The yeast Saccharomyces kluyveri (Lachancea kluyveri), a far relative of Saccharomyces cerevisiae, is not a widely studied organism in the laboratory. However, significant contributions to the understanding of nucleic acid precursors degradation in eukaryotes have been made using this model organism. Here we review eukaryotic pyrimidine degradation with emphasis on the contributions made with S. kluyveri and how this increases our understanding of human disease. Additionally, we discuss the possibilities and limitations of this nonconventional yeast as a laboratory organism.
Pyrimidines are precursors of nucleic acids and many other biomolecules. The intracellular pools are regulated by the de novo synthesis and salvage pathways as well as catabolism, and have great importance for cellular function, for example replication of DNA, which requires well-balanced nucleotide pools.
In animals, pyrimidines are degraded by the reductive pathway, which is composed of three enzymatic steps converting uracil and thymine to β-alanine and β-aminoisobutyric acid, respectively (Traut & Loechel et al., 1984; Piskur et al., 1993). The first reaction is catalyzed by dihydropyrimidine dehydrogenase (DPD, EC 126.96.36.199/EC 188.8.131.52) yielding dihydropyrimidines. In the second reaction, dihydropyrimidinase (also dihydropyrimidine amidohydrolase, DHPase, EC 184.108.40.206) converts the dihydropyrimidines to β-ureidopropionic/β-ureidoisobutyric acid (Fig. 1). These compounds are subsequently converted to the corresponding β-amino acids by β-ureidopropionase (β-alanine synthase, βASase, EC 220.127.116.11). Furthermore, β-alanine may be catabolized to malonate semialdehyde by β-alanine-aminotransferase (βAAase, EC 18.104.22.168).
β-alanine is a structural analogue of the mammalian neurotransmitters γ-aminobutyrate and glycine and it has also been suggested that β-alanine may act as a functional neurotransmitter (Sandberg & Jacobson et al., 1981; Wang et al., 2003). In mammals, the main source of β-alanine is the catabolism of pyrimidines. Defects in the first enzymatic step are known to cause disorders in humans. The predominant symptoms are neurological, such as mental retardation, epileptic attacks and convulsions (Fiumara et al., 2003; Al-Sanna'a et al., 2005).
In clinical cancer treatment, the pyrimidine catabolism is highly relevant. Pyrimidine analogues such as 5-fluorouracil (5FU) are used for treatment of colorectal, breast, as well as head and neck cancer (van Kuilenburg et al., 2004), and it remains one of the most frequently prescribed chemotherapeutic anticancer drugs. In most patients, >80% of the administered 5FU is rapidly catabolized by the pyrimidine-degrading enzymes (Heggie et al., 1987). DPD, which catalyzes the first enzymatic step, has been shown to be a key determinant of toxicity of the treatment. Evidently, the study of this pathway could lead to improved treatment in clinics.
In yeast, β-alanine is required for biosynthesis of pantothenate and coenzyme A (White et al., 2001). The biosynthetic pathway of coenzyme A from pantothenate is currently considered a promising target for potential novel antimicrobial drugs (Spry et al., 2008).
Yeast as a model organism
In the late nineties, pyrimidine catabolism was still relatively poorly studied. Therefore, Gojkovic (1998) attempted to establish yeast as a simple eukaryotic model organism for pyrimidine catabolism. Among the Saccharomyces genus, only Saccharomyces kluyveri (Lachancea kluyveri) degrades uracil and the intermediates of the reductive pathway (Andersen et al., 2006). However, many other yeast and fungal genera can well degrade pyrimidines (Andersen et al., 2006; Piskur et al., 2007). Apparently, after the whole genome duplication event, the Saccharomyces cerevisiae lineage lost this ability. Through mutagenesis of S. kluyveri, three mutation groups were identified, pyd1, pyd2 and pyd3, which were unable to grow on uracil, dihydrouracil and β-ureidopropionic acid as sole nitrogen source, respectively. The natures of the PYD2 and PYD3 loci were identified by complementation with a genomic library and subsequent gene identification, and further with subcloning, overexpression and enzymatic studies.
In S. kluyveri, the PYD2 gene encodes a DHPase. Slime mold, plant and insect DHPases have been successfully cloned and shown to functionally replace a defective DHPase in S. kluyveri (Gojkovic et al., 2003). The three-dimensional structures of S. kluyveri DHPase and later also Dictyostelium discoideum (slime mold) DHPase (Fig. 2) have been determined and refined to 2.4 and 2.05 Å, respectively (Lohkamp et al., 2006). Both enzymes contain a dizinc catalytic center embedded in a (β/α)8-barrel structural core that is accompanied by a smaller β-sandwich domain. The two structures and their active sites are remarkably similar to each other, as well as to those of hydantoinases, dihydroorotases and other members of the amidohydrolase superfamily of enzymes. Complexes of yeast DHPase with dihydrouracil and N-carbamyl-β-alanine revealed the mode of substrate and product binding and allowed conclusions about what determines substrate specificity, stereoselectivity and the reaction direction among cyclic amidohydrolases (Lohkamp et al., 2006). Based on the observed similarities, catalysis is expected to follow a dihydroorotase-like mechanism but in the opposite direction and with a slightly different substrate. The knowledge gained from the Sk DHPase crystal structure has allowed structural modeling of human DHPase (Fig. 2) and clinically relevant genetic mutations can now be understood at the molecular level. For example, based on the crystal structures of eukaryotic DHP from S. kluyveri and D. discoideum, it has been suggested that the W360R and R412M mutations, found in several patients, lead to structural instability of the enzyme, which could potentially impair the assembly of the tetramer (van Kuilenburg et al., 2007).
