; Stanbrough & Magasanik, 1995; Springael & André, 1998). Gap1p function is also regulated through intracellular sorting systems: poly-ubiquitination of Gap1p at the cytoplasmic N-terminus, which requires Npi1p and Bul1p/Bul2p activities, serves as a signal for the direct sorting of Gap1p from the Golgi to the vacuole without reaching the plasma membrane (Helliwell et al., 2001; Soetens et al., 2001). Furthermore, a nonubiquitinateable form of Gap1p (Lys9→Arg, Lys16→Arg), which is constitutively localized at the plasma membrane, can be rapidly and reversibly inactivated by its own activity of amino acid transport (Risinger et al., 2006). Tat2p, a tryptophan permease, is also downregulated posttranslationally in a ubiquitination-dependent manner (Beck et al., 1999). Upon starvation for nitrogen or treatment with rapamycin of the cells, Tat2p is delivered directly from the Golgi to the vacuole without being routed to the plasma membrane. In addition, when the cells are subjected to hydrostatic pressure, the Tat2p activity is inhibited by a pressure-induced degradation governed by multiple ubiquitin-specific proteases (Abe & Horikoshi, 2000; Miura & Abe, 2004). The degradation of Bap2p is induced by nutrient starvation, and is largely dependent on cellular ubiquitination and endocytosis machineries (Omura et al., 2001a). Finally, a putative serine/threonine protein kinase, Npr1p, controls the activity of Gap1p (Grenson, 1983), Put4p (Vandenbol et al., 1987, 1990), and Tat2p (Schmidt et al., 1998), and it modifies the phosphorylation status of the N-terminal domain of Bap2p in response to starvation (Omura & Kodama, 2003).

The genome sequencing of the yeast S. cerevisiae has shown the existence of duplicated genes, constituting more than 30% of the genome, leading to the hypothesis of a whole-genome duplication (Wolfe & Shields, 1997; Wong et al., 2002). The sequence comparison of a selected set of 38 genes of six different Hemiascomycete yeasts has revealed that a whole-genome duplication took place in the progenitor of the modern Saccharomyces and Kluyveromyces yeast lineages about 150 million years ago, and that the duplication was followed by asynchronous differentiation and specialization (Langkjær et al., 2003). The differential evolutional drift of one of the two gene copies is thought to be a major force for adaptation to novel ecological niches and further speciation (Kellis et al., 2003). Analysis of transmembrane transporters from five complete genome sequences spanning the entire Hemiascomycete phylum, using a semiautomatic classification system, has clarified that AAPs are included in one of the most variable subfamilies belonging to the amino acid-polyamine-organocation family (De Hertogh et al., 2006).

Lager brewing yeast, one of the most important industrial yeasts, is an allopolyploid hybrid species formed by S. cerevisiae and a closely related Saccharomyces species. It therefore contains parts of two diverged genomes: one derived from S. cerevisiae and the other derived from a different Saccharomyces yeast (reviewed in Kodama et al., 2006). Accordingly, in almost all cases, two divergent alleles of a gene of interest are found in lager brewing yeast. For instance, the identity between the S. cerevisiae-type BAP2 permease gene and the non-S. cerevisiae-type counterpart, LgBAP2, is 80.3% at the nucleotide level. However, the intergenic regions are more diverged, and consequently, clear differential expression patterns between the two orthologues are observed (Kodama et al., 2001, 2006).

Here, a novel AAP is reported with a tyrosine transport activity found during the course of the whole genome sequencing of lager brewing yeast Weihenstephan Nr.34 (34/70). The corresponding gene TAT3 (Tyrosine Amino acid Transporter) does not exist in the genome of S. cerevisiae standard laboratory strain S288C, but does exist in another S. cerevisiae strain, RM11-1a, originating from a natural isolate (Mortimer et al., 1994). The phylogenetic aspects and possible functions of Tat3p permease in lager brewing yeast are discussed.

Materials and methods

Yeast strains, media, and culture conditions

The S. cerevisiae strains used in this work are listed in Table 1. Yeast cells were maintained on YPD (Rose et al., 1990). Synthetic complete (SC) medium and synthetic dextrose minimal medium (Fig. 7, Am) were prepared as described elsewhere (Rose et al., 1990). Medium B was identical to SC medium, except that the concentrations of all the amino acids were twofold higher, and that ammonium sulfate was not included. SD-Pro medium and SD-urea medium were prepared using a yeast nitrogen base without ammonium sulfate and amino acids (Difco) supplemented with 2% glucose, and 0.1% proline or urea, respectively. Minimal media with amino acids as the sole nitrogen source were prepared as SD-Pro, except that proline was replaced by one of other amino acids (0.1%). Cultures of lager brewing yeast and a temperature-sensitive strain RH1597 were carried out at 25 °C. Other S. cerevisiae strains were cultured at 30 °C, or as otherwise indicated.

