The synthesis of biotin, a vitamin required for many carboxylation reactions, is a variable trait in Saccharomyces cerevisiae. Many S. cerevisiae strains, including common laboratory strains, contain only a partial biotin synthesis pathway. We here report the identification of the first step necessary for the biotin synthesis pathway in S. cerevisiae. The biotin auxotroph strain S288c was able to grow on media lacking biotin when BIO1 and the known biotin synthesis gene BIO6 were introduced together on a plasmid vector. BIO1 is a paralog of YJR154W, a gene of unknown function and adjacent to BIO6. The nature of BIO1 illuminates the remarkable evolutionary history of the biotin biosynthesis pathway in S. cerevisiae. This pathway appears to have been lost in an ancestor of S. cerevisiae and subsequently rebuilt by a combination of horizontal gene transfer and gene duplication followed by neofunctionalization. Unusually, for S. cerevisiae, most of the genes required for biotin synthesis in S. cerevisiae are grouped in two subtelomeric gene clusters. The BIO1–BIO6 functional cluster is an example of a cluster of genes of "dispensable function," one of the few categories of genes in S. cerevisiae that are positionally clustered.

---

BIOTIN (vitamin H) is an essential vitamin. In Saccharomyces cerevisiae it is required for lipid metabolism (ROGGENKAMP et al. 1980), leucine metabolism (OHSUGI and IMANISHI 1985), and is a substrate of the biotin protein ligase (BPL1) (HOJA et al. 1998). Biotin is taken up by VHT1, a plasma membrane biotin transporter (STOLZ et al. 1999). Many prokaryotes are capable of synthesizing biotin, as are plants and some fungi. No animal is capable of synthesizing biotin. Some fungal species including Aspergillus sp. (SHCHELOKOVA and VOROB'EVA 1982) are capable of synthesizing biotin, and some such as Ashbya gossypii (DIETRICH et al. 2004) lack the genes necessary to synthesize biotin. Some fungal species that are incapable of synthesizing biotin still have portions of the biotin biosynthetic pathway and are capable of utilizing intermediates to synthesize biotin (PHALIP et al. 1999; WU et al. 2005). In S. cerevisiae, many wild isolates cannot synthesize biotin but nevertheless retain the BIO2, BIO3, and BIO4 genes from this biosynthetic pathway. Other isolates of this species are biotin prototrophs. The steps in biotin biosynthesis have been determined in Escherichia coli and Bacillus sp. (STREIT and ENTCHEVA 2003) (Figure 1). The first step of biotin biosynthesis, the formation of pimeloyl-CoA, is not conserved. Some species including E. coli, form pimeloyl-CoA from the condensation of three molecules of malonyl-coenzyme A (CoA) by the action of the bioC–bioH complex (IFUKU et al. 1994). Other species such as some species of Bacillus are capable of forming pimeloyl-CoA from pimelic acid with a single gene (bioW) (BOWER et al. 1995). The last four steps from pimeloyl-CoA to biotin are generally conserved (STREIT and ENTCHEVA 2003).

View larger version (14K): In this window In a new window Download PPT slide   FIGURE 1.— Diagram of the biotin biosynthesis pathway in S. cerevisiae. (A) Genomic position of the six known genes involved in biotin biosynthesis and intermediate transport. The chromosomal positions of BIO1/BIO6 genes are based on strain S288C and represent duplicated regions on chromosomes I and VIII. (B) Position of the genes in the biotin biosynthesis pathway. The biochemical intermediates in this pathway are based on work in E. coli, Bacillus sp., and S. cerevisiae. The precursor of pimeloyl-CoA is malonyl-CoA in E. coli and pimelic acid in Bacillus sp.

 
In eukaryotes, less is known about the genes involved in biotin formation than in bacteria. Most of what is known about biotin biosynthesis in eukaryotes is derived from work in Arabidopsis thaliana and S. cerevisiae (ZHANG et al. 1994; PHALIP et al. 1999; STREIT and ENTCHEVA 2003; PINON et al. 2005; WU et al. 2005). In S. cerevisiae, three of the genes involved in biotin synthesis and intermediate transport are found within a subtelomeric cluster located on chromosome XIV (BIO3, BIO4, and BIO5) (PHALIP et al. 1999). The final gene in the pathway, encoding biotin synthetase (BIO2), is found on chromosome VII (ZHANG et al. 1994) (Figure 1). A screen for genes of recent bacterial origin in S. cerevisiae and A. gossypii indicated, and subsequent phylogenetic analysis strongly supports, that two of these three known genes for biotin biosynthesis, BIO3 and BIO4, were acquired by horizontal gene transfer (HGT) from bacteria (HALL et al. 2005). WU et al. (2005) identified and characterized the gene immediately upstream of the known S. cerevisiae biotin biosynthesis pathway. This gene, BIO6, was found as a result of genomic sequencing of the sake-brewing, biotin prototrophic strain K7 of S. cerevisiae. This gene is a paralog of the horizontally transferred BIO3, but was shown to have acquired a new function. On the basis of analysis of gene order, this duplication appears to have been independent of the genome duplication (WOLFE and SHIELDS 1997). BIO6 is necessary but not sufficient for biotin production indicating the presence of one or more additional genes. We have identified the open reading frame adjacent to BIO6 as the upstream step of biotin synthesis in S. cerevisiae. This gene, which is now named BIO1, is a paralog of the gene YJR154W and appears to have acquired the function of pimeloyl-CoA synthetase (PCAS) and as such is a functional homolog of the bioW gene in Bacillus sp. and the bioC–bioH complex of E. coli. In S288c, there are two pseudogene copies each of BIO1 and BIO6 located on chromosomes I and VIII given the systematic names YAR069W-A (BIO6) and YHR214W-F (BIO8) and YAR070W-A (BIO1) and YHR214W-G (BIO7).