Eukaryotic βASases can be divided into two subfamilies based on sequence homology. Although the subfamilies are functionally related, they are phylogenetically diverse (Gojkovic et al., 2001). One subfamily consists of βASases from higher eukaryotes, whereas the second consists of fungal βASases. The first structurally characterized βAS was the S. kluyveri enzyme (Lundgren et al., 2003). It consists of two identical subunits composed of a larger catalytic domain and a smaller domain mediating dimerization. Both domains have an α/β-topology. The catalytic domain contains the active site harboring a dizinc metal center. Surprisingly, this enzyme shows high structural homology to a family of dizinc-dependent exopeptidases, which suggests that they have a common origin (Andersen, 2008a).
Saccharomyces kluyveri lacks DPD and the Sk βASase is phylogenetically unrelated to the human counterpart. However, human βASase has been functionally expressed in S. kluyveri (Gojkovic et al., 2001). The first determined crystal structure of a higher eukaryotic βASase was of Drosophila melanogaster (Lundgren et al., 2008). The βASases of higher eukaryotes belong to the nitrilase superfamily and show the characteristic αββα-sandwich architecture (Pace & Brenner et al., 2001). However, the common core of the nitrilase-like enzymes is in D. melanogasterβASase extended by three helices and a β-strand at the N-terminus. The enzyme exists as a homooctamer assembled from tightly packed dimers. The active site contains a catalytic triad composed of Cys234, Glu120 and Lys197, which is conserved in all nitrilases. Drosophila melanogasterβAS shares 58% sequence similarity with the human enzyme. It is therefore likely that its crystal structure can be used to model the effects of mutations naturally occurring in human βASase.
Recently, the S. kluyveri PYD4 gene and its product, β-alanine aminotransferase, have also been characterized (Andersen et al., 2007). The determination of the enzyme's crystal structure may provide insights into this catabolic reaction for other organisms also.
Saccharomyces kluyveri as a laboratory organism
The advances made in nucleic acid precursor degradation reviewed above have established S. kluyveri as a useful laboratory organism. In fact, many methods for S. cerevisiae are applicable to S. kluyveri with little or no modifications. Auxotrophic strains (with mutations such as ura3−) have been developed in our lab and enable many traditional genetic techniques. An electroporation-based transformation protocol is described for transformation of yeast plasmids (Gojkovic et al., 2000) and plasmids commonly used in S. cerevisiae are replicated and maintained in S. kluyveri. Furthermore, a genomic library has been constructed (by F. Lacroute, Gif-sur-Yvette, France) for mutant complementation studies and genetic screens (Gojkovic et al., 2000).
In S. kluyveri, several options for genetic manipulation exist. Double mutants can be constructed by genetically crossing strains by mating and subsequent sporulation. Dissection of tetrads has been reported but rupture of the ascus sac by conventional enzymatic digestion is currently problematic, and in practice only random spore analysis is possible. Furthermore, gene targeting is relatively straightforward, and precise genetic manipulations are possible as integration by nonhomologous end joining appears to be rare. In this respect, S. kluyveri is comparable to S. cerevisiae. Our lab has routinely constructed knockout cassettes, using the kanMX cassette conferring resistance to G418, with c. 500 bp of end homology (Gojković et al., 2004).
In 2003, the entire genome of S. kluyveri was sequenced (Cliften et al., 2003), and since then several bioinformatics comparative studies have been published. Furthermore, the availability of spotted cDNA microarrays allows genomewide transcription analysis (available from M. Johnstons lab, St. Louis, MO). Microarray experiments have been successfully performed in our lab to study transcription in S. kluyveri knockout mutants as well as transcriptional response to growth with pyrimidines as sole nitrogen source (unpublished data).
Saccharomyces kluyveri has also been studied for the generation of mitochondrial respiratory-deficient, petite mutants, the ability to survive without oxygen, Crabtree effect and ethanol production. This yeast can easily be grown at large scale under fully controlled growth conditions (Møller et al., 2001).
Saccharomyces cerevisiae is one of the most studied organisms and the most advanced model to understand the eukaryote cell. So far, only a few yeasts have been thoroughly studied in the laboratory, but the recent advances in molecular biology, genomics and postgenomic approaches have put several nonconventional yeasts in focus. A few of them have now become successful model organisms to study traits, such as osmotolerance in Debaryomyces hansenii, telomere metabolism in Saccharomyces castellii, fungal pathogenicity in Candida glabrata, etc. As reviewed above, S. kluyveri is important as the model to understand nucleic acid precursor metabolism; however, in the search for renewable energy sources, it will also likely play an important role as a model to understand the yeast's ability to produce ethanol in the presence of oxygen.
The Swedish Research Council, Cancerfonden, KI-fonden and Lawski Foundation are acknowledged for their financial support. Anna Rasmussen and Olof Björnberg are acknowledged for their comments on an early version of this manuscript.