View this table:

Yeast strains

StrainRelevant genotypeSource or references
23346cMATaura3Hein . (1995)
27038aMATaura3 npi1Hein . (1995)
RH1602MATaura3 leu2 his4 bar1Raths . (1993)
RH1597MATaura3 leu2 his4 bar1 end4-1Raths . (1993)
YK007MATassy1TDH3p-SFA1 gap1AUR1-CKodama . (2001)
Weihenstephan Nr.34 (34/70)Lager brewing yeast
  • * Provided by Fachhochschule Weihenstephan, Freising, Germany.


Expression of TAT3 mRNA in lager brewing yeast. Total RNA was prepared from strain Weihenstephan Nr.34 (34/70) grown to the logarithmic phase in SC medium, ammonia-based synthetic minimal medium (Am), SD-Pro medium (Pro), ammonia-based medium plus 2 mM tyrosine (Am+Tyr), and SD-Pro plus 2 mM tyrosine (Pro+Tyr). The samples were then processed for Northern blot analysis using labeled DNA fragments corresponding to the entire or part of ORFs of TAT3. GAP1, PUT4, DIP5, BAP2, and ACT1 genes as probes.

Plasmid constructions and yeast transformation

DNA manipulations were performed according to standard protocols (Sambrook et al., 1989). The whole genome of lager brewing yeast Weihenstephan Nr.34 (34/70) was determined by ‘shotgun sequencing’ (Kodama et al., 2006) with the aid of an assembling software arachne (Batzoglou et al., 2002) (Y. Nakao, manuscript in preparation). Based on the sequence information obtained, oligonucleotides 1 and 2 (Table 2) were designed as primers used in the PCR for amplification of the TAT3 ORF sequence. The PCR was performed using the Weihenstephan Nr.34 (34/70) chromosomal DNA as the template. The TAT3 ORF was prepared as a 1783-bp DNA fragment tagged with SacI and NotI restriction sites at the 5′- and 3′-ends, respectively. The integrity of the PCR product containing TAT3 ORF was verified by DNA sequencing. For construction of a 2-μm- type expression vector plasmid, PCR was carried out with the oligonucleotides 3 and 4 (Table 2) as the primers and plasmid pYCGPY2 (Kodama et al., 2002) as the template to obtain a DNA fragment containing multi cloning sites (SacI–NotI–BamHI) and a TDH3 terminator sequence. After digestion with SacI and BglII, the 150-bp fragment was inserted into the SacI–BamHI sites of the 2-μm vector pYE22mG (Omura et al., 1995) to give pYEG-not (Fig. 4a). The 1.8-kb fragment corresponding to TAT3 ORF was digested with SacI and NotI, and ligated with the SacI–NotI gap of pYEG-not to result in pYEG-TAT3 (Fig. 4a). The gene encoding Tat3p–green fluorescent protein (GFP) fusion protein was constructed as follows: PCR was performed with oligonucleotides 5 and 6 as primers, along with the 1.8-kb SacI–NotI TAT3 ORF fragment as the template, to obtain the 840-bp fragment corresponding to the latter half of TAT3 ORF, whose stop codon was converted to a BamHI site allowing in-frame fusion to the GFP gene. The resultant fragment was digested with NarI and BamHI, and was ligated with the 850-bp SacI–NarI fragment for the former half of TAT3 ORF, the 700-bp BamHI–NotI GFP-encoding fragment from pYES-GFP (Omura et al., 2001b), and the NotI–SacI digest of centromeric vector pYCGPY2 (Kodama et al., 2002) to obtain pYC-TAT3-GFP (Fig. 4b).

View this table:

Oligonucleotides used in this study

  • The lowercase nucleotides provide restriction sites.


Schematic representation of plasmid DNAs used in this study. (a) Plasmid pYEG-not is a 2-μm-type shuttle vector harboring ampicillin resistance gene (Ampr), neomycin analogue G418 resistance gene (G418r), and constitutive TDH3 promoter (PTDH3) juxtaposed with multi-cloning sites and TDH3 terminator (T). TAT3 ORF was inserted into the SacI–NotI sites of pYEG-not to yield pYEG-TAT3 for constitutive expression. (b) The DNA fragment (TAT3–GFP) encoding the Tat3p–GFP fusion protein was inserted downstream of the PYK1 promoter (PPYK1) in the centromeric plasmid pYCGPY2 [16] to yield pYC-TAT3-GFP, which drives constitutive expression of the fusion protein. ARS, autonomously replicating sequence; CEN, centromeric sequence; ori, replication origin. Restriction enzymes that cut the plasmid once are shown.

Transformation of yeast cells was performed using the standard lithium acetate method (Rose et al., 1990). Selection for the positive clones was carried out on a YPD plate supplemented with 300 μg mL−1 of the neomycin analogue G418.