Phylogenetic methods:

Accession numbers for all sequences used in this analysis can be found in supplemental Table 6 at http://www.genetics.org/supplemental/. Gene sequences and ribosomal small subunit ribosomal (SSU) DNA sequences used in this analysis were acquired from GenBank (BENSON et al. 1999). Ribosomal SSU sequences were aligned by primary structure using ClustalX (THOMPSON et al. 1997). Amino acid sequences for all biotin biosynthesis genes were aligned by primary structure using ClustalX. Ambiguously aligned regions were excluded from analysis. Estimates of phylogenetic relatedness among species were determined using neighbor-joining (NJ) (SAITOU and NEI 1987) analysis of SSU sequences. NJ trees were constructed in ClustalX using the IUB matrix. NJ trees were bootstrapped in ClustalX using 1000 replicates. Estimates of phylogenetic relatedness among biotin biosynthesis genes were determined using NJ analyses of protein sequences. NJ trees were constructed in ClustalX using the method of SAITOU and NEI (1987) and the Gonnet matrix (GONNET et al. 1992).

Yeast strains and plasmids:

The S. cerevisiae strains used in this study are listed in supplemental Table 7 at http://www.genetics.org/supplemental/. Strains SCCH015 (bio1::HygB), SCCH016 (bio3::HygB), and SCCH017 (bio6::HygB) were created by the PCR-mediated gene disruption technique, using the hygromycin B resistance cassette from plasmid pGA32 and the primer pairs BIO1KO1–BIO1KO2, BIO3KO1–BIO3KO2, and BIO6KO1–BIO6KO2, respectively (WACH et al. 1994; LORENZ et al. 1995). All gene disruptions were confirmed by PCR and sequencing by standard methods using primer pairs BIO6PFX–HYGB and HYGC–PHO11R for BIO1, BIO3A–HYGB and HYGC–BIO3D for BIO3, and BIO6A–HYGB and HYGC–BIO6D for BIO6 (MULLIS et al. 1986). Plasmids used in this study are listed in supplemental Table 8 at http://www.genetics.org/supplemental/. Primer sequences used in this study are listed in supplemental Table 9 at http://www.genetics.org/supplemental/.

To test whether BIO1 is necessary for biotin biosynthesis in S. cerevisiae, BIO1, BIO6, and both genes together were expressed in S288c on derivatives of the 2 µ plasmid YEplac195 containing the BIO1, BIO6, and BIO6–BIO1 (with intergenic sequence) genes, respectively. BIO1, BIO6, and BIO6–BIO1 were PCR amplified from A364a genomic DNA using primers BIO1PF–BIO1PR, BIO6PF–BIO6PR, and BIO6PF–BIO1PR, respectively. PCR products were inserted by homologous plasmid gap repair in vivo (LORENZ et al. 1995). Plasmid PPH001 was digested with EcoRI. Linearized plasmid and PCR products were used to transform S. cerevisiae strain BY4743. Cells were selected on minimal media lacking uracil (BURKE et al. 2000). Plasmid inserts were confirmed by PCR and sequencing using primers PTHQOUT1, PTHQOUT2, PTHQOUT3, PTHQIN1, PTHQIN2, and PTHQIN3.

To determine if YJR154W can complement a BIO1 deletion, BIO1 and YJR154W were PCR amplified and cloned into the 2 µ plasmid vector (p427-TEF) under the control of a TEF1 promoter and introduced into the bio1 (SCCH015) deletion strain. BIO1 and YJR154W were PCR amplified from A364a genomic DNA using primer pairs BIO1PF427–BIO1PR427 and YJR154WPF427–YJR154WPR427, respectively. PCR products were inserted by homologous plasmid gap repair in vivo (LORENZ et al. 1995). Plasmid p427-TEF was digested with EcoRI. Linearized plasmid and PCR products were used to transform S. cerevisiae strain SCCH015. Cells were selected on yeast peptone dextrose (YPD) with G418. PCR and sequencing using primers TEFfw, CYCtREV, and CYCtREV2 confirmed plasmid inserts.

To determine if BIO3, BIO6, and the BIO3 homolog from K. lactis have overlapping function, BIO3, BIO6, and the BIO3 homolog from K. lactis were PCR amplified and cloned into the 2 µ plasmid vector (p427-TEF) under the control of a TEF1 promoter and introduced into the bio3 (SCCH016) and bio6 (SCCH017) deletion strains. BIO3, BIO6, and the BIO3 homolog from K. lactis were PCR amplified from S. cerevisiae strain A364a and K. lactis strain CBS 683 genomic DNA using primer pairs BIO3PF427–BIO3PR427, BIO6PF427–BIO6PR427, and KLACBIOPF427–KLACBIOPR427, respectively. PCR products were inserted by homologous plasmid gap repair in vivo (LORENZ et al. 1995). Plasmid p427-TEF was digested with EcoRI. Linearized plasmid and PCR products were used to transform S. cerevisiae strains SCCH016 and SCCH017. Cells were selected on YPD with G418. PCR and sequencing using primers TEFfw, CYCtREV, and CYCtREV2 confirmed plasmid inserts.

Media and growth conditions:

For general cultivation purposes all strains were grown on YPD broth and agar (BURKE et al. 2000). Selection for plasmids PPH001, PCH001, PCH002, and PCH003 was performed on synthetic complete minus uracil media (SC –Ura) (BURKE et al. 2000). Selection for plasmids p427-TEF, PCH004, PCH005, PCH006, PCH007, and PCH008 was performed on YPD media with G418 (BURKE et al. 2000). Tests for biotin prototrophy were performed on biotin-free media composed of ammonium sulfate (15 mM), monopotassium phosphate (6.6 mM), dipotassium phosphate (0.5 mM), sodium chloride (1.7 mM), calcium chloride (0.7 mM), magnesium chloride (2 mM), boric acid (0.5 µg/ml), copper chloride (0.04 µg/ml), potassium iodide (0.1 µg/ml), zinc chloride (0.19 µg/ml), calcium pantothenate (2 µg/ml), thiamine (2 µg/ml), pyridoxine (2 µg/ml), inositol (20 µg/ml), and glucose (2%). It was found that agar contained too much contaminating biotin to prevent background growth of auxotrophic strains of S. cerevisiae. When plates were required, agarose was used as a gelling agent to a final concentration of 2%. For growth of strain A364a this media was supplemented with adenine sulfate (20 mg/liter), histidine (20 mg/liter), lysine (30 mg/liter), and tyrosine (30 mg/liter). We found it necessary to titrate the level of biotin by transferring from two or more plates of media lacking biotin to eliminate background growth. For media containing biotin, biotin was added to a final concentration of 2 µg/liter.