Amino acid assimilation test

Medium B supplemented with 300 μg mL−1 G418 (10 mL) was inoculated with YK007 cells carrying either pYEG-TAT3 or control plasmid pYEG-not to give 2 × 107 cells mL−1, and cultivated aerobically at 30 °C for 24 h. The cultures were centrifuged to remove the cells. The supernatants were diluted 10-fold with 0.02 N HCl and analyzed with an amino acid analyzer (model L-8800, Hitachi, Tokyo, Japan). The amount of assimilated amino acids was estimated by calculating the difference in the concentration of amino acids in the medium before and after yeast cultivation.

Tyrosine transport assay

Overnight cultures of YK007 cells containing either plasmid pYEG-TAT3 or pYEG-not were grown in SD-urea medium, diluted in the same medium, and grown to the logarithmic phase (OD660 nm of 0.8) at 30 °C. Cells were harvested from 90 mL of culture, washed in minimal medium without any nitrogen source, and resuspended to an OD660 nm of 100 in the same medium. After 10 min of preincubation at 30 °C, 100 μL of cells were added to 120 μL of the same medium containing 50 μM to 2 mM l-tyrosine with 0.075 μCi of [14C]l-tyrosine. The assay was performed at 30 °C, and stopped after a 5-min incubation by the addition of 7 mL of ice-cold water, immediately followed by filtration through a glass fiber filter. The collected cells were washed twice with 3 mL of ice-cold water, dried, and counted in a scintillation counter.

Northern blot analysis

Isolation of total RNA and subsequent agarose gel electrophoresis and blotting were performed according to the standard method (Rose et al., 1990). After hybridization, the blot was washed for 30 min with 0.1 × SSC (1 × SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at 70 °C, which was stringent enough to distinguish the transcripts originating from S. cerevisiae- and non-S. cerevisiae-type subgenomes (Kodama et al., 2001). Signal detection was performed with the aid of a digoxigenin chemiluminescent detection system (Roche Diagnostic). The entire ORF sequences of PUT4, DIP5, BAP2, and ACT1, and part of GAP1 ORF (nucleotide no. 689–1809) and TAT3 ORF (no. 840–1674) were used for the preparation of labeled probes, according to the manufacturer's instructions.

Fluorescence microscopy

Saccharomyces cerevisiae strains RH1597, RH1602, 27038a, and 23346c were transformed with plasmid pYC-TAT3-GFP. The transformant cells derived from 27038a and 23346c were grown in the media indicated in Fig. 8 at 30 °C to the logarithmic phase, and were subjected to fluorescence microscopy. The transformant clones from the temperature-sensitive strains RH1597 and RH1602 were grown at 25 °C and shifted to 37 °C for 1 h before the microscopy. Cells were visualized with a fluorescence microscope (model Eclipse E600, Nikon, Tokyo, Japan), and images were taken at × 100 using a color-chilled 3CCD camera (model C5810, Hamamatsu Photonics, Shizuoka, Japan).


Localization of the Tat3p–GFP fusion protein. Strains 23346c (NPI1), 27038a (npi1), RH1602 (END4), and RH1597 (end4) were transformed with pYC-TAT3-GFP. The 23346c transformant cells were grown at 30°C to the logarithmic phase in SD-Pro medium (Pro), SD-urea medium (Urea), or SC medium, and were subjected to microscopy. The 27038a, RH1602, and RH1597 transformant cells were grown in SC medium. RH1602- and RH1597-derived cells were grown at 25°C, and shifted to 37°C for 1 h before observation. Scale bar, 5 μm.

Nucleotide sequence accession numbers

The nucleotide sequence data of TAT3 from lager brewing yeast have been deposited with DDBJ (http://getentry.ddbj.nig.ac.jp/) under the accession number AB302221.


Identification of an ArlP homologue in lager brewing yeast

Lager brewing yeast (or Saccharomyces pastorianus) is allotetraploid, and is believed to originate from a natural hybrid between S. cerevisiae and a closely related Saccharomyces species (Kodama et al., 2006). In order to gain a better understanding of the lager brewing yeast, the whole genome sequence of one S. pastorianus strain, Weihenstephan Nr.34 (34/70), was determined by ‘shotgun sequencing’ (Kodama et al., 2006). To this end, a total of 348 001 sequence reads were performed, and the sequences obtained constituted c. 160 million base (bp) of DNA, corresponding to a 6.5-fold coverage of the genome. The contigs obtained were searched for possible syntenies by comparing with a database of S. cerevisiae strain S288C (Saccharomyces Genome Database; SGD), and a lager brewing yeast/S. cerevisiae comparative genomic map was constructed (Y. Nakao, manuscript in preparation). The lager brewing yeast genome turned out to be 23.2 million bp large, and to consist of subsets of an S. cerevisiae-like genome and a non-S. cerevisiae-like genome, whose average sequence identity in ORFs to the S. cerevisiae genome was 98.9% and 84.4%, respectively. As to the chromosomal structure, there were three kinds of chromosomes in the sequenced strain: S. cerevisiae-type and non-S. cerevisiae-type chromosomes, and various mosaic chromosomes composed of the former two types (Kodama et al., 2006). During the process of gene annotation, some ORFs found in the lager brewing yeast genome exhibited no significant identities to any of the reported ORFs in S. cerevisiae S288C. One of these genes, later designated TAT3, was located in the subtelomeric region of the right arm of S. cerevisiae-type chromosome XIII in lager brewing yeast (Fig. 1). Although the TAT3 sequence does not exist in the S. cerevisiae S288C genome, the region encompassing SAM4, SAM3, TAT3, HXT13, and YEL070W (DSF1) (Fig. 1) was found as a conserved synteny in the subtelomeric region of chromosome XVI left arm in another strain of S. cerevisiae, RM11-1a, originating from a natural isolate found in fermenting grape musts (http://www.broad.mit.edu/annotation/genome/saccharomyces_cerevisiae/).In S. cerevisiae S288C, on the other hand, only the SAM4 and SAM3 genes are located on the left arm of chromosome XVI, while HXT13 and YEL070W (DSF1) are on the left arm of chromosome V. To achieve an understanding of the function of the TAT3 gene product (Tat3p), all nonredundant sequences in the GenBank CDS translations database were searched for possible Tat3p homologues. Tat3p showed a significant similarity (E-value lower than 10−94) to Penicillium chrysogenum ArlP permease with a transport activity for tyrosine, phenylalanine, tryptophan, and leucine (Trip et al., 2002). Figure 2 shows the overall amino acid sequence alignment of Tat3p and ArlP. The homology search also revealed that Tat3p shares its sequence features with putative AAPs from Candida albicans (E-value lower than 10−141) and Aspergillus fumigatus (E-value lower than 10−96). The phylogenetic relationships among Tat3p and these fungal permeases, along with other known S. cerevisiae permeases and lager brewing yeast's LgBap2p, are shown in Fig. 3.


Schematic illustration of Saccharomyces cerevisiae-type and non-S. cerevisiae-type structures of chromosome XIII in lager brewing yeast Weihenstephan Nr.34 (34/70). The open bar with a black circle (which depicts the centromere) indicates the S. cerevisiae-type chromosome XIII. The dotted bar indicates the non-S. cerevisiae-type chromosome XIII with a translocation to the S. cerevisiae-type at YMR302C locus (arrow). Subtelomeric regions of the right arm of chromosome XIII are shown in detail below. Black boxes represent the genes whose synteny is conserved in the right arm of chromosome XIII in S. cerevisiae S288C; the genes shown as open boxes are found in this order in the left arm of chromosome XVI in S. cerevisiae RM11-1a. The hatched boxes represent the genes whose homologous sequences are found in S. cerevisiae chromosomes other than XIII or XVI.


Amino acid sequence comparison of lager brewing yeast Tat3p and Penicillium chrysogenum ArlP. The sequences were aligned with DNA analysis software vector nti suite 7.0 (World Fusion). Identical amino acids between the two proteins are shaded in black, and similar amino acids are shaded in gray. The amino acid sequence number of Tat3p is indicated.


Phylogenetic relationship among Tat3p and related permeases. The predicted amino acid sequences of permeases from Candida albicans (NCBI database XP_718551) and Aspergillus fumigatus (XP_748248), Penicillium chrysogenum ArlP (Trip et al., 2002), Saccharomyces cerevisiae AAPs (Put4p, Dip5p, Agp3p, Tat2p, Gap1p, Bap3p, and Bap2p), and lager brewing yeast permeases LgBap2p (Kodama et al., 2001) and Tat3p were aligned with vector nti software. The phylogenetic tree was drawn with treeview software version 1.6.6 (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html).