To identify the remaining gene(s) necessary for biotin biosynthesis, crosses were carried out between the biotin producing strains YJM627 and A364a and the biotin auxotroph S288c. Thirty tetrads from the YJM627–S288c cross yielded a complex segregation pattern of 4 4:0 (+, –), 17 3:1 (+, –), and 9 2:2 (+, –) for biotin production. This pattern is consistent with two redundant and unlinked loci. Results from 25 tetrads of the A364a–S288c cross showed 2:2 segregation in all cases. S288c lacks at least two biotin biosynthesis genes, one of which is BIO6. The observed segregation pattern suggested a single-gene trait, and indicated that the missing gene(s) was closely linked to BIO6. Analysis of the genome sequences of the Saccharomyces sensu stricto species as well as compiled data from the S. cerevisiae genome projects ongoing at the Sanger Centre (http://www.sanger.ac.uk/Teams/Team71/durbin/sgrp/) allowed construction of a gene map of the region surrounding BIO6 (Figure 1). Several genes neighbor BIO6, including members of the multigene families FLO1 (flocculation), PHO11 (acid phosphatase), and IMD1 (inosine monophosphate dehydrogenase), as well as genes of unknown function. Examination of the map suggested that the most likely candidate for "BIO1" was the gene immediately flanking BIO6. This gene is a paralog copy of YJR154W, a gene of unknown function but with sequence similarity to genes that bind phytanoyl-CoA, a molecule structurally similar to pimeloyl-CoA (Figure 2 and supplemental Figure 1 at http://www.genetics.org/supplemental/), the likely product of the missing biotin biosynthesis gene. PCR and sequencing confirmed that BIO6 is flanked by this putative BIO1 in S. cerevisiae strains A364a and YJM627.

View larger version (39K): In this window In a new window Download PPT slide   FIGURE 2.— Phylogeny of BIO1 from S. cerevisiae and related genes and pimeloyl-CoA synthetase genes from prokaryotes (left) and species phylogeny based on SSU rDNA (right). BIO1 (clade a) is a duplicate copy of YJR154W. The pimeloyl-CoA synthetase activity of the genes in species in boldface type has been experimentally verified (STREIT and ENTCHEVA 2003). Pimeloyl-CoA synthetase activity appears to be highly divergent among the prokaryotes with some bacteria requiring the two-gene bioC–bioH complex (clades b and e) and others requiring a single gene such as bioW or bioZ (clades c and d). Trees were constructed by NJ in ClustalX. Numbers indicate bootstrap support for each node of 1000 replicates. Taxa marked with asterisks indicate putative phytanoyl-CoA dioxygenases.

 
To test whether BIO1 is necessary for biotin biosynthesis in S. cerevisiae, BIO1 and BIO6 were PCR amplified and cloned into a plasmid vector (YEplac195 derived), both individually under the control of an ADH1 promoter and together on the same plasmid with BIO6 under the control of an ADH1 promoter and BIO1 under the control of its native promoter. S288c strains with both BIO1 and BIO6, but not with either individually or with the vector control, became biotin prototrophs (Figure 3). Growth on 5-FOA to select for loss of the plasmid resulted in loss of biotin prototrophy (data not shown). These results indicate that BIO1 and BIO6 are together necessary and sufficient for the conversion of S288c into a biotin prototroph and suggests that BIO1 is the likely pimeloyl-CoA synthetase of S. cerevisiae.

View larger version (66K): In this window In a new window Download PPT slide   FIGURE 3.— Growth of S. cerevisiae strains on media containing 2 µg/liter biotin (A) and lacking biotin (B). Cells were grown overnight in YPD, transferred to liquid minimal media lacking biotin, grown overnight, transferred a second time to liquid minimal media lacking biotin, and then spotted onto plates. S288c is a biotin auxotroph. YJM627 is a biotin prototroph. +vector indicates vector alone (plasmid PPH001), +BIO1 includes BIO1 gene (plasmid PCH001), +BIO6 includes BIO6 gene (plasmid PCH002), and +BIO1BIO6 includes both genes on a single plasmid (plasmid PCH003).

 
To determine if BIO3 and BIO6 have overlapping function, and if YJR154W can complement a BIO1 deletion, we constructed deletions of BIO1, BIO3, and BIO6 in the biotin prototrophic strain A364a. BIO1, BIO3, and BIO6 were disrupted by the replacement of the open reading frame with the hphMX4 (GOLDSTEIN and MCCUSKER 1999) cassette by homologous recombination. All three deletions were confirmed by PCR and sequencing. All three deletions resulted in a strain auxotrophic for biotin.

BIO1 and YJR154W were PCR amplified and cloned into a plasmid vector (p427-TEF) under the control of a TEF1 promoter and introduced into the bio1 deletion strain. The plasmid expressing BIO1 was able to complement the bio1 strain, while the plasmid expressing YJR154W was not able to complement the bio1 strain (Figure 4). Similarly, BIO3, BIO6, and the BIO3/BIO6 homolog from K. lactis were PCR amplified and cloned into a plasmid vector (p427-TEF) under the control of a TEF1 promoter and introduced into the bio3 and bio6 deletion strains. The plasmids expressing BIO3 from S. cerevisiae and K. lactis were able to complement the bio3 strain, while the plasmid expressing BIO6 did not complement (Figure 5). In the bio6 strain, the plasmids expressing BIO3 from S. cerevisiae and K. lactis were unable to complement the bio6 strain, while the plasmid expressing BIO6 was able to complement (Figure 6), confirming the results reported by WU et al. (2005).

View larger version (55K): In this window In a new window Download PPT slide   FIGURE 4.— Growth of S. cerevisiae strains on media containing 2 µg/liter biotin (A) and lacking biotin (B). Cells were grown overnight in YPD, transferred to liquid minimal media lacking biotin, grown overnight, transferred a second time to liquid minimal media lacking biotin, and then spotted onto plates. S288c is a biotin auxotroph. A364a is a biotin prototroph. All plasmids were transformed into strain SCCH015 (A364a derived, bio1::HygB). +vector indicates vector alone (plasmid p427-TEF), +BIO1 includes BIO1 gene (plasmid PCH004), and +YJR154W includes the YJR154W gene (plasmid PCH005).

 

View larger version (48K): In this window In a new window Download PPT slide   FIGURE 5.— Growth of S. cerevisiae strains on media containing 2 µg/liter biotin (A) and lacking biotin (B). Cells were grown overnight in YPD, transferred to liquid minimal media lacking biotin, grown overnight, transferred a second time to liquid minimal media lacking biotin, and then spotted onto plates. S288c is a biotin auxotroph. A364a is a biotin prototroph. All plasmids were transformed into strain SCCH016 (A364a derived, bio3::HygB). +vector indicates vector alone (p427-TEF), +BIO3 and +BIO3 K. lactis includes the BIO3 gene from S. cerevisiae and K. lactis, respectively (plasmids PCH006 and PCH007), and +BIO6 includes the BIO6 gene (plasmid PCH008).

 

View larger version (47K): In this window In a new window Download PPT slide   FIGURE 6.— Growth of S. cerevisiae strains on media containing 2 µg/liter biotin (A) and lacking biotin (B). Cells were grown overnight in YPD, transferred to liquid minimal media lacking biotin, grown overnight, transferred a second time to liquid minimal media lacking biotin, and then spotted onto plates. S288c is a biotin auxotroph. A364a is a biotin prototroph. All plasmids were transformed into strain SCCH017 (A364a derived, bio6::HygB). +vector indicates vector alone (p427-TEF), +BIO3 and +BIO3 K. lactis includes the BIO3 gene from S. cerevisiae and K. lactis, respectively (plasmids PCH006 and PCH007), and +BIO6 includes the BIO6 gene (plasmid PCH008).

 

A controversy arose following the seminal work of PASTEUR (1860) as to the nutrient requirements of yeast. While some supported the observation of Pasteur that fermentation could be obtained with a media of mineral salts in the form of yeast ash, sucrose, ammonium tartrate, and a pinhead-sized inoculum of levure (DUCLAUX 1864), others reported that this mixture was insufficient to allow growth (VON LEIBIG 1869). Extensive work by WILDIERS (1901) suggested the existence of an essential micronutrient that is water soluble, insoluble in alcohol, absent from yeast ash, dialyzable, and while present in peptone and beer wort, absent from urea, asparagine, and several other nitrogenous compounds. He named this compound "bios." Efforts to purify this compound failed, and controversy continued over the nature of this compound and even its existence (TANNER 1924). Work by FARRIES and BELL (1930) showed that A. gossypii and two related species are unable to grow on asparagine as a sole nitrogen source, while some other fungi including Fusarium fructigenum were capable of growth. They found that a growth factor was necessary for growth of A. gossypii, which they furthermore fractionated and extensively analyzed. Work by BUSTON and PRAMANIK (1931a,b) showed that this factor necessary for growth of A. gossypii contained two components, one being inositol, and the other being identified several years later by purification, crystallization, and compositional determination as biotin (KOGL and TONNIS 1936).

In retrospect, much of the confusion over the existence and nature of bios may have stemmed from variability of the organisms used—we now know that some S. cerevisiae strains are biotin auxotrophs and that a few S. cerevisiae strains and many other fungal species are prototrophic (SHCHELOKOVA and VOROB'EVA 1982) (Figure 7 and Table 1). Furthermore, many S. cerevisiae autotrophic strains can utilize not only biotin, but also certain intermediates in the biotin biosynthetic pathway (PHALIP et al. 1999), and even a dilution of 4 x 10^–11 of crystallized biotin is sufficient to allow yeast growth (KOGL et al. 1937). It is possible that either the strain studied by Pasteur was a biotin prototroph or that his pinhead-sized inoculum transferred sufficient biotin to enable growth. It is also clear that A. gossypii was a fortuitous subject for the analysis of biotin by Farries and Bell. The species is not only vigorously growing and a biotin auxotroph, but, as genome sequencing has revealed (DIETRICH et al. 2004), is lacking the biotin synthetic genes BIO2, BIO3, BIO4, and BIO5 found in biotin auxotrophic strains of S. cerevisiae and is thus unable to convert biotin precursors to biotin.

View larger version (155K): In this window In a new window Download PPT slide   FIGURE 7.— Variability of biotin biosynthesis in hemiascomycete fungi. Strains of Candida albicans (2), K. lactis (2), S. cerevisiae (37), S. paradoxus (35), and S. kluyveri (2) were replica plated from minimal media onto media lacking biotin. Some strains show no growth, some weak growth, and others robust growth. This experiment was repeated three times. Growth indicated is an average for all three experiments. It is likely that growth differences are due to copy number differences in BIO6 and BIO1 (WU et al. 2005). Strains used and their relative growth on minimal media lacking biotin are summarized in Table 1.

 

View this table: In this window In a new window   TABLE 1 Strains used in Figure 8 and their relative growth on minimal media lacking biotin

 

View larger version (111K): In this window In a new window Download PPT slide   FIGURE 8.— Phylogeny of KAPA synthetase (bioF and BIO6) and DAPA synthetase (bioA and BIO3) indicates horizontal gene transfer of BIO3 into fungi from proteo bacteria (A) and gain of KAPA synthetase function by duplication of BIO3 to form BIO6 (B). Gene phylogeny KAPA synthetase is shown in a and DAPA synthetase in b. Species phylogeny was determined by ribosomal small subunit (SSU). Both trees were determined by distance methods. The numbers indicate bootstrap support for each node of 1000 replicates.

 
The identification of biotin, its role as a cofactor, and the resolution of its structure had a protracted history. The biotin synthesis pathway in S. cerevisiae has proven difficult to study largely because the evolution of the biotin biosynthesis pathway in S. cerevisiae has been complex. As we discuss in more detail below, the biotin biosynthesis pathway in S. cerevisiae has evolved by mechanisms of gene loss and gene gain by both HGT and gene duplication, making the biotin synthesis pathway an interesting model of gene pathway evolution. Also, unusually for S. cerevisiae, five of the six genes required for biotin synthesis, including all of the horizontally derived and duplicated genes, are localized to two subtelomeric clusters.