Functional analysis of Tat3p permease

The yeast S. cerevisiae possesses a variety of AAPs that exhibit different properties with respect to substrate affinity, specificity, capacity, and regulation (André, 1995; Sophianopoulou & Diallinas, 1995; Nelissen et al., 1997; Regenberg et al., 1999), and one particular amino acid can be taken up by the cells through several distinct permeases (Rogenberg et al., 1999). In order to facilitate the estimation of the amino acid transport activity of Tat3p, TAT3 gene was placed downstream of the constitutive TDH3 promoter (Fig. 4), and expressed in an ssy1Δgap1ΔS. cerevisiae strain YK007 (Iraqui et al., 1999; Kodama et al., 2001). In this strain, intrinsic transport activities for isoleucine, leucine, valine, and aromatic amino acids are largely impaired due to the lack of the general amino acid transporter (Gap1p), and, due to the lack of the Ssy1p-Ptr3p-Ssy5p sensor system necessary for the function of several specific AAPs (Jauniaux & Grenson, 1990; Didion et al., 1998; Forsberg et al., 2001; Kodama et al., 2002). Tat3p-expressing yeast and the control yeast were tested for their ability to grow on minimal media containing one of the naturally occurring amino acids [0.1% (w/v)] as the sole nitrogen source (Fig. 5). As expected, the cell growth of the control strain was poor with isoleucine, leucine, and aromatic amino acids. On the other hand, the growth with aspartic acid, glutamic acid, arginine, serine, asparagine, and glutamine was as good as with SC medium by the ssy1Δgap1Δ background, suggesting that another NCR-sensitive permease, Agp1p (Schreve et al., 1998; Iraqui et al., 1999), was functional in the uptake of some of those amino acids. However, it was obvious that the expression of Tat3p significantly restored the growth of this strain with tyrosine as a nitrogen source (Fig. 5).


Growth test of ssy1Δgap1Δ cells carrying TAT3 expression plasmid. YK007 cells with TAT3-expression plasmid pYEG-TAT3 (lower panel) or with empty vector DNA (upper panel) were serially diluted with water and spotted onto minimal media with the indicated amino acid (0.1%) as the sole nitrogen source. The plates were incubated at 30°C for up to 6 days. The SC plate was used as a control.

With these experiments, however, it was impossible to judge whether or not Tat3p had a transport activity for amino acids other than tyrosine because the growth of the YK007 cells with or without the TAT3 gene did not show significant difference upon cultivation with all other amino acids. Thus, the amino acid assimilation of Tat3p-expressing cells cultivated in a medium containing a mixture of all 20 amino acids was investigated. The yeast cells were cultivated aerobically in a synthetic liquid medium (medium B, see ‘Materials and methods’) at 30 °C for 24 h, and the decrease of each amino acid in the medium during cultivation was analyzed, and compared with the same numbers for YK007 control cells with no TAT3 gene expression (Fig. 6). Owing to poor separation of the peaks corresponding to serine and asparagine with the present HPLC system, these amino acids are shown as the sum of the two, and the same applies in the case of threonine and glutamine. By the end of the cultivation, only two amino acids, arginine and lysine, were completely exhausted from the culture (data not shown). The amino acid assimilation profile in this assay with ssyΔgap1Δ control cells showed good agreement with the results from an uptake analysis using 14C-labeled amino acids as reported previously (Regenberg et al., 1999). Clearly, the expression of Tat3p led to a noticeable improvement in the uptake of tyrosine, phenylalanine, tryptophan, and leucine, strongly suggesting that Tat3p is capable of transporting these aromatic and hydrophobic amino acids. On the contrary, the uptake of proline and aspartic acid by the Tat3p-expressing cells was lower than that of the control cells. This might be explained by repression of the NCR-sensitive PUT4 and DIP5 (responsible for the transport of proline and aspartic acid, respectively) by the increased intracellular pools of amino acids taken up by Tat3p.


Amino acid assimilation in the ssy1Δgap1Δ strain with TAT3 expression plasmid. YK007 cells with pYEG-TAT3 (black bar) or with empty vector DNA (open bar) were cultivated aerobically in medium B at 30°C for 24 h. The amount of assimilation of amino acids was estimated as the differences between the amounts in the medium before and after yeast cultivation. Data are presented as means of three independent experiments with SDs.

To determine the kinetic characteristics of the Tat3p permease, the uptake of 14C-labeled tyrosine was analyzed with the YK007 cells in which Tat3p is constitutively expressed. The cells were grown to the logarithmic phase in minimal medium with urea as the sole nitrogen source. The control cells harboring an empty vector were also subjected to the tyrosine-uptake assay, and the net contribution to tyrosine uptake by Tat3p was obtained by subtraction of the background value given by the control cells. The apparent Km of tyrosine uptake as derived from Lineweaver–Burk plots was c. 160 μM (data provided as supplementary material Fig. S1).

Expression of Tat3p in lager brewing yeast

Some AAPs are preferentially active when only so-called ‘nonpreferred’ nitrogen sources are present (e.g. urea and proline) and inactive with ‘preferred’ sources (e.g. glutamine and ammonia). Nitrogen regulation at the transcriptional level is known as NCR, and a sophisticated network of transcriptional regulators is involved in NCR (Gln3p, Nil1p, Nil2p, Dal80p, and Ure2p) (Magasanik & Kaiser, 2002). On the other hand, another group of AAPs, exemplified by Bap2p, Bap3p, and Tat2p, are transcriptionally induced by particular extracellular amino acids, through the Ssy1p-Ptr3p-Ssy5p sensor machinery (Didion et al., 1998; Forsberg et al., 2001; Kodama et al., 2002; Wu et al., 2006). In order to examine the transcriptional regulation of TAT3 in response to nitrogen availability, total RNA was prepared from lager brewing yeast cells grown in either an amino acid-rich medium (SC medium), or a synthetic medium with ammonia or proline as the sole nitrogen source, and the RNA was subjected to Northern analysis with probes designed to detect transcripts of GAP1, PUT4, DIP5, BAP2, and TAT3 (Fig. 7). The expression of NCR-sensitive genes GAP1, PUT4, and DIP5 was induced by cultivation with proline. However, GAP1 and DIP5 were partially derepressed by ammonia, whereas PUT4 was not. The TAT3 expression profile was quite similar to that of GAP1, indicating that TAT3 expression is subjected to NCR. Addition of 2 mM tyrosine to ammonia-based medium resulted in even further repression of TAT3 transcription although tyrosine seemed to be one of the major substrates to be transported by Tat3p itself (Figs 5 and 6). The addition of tyrosine to the ammonia medium caused the same repressive effect in GAP1 expression, whereas the DIP5 expression was rather derepressed by the added tyrosine. The NCR-insensitive permease gene BAP2 was expressed well in the presence of the abundance of amino acids in SC medium, and the addition of tyrosine to ammonia-based medium increased the expression significantly (Fig. 7).

After establishing that TAT3 expression is under the control of NCR, the upstream region of the TAT3 gene was studied. NCR-sensitive genes contain specific binding sites for GATA factors and for some auxiliary activators responsible for NCR. However, under certain conditions, a series of three 5′-GATAAG-3′ or 5′-GATAAGATAAG-3′ sites have been demonstrated to be adequate to exert a basic NCR regulation (Minehart & Magasanik, 1992; Blinder & Magasanik, 1995; Magasanik & Kaiser, 2002). It appeared that the TAT3 upstream region contains two complete 5′-GATAAG-3′ sites, at the positions −168 to −163 and −144 to −139, and an incomplete sequence 5′-GATAAc-3′ at −193 to −188. As to the 11-nucleotide consensus sequence (5′-GATAAGATAAG-3′), two incompletely matched sequences, 5′-GATAAGcaAAG-3′ (−168 to −158) and 5′-GATtcGATAAG-3′ (−149 to −139), were identified. It is highly likely that these sequences are functioning as NCR-responsible cis-elements upstream of the TAT3 ORF.

Intracellular localization of Tat3p

In addition to its transcriptional control, Gap1p permease undergoes posttranslational regulations depending on the quality of the nitrogen source available (Magasanik & Kaiser, 2002). Delivery to the plasma membrane of Gap1p is controlled by intracellular sorting mechanisms. With nonpreferred nitrogen sources, Gap1p engages in a recycling loop between the trans-Golgi and prevacuolar compartments, and occasionally is transported to the vacuole for degradation without being delivered to the plasma membrane (Helliwell et al., 2001; Soetens et al., 2001). On the other hand, Tat2p permease is regulated differently from Gap1p in terms of response to the nitrogen source, but it shares some of the processes for the intracellular sorting mechanism with Gap1p; vacuolar degradation of Tat2p permease depends on its ubiquitination state, and newly synthesized Tat2p can be targeted directly to the vacuole without reaching the plasma membrane (Beck et al., 1999). Tat3p was tagged with GFP at its carboxyl terminus, so as to be detected with ease. The gene construct encoding the Tat3p–GFP fusion protein was placed under a constitutive PYK1 promoter and was introduced into S. cerevisiae strains 23346c (NPI1), 27038a (npi1), RH1602 (END4), and RH1597 (end4). In wild-type cells (23346c) grown in proline or urea as the sole nitrogen source, the Tat3p-GFP fusion protein localized predominantly at the plasma membrane (Fig. 8). However, when the same cells were grown in the SC medium containing ‘preferred’ nitrogen sources, the fusion protein could only barely be detected at the plasma membrane, and instead, reduced amount of fluorescence was detected in granular and vacuole-like structures in the cytoplasm (Fig. 8). In case Tat3p–GFP was expressed in npi1 mutant cells (Hein et al., 1995; Dunn & Hicke, 2001), where the function of E3 ubiquitin ligase needed for nitrogen catabolite inactivation is impaired, the fusion protein mainly localized at the plasma membrane even when the cells were grown in the SC medium (Fig. 8). Accumulation of the Tat3p–GFP fusion protein at the plasma membrane was also observed when the fusion construct was expressed in a temperature-sensitive endocytotic mutant (end4) at a restrictive temperature. These observations indicate that the Tat3p function is regulated posttranslationally as well as transcriptionally in response to the quality of the nitrogen source, and that the down-regulation of Tat3p requires cellular functions of ubiquitination and endocytosis. However, it is unlikely that the majority of Tat3p is directly sent to the vacuole for degradation without being routed to the plasma membrane upon exposure to preferred nitrogen sources. It is more likely that Tat3p inactivation can be accomplished by acceleration of its endocytosis from the plasma membrane, followed by the vacuolar degradation.