Gene clustering:

The BIO1 and BIO6 genes are adjacent, as are the BIO3, BIO4, and BIO5 genes in S. cerevisiae. We have examined the extent of clustering of genes of related function and found that clusters of genes of related function are unusual, and generally, there is very little clustering of genes of related function in S. cerevisiae. For most pathways, genes are scattered across the genome. With the exception of five specific categories of genes, discussed below, we identified only 14 pairs of adjacent, functionally related genes. This finding contradicts several published claims. COHEN et al. (2000, p. 184) reported that "of the 2081 adjacent pairs examined, 387 fell into the same functional category, significantly more than would be expected by chance (P = 10^–8)." LEE and SONNHAMMER (2003, p. 875) reported on the basis of the KEGG pathways (KANEHISA et al. 2006) that "virtually all pathways in S. cerevisiae showed significant clustering." TEICHMANN and VEITIA (2004, p. 2124) reported that "a significant fraction of genes encoding subunits of stable complexes are close to each other on yeast chromosomes."

Analysis of adjacent genes in S. cerevisiae using the KEGG pathways (KANEHISA et al. 2006), GO annotation (ASHBURNER et al. 2000), gene names, and Saccharomyces Genome Database (SGD) descriptions (CHERRY et al. 1997) identifies five categories of genes that are clustered by function. The genes involved in biotin synthesis are an example of a "dispensable pathway," a pathway that is found in some strains and species but not in others within a phylogenetic clade. In S. cerevisiae this category includes specific carbon source utilization clusters, aryl-sulfate utilization, siderophore-bound iron utilization, utilization of S-adenosylmethionine, and the utilization of allantoin as a nitrogen source (WONG and WOLFE 2005). This category furthermore includes arsenic resistance and vitamin B1/B6 metabolism (Table 2). The limited number of dispensable pathway clusters in S. cerevisiae is likely a reflection of the absence of secondary metabolism. Dispensable pathway clusters have been widely reported in other fungi including indole-diterpene biosynthesis in Neotyphodium lolii (YOUNG et al. 2006) and Aspergillus flavus (ZHANG et al. 2004), the penicillin biosynthetic gene cluster in Penicillium chrysogenum (SMITH et al.1990; ABE et al. 2002), P. nalgiovense (LAICH et al. 1999), P. griseofulvum, and P. verrucosum (LAICH et al. 2002), the aflatoxin pathway in A. nidulans (KELLER and ADAMS 1995) and A. parasiticus (TRAIL et al. 1995), ergot synthesis in Claviceps purpurea (TUDZYNSKI et al. 1999), and genes involved in plant pathogenesis in Ustilago maydis (KAMPER et al. 2006).

View this table: In this window In a new window   TABLE 2 Dispensable pathway genes that are clustered in S. cerevisiae

 
The additional four categories of clustered genes as well as two categories of artifacts that have possibly been confused with functionally related gene clusters in previous analyses, are summarized in Tables 1–3, and in supplemental Table 4 at http://www.genetics.org/supplemental/. Beyond these five categories of clustered genes, only 14 pairs of adjacent genes of related function were identified (Table 3). Of these 14 pairs only 8 pairs are also adjacent genes in A. gossypii.

View this table: In this window In a new window   TABLE 3 Functionally related pairs of adjacent genes in S. cerevisae

 
The clustering of the biotin biosynthesis genes is consistent with the observation that these genes are not ubiquitous in the hemiascomycetes nor are they consistently found in the species S. cerevisiae. It is unclear whether the clustering of genes in dispensable pathways is a result of selection for more efficient transmission of the trait to meiotic progeny, or of selection against toxic intermediates produced by incomplete pathways in meiotic progeny, or some other explanation.

Horizontal gene transfer:

A comparison of S. cerevisiae and A. gossypii (HALL et al. 2005) identified 10 genes horizontally acquired from prokaryotes by S. cerevisiae since the divergence of these species, including BIO3 and BIO4. We here report an updated analysis of horizontally acquired genes in S. cerevisiae including an additional 3 genes (supplemental Table 5 and supplemental Figures 3, 4, and 5 at http://www.genetics.org/supplemental/).

In our initial analysis of HGT in S. cerevisiae and A. gossypii, the majority of the genes of recent bacterial origin in S. cerevisiae appeared to be enzymes involved in the scavenging of nutrients such as the horizontally acquired alkyl-aryl sulfatase BDS1. This observation fits well with our understanding of HGT in bacteria in which HGT appears to be a mechanism for the acquisition of novel metabolic characteristics. These genes appear to be single-gain events that provide an immediate selective advantage. Uniquely, the two genes of the biotin biosynthesis pathway identified in our screen as likely recent horizontal acquisitions, BIO3 and BIO4, encode two adjacent steps of the same biosynthetic pathway. Phylogenetic analysis strongly supported the view that these two genes were of recent bacterial origin (Figures 8 and 9). Phylogenetic analysis also indicated that these two genes were not acquired simultaneously as part of an "operon transfer" but individually from different prokaryotic donors. BIO3 appears to come from gamma-proteo bacteria and BIO4 from alpha-proteo bacteria. Further evidence for HGT of BIO3 and BIO4 is that in the eukaryotic biotin biosynthesis pathway, the activities of dethiobiotin synthetase (DTBS) and DAPA synthetase (DS) appear to be performed by a single protein encoded by a single gene based on sequence homology. Species with eukaryotic type DTBS and DS show this dual function enzyme structure. In prokaryotes and in some hemiascomycetes, these two activities are encoded on separate genes (supplemental Figure 6 at http://www.genetics.org/supplemental/).

View larger version (17K): In this window In a new window Download PPT slide   FIGURE 9.— Phylogeny of dethiobiotin synthetase (bioD and BIO4) indicates horizontal gene transfer of BIO4 into fungi from bacteria. Gene phylogeny of dethiobiotin synthetase (left) and species phylogeny (right) were determined by ribosomal small subunit (SSU). Both trees were determined by distance methods. The numbers indicate bootstrap support for each node of 1000 replicates.

 

Reconstruction of the biotin synthesis pathway:

Identification of S. cerevisiae genes in amino acid and nucleotide synthesis pathways dates to the 1950s, yet the first gene identified in biotin biosynthesis was BIO2 in 1995 (ZHANG et al. 1994). With the identification of BIO3, BIO4, BIO5 (PHALIP et al. 1999), BIO6 (WU et al. 2005), and BIO1 (this report), it now appears that the complete set of genes for biotin biosynthesis in S. cerevisiae has been identified.

The identification of BIO6 and now BIO1 allows us to reconstruct the evolutionary history of the entire pathway. This history is detailed in Figure 10. All of the known genes for biotin biosynthesis form distinct eukaryotic and prokaryotic clades (Figures 2, 8, 9, 10 and supplemental Figure 2 at http://www.genetics.org/supplemental/). Construction of gene phylogenies for each step of biotin biosynthesis reveals that eukaryotic-type biotin biosynthesis genes are found among plants and most fungi, but disappear after the divergence of Yarrowia lipolytica from the rest of the hemiascomycetes except for the final step biotin synthetase (BS). Other than Y. lipolytica, most hemiascomycete fungi possess prokaryotic-type DTBS and DS genes. The most likely explanation for this observation is that the eukaryotic biotin biosynthetic pathway was lost in the ancestor of Candida/Debaryomyces/Kluyveromyces/Saccharomyces after the divergence with the ancestor of Y. lipolytica except for the final step, BS (Figure 10 and supplemental Figure 2 at http://www.genetics.org/supplemental/). The next two steps, DTBS and DS, were then reacquired by HGT from prokaryotes. It is difficult to precisely determine at which point in the evolution of the hemiascomycetes the loss of the pathway occurred or when prokaryotic genes were acquired. This uncertainty is due to limited genomic sampling of more basal hemiascomycetes.

View larger version (13K): In this window In a new window Download PPT slide   FIGURE 10.— Loss of the eukaryotic biotin biosynthesis pathway in the hemiascomycete lineage and reconstruction by horizontal gene transfer and gene duplication. Phylogram represents the evolutionary relationship between fungi. Rows indicate enzyme type. E indicates eukaryotic-type enzyme and P indicates prokaryotic-type enzyme. D indicates genes formed by duplication and neofunctionalization. Question marks indicate that the gene responsible for this step is unknown. Columns indicate enzyme function. PCAS indicates pimeloyl-CoA synthetase, KAPAS indicates KAPA synthetase, DAPAS indicates DAPA synthetase, DTBS indicates dethiobiotin synthetase, and BS indicates biotin synthetase. While S. kluyveri is a biotin prototroph (Figure 8), the genome of this organism does not appear to encode either a eukaryotic or a prokaryotic type PCAS or KAPAS.

 
The partial construction of the biotin biosynthesis pathway accomplished by the acquisition of bacterial genes left the pathway still incomplete, yet this partial pathway appears to have been maintained (Figure 10). With the appearance of the Saccharomyces sensu stricto complex, reconstruction of the biotin biosynthesis pathways was completed with the appearance of BIO6, a KAPA synthetase (KS), and BIO1, a putative PCAS. Phylogenetic construction indicates that these genes are the result of duplication of BIO3 and YJR154W, respectively. BIO3 does not have the function of BIO6 (WU et al. 2005; this study), nor does YJR154W have the function of BIO1 (this study). These two genes represent cases of gene duplication with subsequent neofunctionalization.

One of the most important largely unresolved issues in evolutionary biology concerns the genetic origins of morphological and biochemical novelties. The most common source of new genetic material in eukaryotes appears to be gene duplication. Gene duplication refers to the production of two duplicate loci, through unequal crossing over, tandem duplication, or other illegitimate chromosomal rearrangements. Gene duplication was originally discussed in the work of MULLER (1936), but its full importance in the evolutionary process was not generally realized until Ohno's seminal book >30 years later (OHNO 1970). Previously, models have predicted that it is rarer for a duplicated gene to reach a stable frequency in a population than to be silenced through mutation (made into a nonfunctional "pseudogene") (HALDANE 1933; OHNO 1970). Some insight into the extent of gene loss after duplication can be gained by examining the whole-genome duplication in the lineage of S. cerevisiae. In that case it has been shown that 92% of the duplicated genes were lost within 100 million years after genome duplication (SEOIGHE and WOLFE 1998). It seems more likely for a gene to be lost after duplication than to acquire a new function. The reconstructed biotin biosynthesis pathway observed in S. cerevisiae acquired two genes by this process. It is also interesting to note that BIO5, although not part of the core biosynthetic pathway, also appears to have formed as the result of gene duplication and possible neofunctionalization (supplemental Figure 7 at http://www.genetics.org/supplemental/).

In this work we report the identification of BIO1 as the first step of the known biotin biosynthetic pathway in S. cerevisiae. We also examine the remarkable evolutionary history of biotin biosynthesis in S. cerevisiae and related fungi. We conclude that the biotin biosynthetic pathway represents an example of gene pathway evolution. In this lineage, biotin biosynthesis has been nearly lost and then rebuilt by two rare genetic phenomena: horizontal gene transfer and gene neofunctionalization after duplication. More work is needed to determine whether pimelic acid, malonyl-CoA, or another molecule functions as the first intermediate in biotin synthesis in S. cerevisiae.

Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under accession no. EF567080.

ABE, Y., T. SUZUKI, C. ONO, K. IWAMOTO, M. HOSOBUCHI et al., 2002 Molecular cloning and characterization of an ML-236B (compactin) biosynthetic gene cluster in Penicillium citrinum. Mol. Genet Genomics 267: 636–646.[CrossRef][Medline]

ASHBURNER, M., C. BALL, J. A. BLAKE, D. BOTSTEIN, H. BUTLER et al., 2000 Gene ontology: tool for the unification of biology. Nat. Genet. 25: 25–29.[CrossRef][Medline]

BENSON, D. A., M. S. BOGUSKI, D. J. LIPMAN, J. OSTELL, B. F. OUELLETTE et al., 1999 GenBank. Nucleic Acids Res. 27: 12–17.[Abstract/Free Full Text]

BOWER, S., J. PERKINS, R. R. YOCUM, P. SERROR, A. SOROKIN et al., 1995 Cloning and characterization of the Bacillus subtilis birA gene encoding a repressor of the biotin operon. J. Bacteriol. 177: 2572–2575.[Abstract/Free Full Text]