In this work, a novel high-affinity tryptophan permease gene, TAT3, was identified in a lager brewing yeast genome. Although no sequences identical to TAT3 appeared to exist in the S. cerevisiae S288C genome, an identical ORF was found in the genome of S. cerevisiae strain RM11-1a, a natural isolate from fermenting grape musts (Mortimer et al., 1994). In the lager brewing yeast genome, the subtelomeric region of the right arm of S. cerevisiae-type chromosome XIII is translocated to the region encompassing SAM4, SAM3, TAT3, HXT13, and YEL070W (DSF1) found in the left arm of chromosome XVI of S. cerevisiae RM11-1a (Fig. 1). This translocation seems to have occurred at the breakpoint within the homologous region between the two ORFs: YMR321C and SAM4. In general, lager brewing yeast contains S. cerevisiae-type and non-S. cerevisiae-type subsets of genome. However, the non-S. cerevisiae-type chromosome XIII is translocated to the S. cerevisiae-type chromosome XIII at YMR302C locus as shown in Fig. 1 (Kodama et al., 2006), which shows a possibility that both S. cerevisiae-type and non-S. cerevisiae-type chromosomes XIII bear the above-mentioned TAT3-containing region.

In comparison with other S. cerevisiae AAPs, Tat3p is rather distantly related to high-affinity specific permeases Bap2p and Bap3p (Fig. 3), both of which are under control of the Ssy1p-Ptr3p-Ssy5p sensor system. Tat3p seems to be structurally closer related to the NCR-sensitive Put4p and Dip5p permeases. In addition, Tat3p shows significant similarity to permeases from filamentous fungi and C. albicans (Figs 2 and 3). AAPs are known as one of the most variable groups among yeast transmembrane transporters (De Hertogh et al., 2006). One of the fungal orthologues of Tat3p, P. chrysogenum ArlP, has been shown to act as a transporter of aromatic amino acids and leucine (Trip et al., 2002), and this may imply that an aromatic amino acid-transporting activity implemented by Tat3p has at some point been physiologically meaningful, and as a consequence, the TAT3 gene has been conserved throughout the evolutionary period after the segregation of the progenitors of P. chrysogenum and Saccharomyces species. Although selective transport of aromatic amino acids is performed by Tat1p and Tat2p as well, these two permeases are both regulated in a different way than Tat3p. Accordingly, Tat3p possibly plays a role distinct from that of Tat1p and Tat2p, which constitutes part of the whole aromatic amino acid uptake system. On the contrary, S. cerevisiae S288C seems to have abandoned the TAT3 gene at some point in its evolutionary history, likely as a result of the unusual circumstances provided for this well-established laboratory strain. Although non-S. cerevisiae-type TAT3 was not found in lager brewing yeast genome, TAT3 orthologues were found in other Saccharomyces sensu stricto species with low E values (i.e. S. mikatae with an E-value of lower than 10−263; Saccharomyces kudriavzevii, <10−243; Saccharomyces bayanus, <10−229).

Tat3p was shown to transport tyrosine, using an ssy1Δgap1Δ strain with constitutive Tat3p expression (Fig. 4), by (1) growth assay on a tyrosine plate (Fig. 5), (2) measurement of amino acid assimilation during a 24-h cultivation in the presence of all 20 amino acids (Fig. 6), and (3) direct measurement of 14C-labeled tyrosine uptake. The results from Fig. 6 suggest that Tat3p likely has the ability to transport tryptophan, phenylalanine, and leucine in addition to tyrosine. These observations are well in line with the fact that the structurally similar P. chrysogenum ArlP (Fig. 2) recognizes those four amino acids, but neither isoleucine nor valine, which have a branched β-carbon atom (Trip et al., 2002).

NCR control and Ssy1p-Ptr3p-Ssy5p sensor-mediated induction by the presence of amino acids are two major transcriptional regulations for yeast AAPs (Didion et al., 1998; Iraqui et al., 1999; Forsberg et al., 2001; Kodama et al., 2002). Consistent with the literature evidence, GAP1, PUT4, and DIP5 were shown to be under control of NCR, whereas BAP2 was induced by the addition of an extra amino acid (Fig. 7). The transcriptional regulation of TAT3 was similar to that of GAP1, as TAT3 transcription was largely repressed when ammonia was used as a nitrogen source (Fig. 7). Interestingly, TAT3 transcription was subjected to stronger repression when tyrosine was added to the ammonia-based medium, whereas no such effect was observed when tyrosine was present in combination with proline, showing the complexity of the transcriptional regulation of AAPs. The cis-acting regulatory elements in the 5′-upstream region of NCR-sensitive genes have been reported (Minehart & Magasanik, 1992; Blinder & Magasanik, 1995), and it appeared that the region upstream of the TAT3 ORF contains two complete sequences and one incomplete 5′-GATAAG-3′ consensus sequence and two incomplete 5′-GATAAGATAAG-3′ sequences. It is plausible that some of these elements are functioning in lager brewing yeast as the cis-acting elements responsible for NCR regulation of the TAT3 gene.