BRACHMANN, C. B., A. DAVIES, G. J. COST, E. CAPUTO, J. LI et al., 1998 Designer deletion strains derived from Saccharomyces cerevisiae S288c: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14: 115–132.[CrossRef][Medline]

BURKE, D. T., D. DAWSON and T. STEARNS, 2000 Methods in Yeast Genetics: A Cold Spring Harbor Laboratory Course Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

BUSTON, H., and B. N. PRAMANIK, 1931a The accessory factor necessary for the growth of Nematospora gossypii: part I. Biochem. J. 25: 1656–1670.[Medline]

BUSTON, H., and B. N. PRAMANIK, 1931b The accessory factor necessary for the growth of Nematospora gossypii: part II. Biochem. J. 25: 1671–1673.[Medline]

CHERRY, J. M., C. BALL, S. WENG, G. JUVIK, R. SCHMIDT et al., 1997 Genetic and physical maps of Saccharomyces cerevisiae. Nature 387: 67–73.[CrossRef]

COHEN, B. A., R. D. MITRA, J. D. HUGHES and G. M. CHURCH, 2000 A computational analysis of whole-genome expression data reveals chromosomal domains of gene expression. Nat. Genet. 26: 183–186.[CrossRef][Medline]

DIETRICH, F. S., S. VOEGELI, S. BRACHAT, A. LERCH, K. GATES et al., 2004 The Ashbya gossypii genome as a tool for mapping the ancient Saccharomyces cerevisiae genome. Science 304: 304–307.[Abstract/Free Full Text]

DUCLAUX, M., 1864 Sur la fermentation alcoolique. Comp. Rend. Acad. Sci. 58: 1114–1116.

FARRIES, E. H. M., and A. F. BELL, 1930 On the metabolism of Nematospora gossypii and related fungi, with special reference to the source of nitrogen. Ann. Bot. 44: 423–455.

GOLDSTEIN, A. L., and J. H. MCCUSKER, 1999 Three new dominant drug resistance cassettes for gene disruption in Saccharomyces cerevisiae. Yeast 15: 1541–1553.[CrossRef][Medline]

GONNET, G. H., M. A. COHEN and S. A. BENNER, 1992 Exhaustive matching of the entire protein sequence database. Science 256: 1443–1445.[Abstract/Free Full Text]

HALDANE, J., 1933 The part played by recurrent mutation by evolution. Am. Nat. 67: 5–9.[CrossRef]

HALL, C., S. BRACHAT and F. S. DIETRICH, 2005 Contribution of horizontal gene transfer to the evolution of Saccharomyces cerevisiae. Eukaryot. Cell 4: 1102–1115.[Abstract/Free Full Text]

HOJA, U., C. WELLEIN, E. GREINER and E. SCHWEIZER, 1998 Pleiotropic phenotype of acetyl-CoA-carboxylase-defective yeast cells–viability of a BPL1-amber mutation depending on its readthrough by normal tRNA(Gln)(CAG). Eur. J. Biochem. 254: 520–526.[Medline]

IFUKU, O., H. MIYAOKA, N. KOGA, J. KISHIMOTO, S. HAZE et al., 1994 Origin of carbon atoms of biotin. 13C-NMR studies on biotin biosynthesis in Escherichia coli. Eur. J. Biochem. 220: 585–591.[Medline]

KAMPER, J., R. KAHMANN, M. BOLKER, L. J. MA, T. BREFORT et al., 2006 Insights from the genome of the biotrophic fungal plant pathogen Ustilago maydis. Nature 444: 97–101.[CrossRef][Medline]

KANEHISA, M., S. GOTO, M. HATTORI, K. F. AOKI-KINOSHITA, M. ITOH et al., 2006 From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Res. 34: D354–D357.[Abstract/Free Full Text]

KELLER, N. P., and T. H. ADAMS, 1995 Analysis of a mycotoxin gene cluster in Aspergillus nidulans. SAAS Bull. Biochem. Biotechnol. 8: 14–21.[Medline]

KOGL, F., P. FILDES, A. LWOFF, B. C. J. G. KNIGHT and G. M. RICHARDSON, 1937 Discussion Meeting on Growth Factors. Proc. R. Soc. Lond. B Biol. Sci. 124: 1–13.

KOGL, F., and B. TONNIS, 1936 Uber das bios-problem. Darstellung von krystallisiertem biotin aus eigelb. Mitteilung uber pflanzliche wachstumsstoffe. Zeitschrift für Physiologie Chemie 241: 43–73.

LAICH, F., F. FIERRO, R. E. CARDOZA and J. F. MARTIN, 1999 Organization of the gene cluster for biosynthesis of penicillin in Penicillium nalgiovense and antibiotic production in cured dry sausages. Appl. Environ. Microbiol. 65: 1236–1240.[Abstract/Free Full Text]

LAICH, F., F. FIERRO and J. F. MARTIN, 2002 Production of penicillin by fungi growing on food products: identification of a complete penicillin gene cluster in Penicillium griseofulvum and a truncated cluster in Penicillium verrucosum. Appl. Environ. Microbiol. 68: 1211–1219.[Abstract/Free Full Text]

LEE, J. M., and E. L. SONNHAMMER, 2003 Genomic gene clustering analysis of pathways in eukaryotes. Genome Res. 13: 875–882.[Abstract/Free Full Text]

LORENZ, M. C., R. S. MUIR, E. LIM, J. MCELVER, S. C. WEBER et al., 1995 Gene disruption with PCR products in Saccharomyces cerevisiae. Gene 158: 113–117.[CrossRef][Medline]

MULLER, H., 1936 Bar duplication. Science 83: 528–530.[Free Full Text]

MULLIS, K., F. FALOONA, S. SCHARF, R. SAIKI, G. HORN et al., 1986 Specific enzymatic amplification of DNA in vitro: the polymerase chain reaction. Cold Spring Harb Symp. Quant. Biol. 51(Pt 1): 263–273.

OHNO, S., 1970 Evolution By Gene Duplication. Springer, Berlin.