Microscopic observation of the Tat3p–GFP fusion protein revealed that Tat3p undergoes a posttranslational regulation, depending on the nitrogen source available (Fig. 8). Thus, cultivation in nonpreferred nitrogen sources (proline or urea) led to efficient expression of Tat3p at the plasma membrane. Combined with the results from the transcriptional regulation of the TAT3 gene (Fig. 7), it can be concluded that Tat3p mainly functions when the cells are short of preferred nitrogen sources, whereas it is scarce when preferred nitrogen sources are abundant. Posttranslational regulation of Tat3p seems to be dependent on the cellular functions of the ubiquitination and endocytosis (Fig. 8). The intracellular localization of Tat3p was likely to be regulated by means of ubiquitination state of the permease as in the case of Gap1p and Tat2p (Magasanik & Kaiser, 2002). However, it seems that the majority of Tat3p is downregulated after being sent to the vacuole via the plasma membrane, because a mutation in the endocytosis pathway resulted in accumulation of Tat3p in the plasma membrane. It has been reported that the downregulation of Tat2p is promoted when the cells are placed under hydrostatic pressure, lower cultivation temperature, and weak organic acids (Abe & Horikoshi, 2000; Bauer et al., 2003; Miura & Abe, 2004). Tat2p function is also hampered by a mutation that affects the sterol synthesis pathway (Umebayashi & Nakano, 2003), and by treatment of the cells with the immunosuppressant FK506 or rapamycin (Schmidt et al., 1994; Beck et al., 1999). For a better understanding of the transporter system of aromatic amino acids in lager brewing yeast, it would be necessary to explore whether or not Tat3p function is also affected posttranslationally by those environmental and genetic factors.

The concentration of tyrosine in a standard wort prepared with 100% malt with no adjunct is normally slightly higher than 1 mM, and at the end of beer fermentation it usually decreases by half (data not shown). In the case of wort with an adjunct, the amino acid concentrations are lowered according to the wort/adjunct ratio. Considering that the apparent Km of Tat3p for tyrosine is 160 μM, and that this permease undergoes both NCR and posttranslational regulations, it is conceivable that Tat3p possibly functions later in the beer fermentation period when amino acids are less abundant. Tyrosine is also taken up by the cells through other permeases such as Tat1p, Tat2p, Bap2p, Bap3p, Gap1p, and Agp1p (Regenberg et al., 1999), the former four of which are induced by the presence of amino acids while Gap1p and Agp1p are downregulated. The results suggest that the Tat3p permease plays a role in transporting tyrosine and probably other aromatic amino acids, in concert with two other permeases, Gap1p and Agp1p, when most of the amino acids but proline are about to run out around the end of beer fermentation. Because the substrate specificities of Gap1p and Agp1p are considerably broad, it could be beneficial to have Tat3p as a specific permease for aromatic amino acids and leucine in case they are the majority of the remaining amino acids.

As mentioned above, the major tryptophan permease Tat2p is inactivated by various kinds of cellular stress (Schmidt et al., 1994; Beck et al., 1999; Abe & Horikoshi, 2000; Bauer et al., 2003). The beer fermentation process can provide larger brewing yeast cells with different types of stress, e.g. lower temperature, higher osmolarity of wort, higher ethanol concentration, and lower pH at the end of fermentation. It is plausible that having a differently regulated permease, Tat3p, can alleviate the impairment of aromatic amino acid uptake system when the Tat2p-based system is inactivated.

It is of great interest to understand the functions of as yet uncharacterized AAPs in relation to beer brewing conditions, which will help improve the quality of beer production. In order to improve the quality and efficacy of beer production, it is crucial to understand in detail the cellular uptake of nutrients, of which amino acids form a very important group. A first step towards clarification of the complex network of gene expression underlying amino acid accumulation by brewing yeast is to identify and characterize all participating AAPs. It is believed that the present work, describing a novel tyrosine permease, Tat3p, may contribute significantly to this understanding.

Supplementary material

The following supplementary material is available for this article:

Fig. S1. Lineweaver-Burk plots.


The authors thank Bruno André and Howard Riezman for the yeast strains, and Sandra Rainieri and Jørgen Hansen for critically reading the manuscript. The authors are grateful to Tomoko Shimonaga for her help in the initial part of this study. The expert technical assistance of Yumiko Itokui and Yachiyo Wada is gratefully acknowledged.


View Abstract