OHSUGI, M., and Y. IMANISHI, 1985 Microbiological activity of biotin-vitamers. J. Nutr. Sci. Vitaminol. 31: 563–572.[Medline]

PASTEUR, M. L., 1860 Memoires sur la fermentation alcoolique. Ann. Chim. Phys. 3: 323–426.

PHALIP, V., I. KUHN, Y. LEMOINE and J. M. JELTSCH, 1999 Characterization of the biotin biosynthesis pathway in Saccharomyces cerevisiae and evidence for a cluster containing BIO5, a novel gene involved in vitamer uptake. Gene 232: 43–51.[CrossRef][Medline]

PINON, V., S. RAVANEL, R. DOUCE and C. ALBAN, 2005 Biotin synthesis in plants. The first committed step of the pathway is catalyzed by a cytosolic 7-keto-8-aminopelargonic acid synthase. Plant Physiol. 139: 1666–1676.[Abstract/Free Full Text]

ROGGENKAMP, R., S. NUMA and E. SCHWEIZER, 1980 Fatty acid-requiring mutant of Saccharomyces cerevisiae defective in acetyl-CoA carboxylase. Proc. Natl. Acad. Sci. USA 77: 1814–1817.[Abstract/Free Full Text]

SAITOU, N., and M. NEI, 1987 The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4: 406–425.[Abstract]

SEOIGHE, C., and K. H. WOLFE, 1998 Extent of genomic rearrangement after genome duplication in yeast. Proc. Natl. Acad. Sci. USA 95: 4447–4452.[Abstract/Free Full Text]

SHCHELOKOVA, E., and L. VOROB'EVA, 1982 Biotin formation by the fungus Rhizopus delemar. Prikl. Biokhim. Mikrobiol. 18: 630–635.[Medline]

SMITH, D. J., M. K. BURNHAM, J. EDWARDS, A. J. EARL and G. TURNER, 1990 Cloning and heterologous expression of the penicillin biosynthetic gene cluster from Penicillum chrysogenum. Biotechnology 8: 39–41.[CrossRef][Medline]

STOLZ, J., U. HOJA, S. MEIER, N. SAUER and E. SCHWEIZER, 1999 Identification of the plasma membrane H+-biotin symporter of Saccharomyces cerevisiae by rescue of a fatty acid-auxotrophic mutant. J. Biol. Chem. 274: 18741–18746.[Abstract/Free Full Text]

STREIT, W. R., and P. ENTCHEVA, 2003 Biotin in microbes, the genes involved in its biosynthesis, its biochemical role and perspectives for biotechnological production. Appl. Microbiol. Biotechnol. 61: 21–31.[CrossRef][Medline]

TANNER, F. W., 1924 The "BIOS" question. Chem. Rev. 1: 397–472.[CrossRef]

TEICHMANN, S. A., and R. A. VEITIA, 2004 Genes encoding subunits of stable complexes are clustered on the yeast chromosomes: an interpretation from a dosage balance perspective. Genetics 167: 2121–2125.[Abstract/Free Full Text]

THOMPSON, J. D., T. J. GIBSON, F. PLEWNIAK, F. JEANMOUGIN and D. G. HIGGINS, 1997 The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25: 4876–4882.[Abstract/Free Full Text]

TRAIL, F., N. MAHANTI, M. RARICK, R. MEHIGH, S. H. LIANG et al., 1995 Physical and transcriptional map of an aflatoxin gene cluster in Aspergillus parasiticus and functional disruption of a gene involved early in the aflatoxin pathway. Appl. Environ. Microbiol. 61: 2665–2673.[Abstract]

TUDZYNSKI, P., K. HOLTER, T. CORREIA, C. ARNTZ, N. GRAMMEL et al., 1999 Evidence for an ergot alkaloid gene cluster in Claviceps purpurea. Mol. Gen. Genet. 261: 133–141.[CrossRef][Medline]

VON LEIBIG, J., 1869 Ueber die Gahrung and die Quelle der Muskelkraft. Annalen der Chemie und Pharmacie 153: 1–47.[CrossRef]

WACH, A., A. BRACHAT, R. POHLMANN and P. PHILIPPSEN, 1994 New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 10: 1793–1808.[CrossRef][Medline]

WILDIERS, E., 1901 Nouvelle substance indespensable au developpement de la levure. Le Cellule 18: 313–336.

WOLFE, K. H., and D. C. SHIELDS, 1997 Molecular evidence for an ancient duplication of the entire yeast genome. Nature 387: 708–713.[CrossRef][Medline]

WONG, S., and K. H. WOLFE, 2005 Birth of a metabolic gene cluster in yeast by adaptive gene relocation. Nat. Genet. 37: 777–782.[CrossRef][Medline]

WU, H., K. ITO and H. SHIMOI, 2005 Identification and characterization of a novel biotin biosynthesis gene in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 71: 6845–6855.[Abstract/Free Full Text]

YOUNG, C. A., S. FELITTI, K. SHIELDS, G. SPANGENBERG, R. D. JOHNSON et al., 2006 A complex gene cluster for indole-diterpene biosynthesis in the grass endophyte Neotyphodium lolii. Fungal Genet. Biol. 43: 679–693.[CrossRef][Medline]

ZHANG, S., B. J. MONAHAN, J. S. TKACZ and B. SCOTT, 2004 Indole-diterpene gene cluster from Aspergillus flavus. Appl. Environ. Microbiol. 70: 6875–6883.[Abstract/Free Full Text]

ZHANG, S., I. SANYAL, G. H. BULBOACA, A. RICH and D. H. FLINT, 1994 The gene for biotin synthase from Saccharomyces cerevisiae: cloning, sequencing, and complementation of Escherichia coli strains lacking biotin synthase. Arch. Biochem. Biophys. 309: 29–35.[CrossRef][Medline]

Communicating editor: A. P. MITCHELL

This article has been cited by other articles:

				R. Muralla, E. Chen, C. Sweeney, J. A. Gray, A. Dickerman, B. J. Nikolau, and D. Meinke A Bifunctional Locus (BIO3-BIO1) Required for Biotin Biosynthesis in Arabidopsis Plant Physiology, January 1, 2008; 146(1): 60 - 73. [Abstract] [Full Text] [PDF]

International Access Link

Online ISSN: 1943-2361  Print ISSN: 0016-6731
Copyright 2009 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu