OUP user menu

Why does Kluyveromyces lactis not grow under anaerobic conditions? Comparison of essential anaerobic genes of Saccharomyces cerevisiae with the Kluyveromyces lactis genome

I.S. Ishtar Snoek, H. Yde Steensma
DOI: http://dx.doi.org/10.1111/j.1567-1364.2005.00007.x 393-403 First published online: 1 May 2006


Although some yeast species, e.g. Saccharomyces cerevisiae, can grow under anaerobic conditions, Kluyveromyces lactis cannot. In a systematic study, we have determined which S. cerevisiae genes are required for growth without oxygen. This has been done by using the yeast deletion library. Both aerobically essential and nonessential genes have been tested for their necessity for anaerobic growth. Upon comparison of the K. lactis genome with the genes found to be anaerobically important in S. cerevisiae, which yielded 20 genes that are missing in K. lactis, we hypothesize that lack of import of sterols might be one of the more important reasons that K. lactis cannot grow in the absence of oxygen.

  • Kluyveromyces lactis
  • anaerobic essential genes
  • Saccharomyces cerevisiae
  • yeast deletion strains


The yeast Kluyveromyces lactis is industrially interesting because it is able to grow on lactose as a sole carbon source (Breunig & Steensma et al., 2003). This sugar is one of the main components of whey, which is a waste product of the production of cheese. If the lactose in whey could be converted to ethanol, the costs and environmental strain of waste disposal in this industry could be greatly reduced. The respiro-fermentative nature of metabolism in K. lactis, however, limits the efficiency of this process. Anaerobic growth could lead to full fermentation and thus higher production of ethanol by this yeast. Attempts have been made to transfer the ability of K. lactis to utilize lactose as a carbon source to Saccharomyces cerevisiae, but so far no industrially applicable yeast strain has emerged from this approach (Rubio-Texeira et al., 1998).

Yeast species differ in the ability to grow under anaerobic conditions. Only a few species can grow as successfully under anaerobic as under aerobic conditions, as was demonstrated by Visser (1990). Molecular di-oxygen is needed as the terminal oxidator in the respiratory pathway, leading to the production of energy. Oxygen is also required in several biosynthetic pathways, such as those for heme, sterols, unsaturated fatty acids, pyrimidines and deoxyribonucleotides (Andreasen & Stier et al., 1953; Nagy et al., 1992; Chabes et al., 2000). Cells growing under anaerobic conditions obviously have found ways to circumvent the oxygen dependency of these pathways. Without oxygen, energy can be produced by switching to fermentation. Although K. lactis is able to ferment, it cannot grow under anaerobic conditions (Kiers et al., 1998). The problem may lie in the oxygen dependency of biosynthetic pathways. The different problems arising from the absence of oxygen will be discussed briefly here, in relation to what is known in other organisms, in particular S. cerevisiae.

The synthesis of heme is dependent on traces of molecular oxygen and there is no known way to eliminate this requirement. It has been suggested that in anaerobically growing cells, the heme released by degradation of respiratory cytochromes is recycled in the cytoplasm. In S. cerevisiae, Mdl1 is a putative mitochondrial heme carrier that is upregulated under anaerobic conditions. This protein may be responsible for the transport of heme from the mitochondrial matrix to the cytoplasm (Clarkson et al., 1991; Kwast et al., 2002). The dependency of the biosynthesis of heme on oxygen also implies that production of hemeoproteins, most of which are cytochromes, requires oxygen. There may be anaerobic alternatives for these proteins. One study in S. cerevisiae has shown that the hemeoproteins Erg11, Cyc7, Ole1 and Scs7 are all upregulated under anaerobic batch culture conditions (Kwast et al., 2002). However, only Scs7 was upregulated under anaerobic glucose-limited chemostat culture conditions (ter Linde et al., 1999). ERG11 and CYC7 are known to code for cytochrome P450 and cytochrome c, respectively. Ole1, by contrast, is a fatty-acid desaturase, required for mono-unsaturated fatty-acid synthesis (Stukey et al., 1990), whereas Scs7 is a desaturase/hydroxylase, required for the hydroxylation of very-long-chain fatty acids (VLCFAs) (Dunn et al., 1998). These proteins still need heme and thus oxygen. If the cells are growing, recycled heme cannot account for all heme requirement and cells should have alternative solutions to this problem.

A second pathway that requires oxygen is the biosynthesis of sterols. Under aerobic circumstances, sterols are produced in an oxygen-dependent way, through the activities of six Erg enzymes. Twelve molecules of molecular oxygen are needed for the synthesis of one molecule of ergosterol (Rosenfeld & Beauvoit et al., 2003). Under anaerobic conditions, the cells no longer synthesize sterols, but instead import them. This sterol uptake is essential under anaerobic conditions (Wilcox et al., 2002). Transfer depends on the cellular levels of ergosterol and oleate (Burke et al., 1997; Ness et al., 1998). The transport might be a result of the permeability of the membrane. SUT1, and perhaps also SUT2, has a regulatory effect on this permeability (Ness et al., 2001). The expression of SUT1 increased following a shift to anaerobic conditions. The transcription factor UPC2 is also involved in sterol uptake (Wilcox et al., 2002). Together, these transcription factors upregulate transcription of AUS1, PDR11 and DAN1, the products of which work in synergy to mediate sterol uptake (Wilcox et al., 2002; Alimardani et al., 2004). In another study, ARV1 was identified as being required for sterol uptake and distribution. Strains having a deletion in this gene were unable to grow anaerobically (Tinkelenberg et al., 2000).

Because the production of unsatured fatty acids (UFAs) is oxygen-dependent, the medium for growing cells anaerobically is usually supplemented with Tween 80, which is a source of oleate. The presence of this compound represses the transcription of OLE1, which encodes the Acyl-CoA desaturase, involved in the biosynthesis of palmitoleate and oleate. FAT1 may encode a transporter involved in oleate uptake, which is required for anaerobic growth (Faergeman et al., 1997). The mitochondrial protein Rml2 may also participate in the assimilation (Trotter et al., 1999).

Synthesis of pyrimidines is also oxygen-dependent. The fourth step in the process, the conversion of dihydroorotate to orotate, is catalyzed by dihydroorotate dehydrogenase (DHDODase), which is a respiratory-chain-dependent mitochondrial protein in most yeasts. However, S. cerevisiae, which is able to grow anaerobically, has a cytosolic DHDODase and this enzyme is not dependent on the functionality of the respiratory chain (Gojkovic et al., 2005). Indeed, transfer of the S. cerevisiae DHODase gene (encoded by URA1) into Pichia stipitis transformed this yeast into a facultative anaerobe (Shi & Jeffries et al., 1998).

Biosynthesis of deoxyribonucleotides is catalyzed by ribonucleotide reductases (RNRs) (Kolberg et al., 2004). These enzymes convert the ribonucleotides into their deoxyribonucleotide counterparts. There are three major classes of RNRs. Members of class I are dependent on the presence of oxygen, members of class III function in the absence of oxygen and members of class II can reduce ribonucleotides under both conditions. Until now, only class I RNRs have been found in yeast species. However, because the 3D structures of the three classes are quite similar, although the sequence homology is very low, it could be that a class II or III RNR is present in the yeasts that are able to grow without oxygen.

Nicotinic acid is required for the synthesis of NAD+, and S. cerevisiae can synthesize it from tryptophan via the kynurenine pathway. The nicotinate moiety can also be recycled and be incorporated in NAD+ directly by the activity of nicotinate phosphoribosyl transferase (Npt1). Only the second pathway is oxygen-independent. Because there is no other way to synthesize NAD+, the NPT1 gene is essential under anaerobic conditions (Panozzo et al., 2002).

Under aerobic conditions, the reoxidation of NADH formed during glycolysis occurs through the respiratory chain, transferring the reducing equivalents to oxygen. This is not possible during anaerobiosis. Several ways of reoxidizing NADH are known in S. cerevisiae. The genes FRDS and OSM1 encode fumarate reductases, which irreversibly catalyze the reduction of fumarate to succinate, thereby reoxidizing NADH. FRDS1 (encoded by FRDS) is present in the cytosol and FRDS2 (encoded by OSM1) in the promitochondria, which lack an integrated electron transfer chain and a functional oxidative phosphorylation system and therefore are considered to be inactive for energy production. A mutant with a deletion in both the FRDS and the OSM1 genes is not able to grow under anaerobic conditions (Arikawa et al., 1998; Enomoto et al., 2002). Other ways to reoxidize excess NADH are through the actions of the Gpd2, which is a glycerol-3-phosphate dehydrogenase, and Adh3, which is a mitochondrial alcohol dehydrogenase. However, deletion of these genes only reduced the growth rate, but did not abolish growth under anaerobic conditions (Ansell et al., 1997; Bakker et al., 2000).

ADP/ATP carriers function in aerobic cells to exchange cytoplasmic ADP for intramitochondrially synthesized ATP. Under anaerobic conditions, the same proteins work in the opposite direction, exchanging ATP from glycolysis to the mitochondria. In S. cerevisiae, three genes encode for these transporters, AAC1, AAC2 and AAC3, all of which are transcribed in an oxygen-dependent manner (Sabova et al., 1993; Betina et al., 1995; Gavurnikova et al., 1996). Deletion of AAC2 and AAC3 was anaerobically lethal (Kolarov et al., 1990; Drgon et al., 1991).

In addition to the presence of genes essential for anaerobic growth, redirection of metabolism is required. For instance, owing to the lower yield of fermentation in comparison to that of respiration, a higher glycolytic flux and a higher uptake rate of sugars are necessary to maintain a high growth rate. Therefore, the proper regulatory mechanisms must also be present. In S. cerevisiae, several transcription factors are involved. Hap1 is a factor that has been implicated in the regulation of transcription in response to the availability of oxygen. The protein forms a homodimer in response to heme binding. This complex upregulates the transcription of aerobic genes. One of those genes is ROX1, which represses the transcription of anaerobic genes (Deckert et al., 1995). Both UPC2 and ECM22 are implicated in the induction of an anaerobic sterol import system (Crowley et al., 1998; Shianna et al., 2001). Other proteins that have been reported to influence transcription levels of anaerobic genes are Sut1, Ord1 and the Hap2/3/4/5 complex (Zitomer & Lowry et al., 1992; Lambert et al., 1994; Ness et al., 2001). In our laboratory, SPT3, SPT4, SAC3 and SNF7 have also been found in a search for anaerobic transcription factors (I.S.I. Snoek, unpublished data). The absence of one or more of these genes may also result in the inability to grow anaerobically.

To answer the question of why K. lactis cannot grow without oxygen, whereas other yeasts such as S. cerevisiae can (Fig. 1), we wished to determine whether S. cerevisiae has genes, important for anaerobic growth, that K. lactis lacks. We made use of the collection of S. cerevisiae gene-deletion mutants in strain BY4743 that was created by substituting each known ORF by a KanMX cassette (Giaever et al., 2002). We used the diploid parts of the collection. We tested each strain for its ability to grow anaerobically. This resulted in a list of anaerobically essential genes. In line with the definition by Giaever (2002), we have defined anaerobically essential genes as necessary for growth in YPD, supplemented with ergosterol and Tween 80. By comparing this list with the genome of K. lactis (http://cbi.labri.fr/Genolevures), we were able to identify several genes with little or no similarity in K. lactis. We discuss whether the absence of these genes may explain why K. lactis is not able to grow without oxygen.


Saccharomyces cerevisiae strain CEN.PK 113-7D and Kluyveromyces lactis strains CBS6315, CBS2360, CBS2359, CBS683, JBD100, PM6-7A and JA-6 grown under anaerobic and aerobic conditions. Four microlitres of 10-fold dilutions were spotted onto two MYplus plates. Plates were photographed after 4 days incubation, either aerobically or anaerobically, at 30°C.

Materials and methods


The strains used are listed in Table 1. The Saccharomyces cerevisiae mutant gene deletion collections 95401.H1 (homozygous diploids) and 95401.H4 (heterozygous diploids, essential genes only) were purchased from Research Genetics (Carlsbad, CA).

View this table:

Yeast strains used in this study

Kluyveromyces lactis CBS6315MatαCBS, Utrecht, The Netherlands
Kluyveromyces lactis CBS2360MatαCBS, Utrecht, The Netherlands
Kluyveromyces lactis CBS2359MatαCBS, Utrecht, The Netherlands
Kluyveromyces lactis CBS683CBS, Utrecht, The Netherlands
Kluyveromyces lactis JBD100MATα HO lac4-1 trp1 ara3-100Heus (1990)
Kluyveromyces lactis PM6-7AuraA1-1 adeT-600Wesolowski-Louvel (1992)
Kluyveromyces lactis JA-6MATα ade1-600 adeT-600 trp1-11 ura 3-12 KHT1 KHT2Ter Linde & Steensma (2002)
Saccharomyces cerevisiae CEN.PK 113-7DMatαP. Kötter (J.-W. Goethe Universität, Frankfurt, Germany)
Saccharomyces cerevisiae BY4743MATa/α his3Δ1/his3Δ1 leu2Δ0 /leu2Δ0 lys2Δ0/LYS2 MET15/met15Δ0 ura3Δ0 /ura3Δ0)Euroscarf, Frankfurt, Germany
Lipomyces starkeyi CBS1807CBS, Utrecht, The Netherlands


Yeast cells were grown in YPD (Difco peptone 2%, Difco yeast extract 1%, glucose 2%), MY (Zonneveld et al., 1986), or MYplus. MYplus is MY with 1% casamino acids, adenine, uracil and l-tryptophan at 30 μg mL−1 and 10 μg mL−1 ergosterol and 420 μg mL−1 Tween 80. For anaerobic growth in YPD, 10 μg mL−1 ergosterol and 420 μg mL−1 Tween 80 were added, giving YPDET. When necessary, 150 μg mL−1 G418 was added. Sporulation medium contained 0.1% Difco yeast extract, 1% potassium acetate and 0.05% glucose. Media were solidified by adding 1.5% agar (SpheroQ, Hispanagar, Burgos, Spain).

Anaerobic incubation

For anaerobic incubation of Petri dishes, the Anaerocult IS system (Merck, Darmstadt, Germany) was used. Anaerobicity was monitored both by an indicator strip (Anaerotest, Merck) and by using a Lipomyces starkeyi strain, which cannot grow under anaerobic conditions. Liquid cultures were shaken at 150 r.p.m. in an anaerobic cabinet (Bactron Anaerobic Chamber, Sheldon Inc., Vernon Hills, IL).

Anaerobic growth assay of Kluyveromyces lactis and Saccharomyces cerevisiae

Strains were shaken overnight in 2 mL YPD medium at 30°C. The next day, the strains were used to inoculate 10 mL fresh YPD medium to an absorbance at 655nm (A655) of 0.2. After shaking at 30°C for another 4 h, the cells were diluted in water to an A655 of 0.2 and 4 μL of a 10-fold dilution series in water were spotted onto two MYplus plates. One of the plates was incubated aerobically for 4 days at 30°C; the other was incubated anaerobically also for 4 days at 30°C.

Identification of anaerobically essential genes in Saccharomyces cerevisiae

The collection of homozygous and heterozygous deletion strains obtained from Research Genetics was used. This collection consists of mutants of the strain BY4743 in which each ORF has been replaced by a KanMX cassette as described by Giaever (2002).

The 95401.H1 version of the collection of homozygous deletion strains was grown overnight aerobically in 140 μL of YPD with G418 in flat-bottom 96-well plates (Greiner, Frickenhausen, Germany). About 1–2 μL of culture was transferred with a pin replicator (Nunc, San Diego, CA) to a new plate containing fresh YPDET medium with G418. The cultures were incubated at 30°C for 72 h. Duplicate plates were incubated anaerobically using Anaerocult IS (Merck) for the same period, also at 30°C. Absorbance was then measured at 655 nm in a microtiterplate reader (model 3550, BioRad, Hercules, CA).

The collection of BY4743-derived heterozygous diploid strains (95401.H4) with mutations in essential genes was used to inoculate 200 μL YPD. After overnight incubation at 30°C, a fresh microtiterplate with 200 μL YPD per well was inoculated using a 96-pin replicator. The next day, 2 μL of the cultures were spotted onto sporulation agar in a microtiterplate-sized Petridish (Nunc). After 3–5 days at 30°C, sporulation reached a maximum of only 1–10% for strains derived from BY4743. For other strains, this value was 70–90%. Plates stored at 4°C could be used for at least 1 month. For dissection, a small aliquot of the sporulated culture was resuspended in one drop of a lyticase solution [1 mg lyticase (Sigma, St Louis, MD) in 1 mL of water]. After 3–5 min at room temperature, the suspension was diluted 10-fold with water and used directly or kept on ice. For each strain, 4–6 asci were dissected using a Singer (Watchet, UK) MSM system dissection microscope on two YPDET plates, one of which was incubated aerobically, with the other incubated anaerobically, both at 30°C. Of the strains that did not segregate 2 : 2 for both anaerobic and aerobic growth, another 10 tetrads were dissected. The entire collection was screened twice in this way, starting from the original Genetic Research microtiter plates. The few discrepancies between the first and the second round were tested a third time.


Anaerobically essential genes

Although it is generally accepted that Kluyveromyces lactis is not able to grow under anaerobic conditions, data to support this notion are hard to find. We therefore tested several frequently used K. lactis strains for their ability to grow anaerobically. Figure 1 shows the results on mineral medium supplemented with Tween 80 and ergosterol, but similar results were obtained on rich medium (YPD) with the same supplements. Whereas the two Saccharomyces cerevisiae strains grew abundantly, all seven K. lactis strains only showed some residual growth, probably caused by the initially present oxygen, which would allow growth until essential components were exhausted. Similar effects are observed when S. cerevisiae is incubated anaerobically without Tween 80 or ergosterol. It thus appears that K. lactis, at least the seven strains tested, is not able to sustain growth in the absence of molecular oxygen.

The yield on glucose is much lower during fermentation than during respiration, therefore strains need a high fermentation capacity. Several K. lactis strains, including CBS2360, have the so-called Rag phenotype: they cannot grow on glucose in the presence of the respiration inhibitor antimycin A, owing to a mutation in the RAG1 gene encoding the only low-affinity glucose transporter in this strain (Goffrini et al., 1989, 1990). Obviously, the fermentation rate is too low to support growth. Several other strains, such as JA-6, have two tandemly arranged glucose transporter genes, KHT1 and KHT2, at the RAG1 locus and in these strains, fermentation is enhanced (Breunig et al., 2000). The lack of sufficient fermentation capacity may contribute but cannot be the only explanation for the inability of K. lactis to grow anaerobically, as there was no difference in anaerobic growth among the seven K. lactis strains, including CBS2360 and JA-6. Because S. cerevisiae can grow under anaerobiosis, other factors might be present in S. cerevisiae, which are lacking from K. lactis. As a first approach, we investigated which genes are important for anaerobic growth in S. cerevisiae and then determined the presence of these genes in K. lactis.

Around 1300 S. cerevisiae genes are essential for aerobic growth on rich medium, but it was unknown how many of these are also necessary for anaerobic growth. We therefore sporulated and dissected the 1166 heterozygous diploids with deletions in the essential genes (collection 95401.H4). This test showed that the aerobically essential genes indeed segregated 2 : 2 under aerobic conditions. Most of these genes were also needed for growth under anaerobic conditions. Only 33 genes were not required for anaerobic growth, giving four normal-sized colonies per tetrad, two of which did not grow when restreaked and incubated aerobically. In 32 strains anaerobic growth was retarded, with two normal and two small (<0.5 mm diameter) to very small (<100 cells per colony) colonies per tetrad, making the deleted genes in these strains necessary for optimal anaerobic growth. The results are listed in Table 2. As expected, genes involved in ergosterol synthesis are not necessary when this compound is present in the medium. Similarly, the finding of several mitochondrial genes is not surprising. However, for almost all other genes in the list, even those to which a function has been attributed, it is not clear why they are essential for aerobic but not for anaerobic growth.

View this table:

ORFs essential for aerobic growth

Essential for aerobic growth, but not for anaerobic growth
YGR082wTOM20Transport outer mitochondrial membrane
YGL055wOLE1Stearoyl-CoA desaturase, mitochondrial inheritance, ER
YGL018cJAC1Aerobic respiration, Iron sulfur cluster assembly, mitochondrion
YMR134wIron homeostasis
YDL120wYFH1Yeast frataxin homologue, Iron homeostasis, mitochondrion
YDR353wTRR1Thioredoxin reductase (NADPH) Regulation of redox homeostasis
YBR167cPOP7Ribonuclease P, mitochondrial RNA processing complex
YPL231wFAS23-oxoacyl (acyl carrier protein) reductase/synthetase
YBR192wRIM2Mitochondrial genome maintenance, Transporter
YGL001cERG26Ergosterol biosynthesis
YGR175cERG1Ergosterol biosynthesis
YHR072wERG7Ergosterol biosynthesis
YHR190wERG9Ergosterol biosynthesis
YLR100wERG27Ergosterol biosynthesis
YGR280cPXR1Possible telomerase regulator or RNA-binding protein
YIR008cPRI1Alpha DNA polymerase, DNA replication initiation
YIL118wRHO3Rho small monomeric GTPase, signal transduction
YBR061cTRM7t RNA methyl transferase
YDL212wSHR3Amino acid transport, ER
YEL034wHYP2Translation elongation factor, homologous to ANB1
YER008cSEC3Golgi to plasmamembrane transport
YDR427wRPN919 S proteasome regulatory particle
YER107cGLE2Nuclear pore organization and biogenesis
YKR038cKAE1Kinase associated endopeptidase
YMR239cRNT1Ribonuclease III
Essential for aerobic growth, but related with retarded anaerobic growth
YBL030cPET9/AAC2ATP/ADP antiporter, mitochondrial innermembrane
YGR029wERV1Sulhydryl oxidase, iron homeostasis, mitochondrion organization and biogenesis
YML091cRPM2Ribonuclease P, mitochondrial organization and biogenesis
YMR301cATM1Mitochondrial ABC transporter protein
YER043cSAH1Methionine metabolism
YMR113wFOL3Dihydrofolate synthase
YDR499wLCD1DNA damage checkpoint, telomere maintenance
YER146wLSM5mRNA splicing, snRP
YER159cBUR6Transcription co-repressor
YGL150cINO80ATPase, chromatin remodelling complex
YPR104cFHL1POL III transcription factor
YBL092wRPL32Ribosomal protein
YGL169wSUA5Translation initiation
YNL007cSIS1Chaperone, translational initiation
YER036cKRE30ABC transporter
YDR376wARH1Heme a biosynthesis, Iron homeostasis, mitochondrial inner membrane
YLR259cHSP60Heat shock protein, mitochondrial translocation
YKL192cACP1Fatty acid biosynthesis, cytosol
YNL103wMET4Transcription co-activator, methionine auxotroph
YHR005cGPA1Pheromone response in mating type
YPL020cULP1SUMO specific protease, G2/M transition
  • * Considered viable in the most recent version of SGD.

  • Gave aerobically 2+ : 2 very small colonies.

We next tested the homozygous deletion mutants that could grow aerobically in YPD for growth in YPDET in the absence of molecular oxygen. Although some residual growth to varying degrees was observed, the 23 strains listed in Table 3 consistently did not grow beyond the background.

View this table:

Genes essential for anaerobic growth and not essential for aerobic growth

Systematic nameGeneFunction
YAL026CDRS2Integral membrane Ca(2+)-ATPase
YPL254WHFI1Subunit of SAGA
YBR179CFZO1Mitochondrial integral membrane protein
YDR138WHPR1Subunit of THO/TREX
YDR364CCDC40Splicing factor
YOR209CNPT1Nicotinate phosphoribosyl transferase
YLR242CARV1Sterol metabolism/ transport
YDR173CARG82Transcription factor
YPL069CBTS1Terpenoid biosynthesis
YPR135WCTF4Chromatin-associated protein
YGL025CPGD1Subunit of Mediator
YGL084CGUP1Glycerol transporter
YNL236WSIN4Subunit of Mediator
YNL215WIES2Associates with INO80
YKR024CDBP7ATP-dependent RNA helicase
YDR477WSNF1Protein serine/threonine kinase
YNL284CMRPL10Protein synthesis
YOL148CSPT20Subunit of SAGA

It was expected that at least some of the genes reported in the literature to be of importance for anaerobic growth (see Introduction) would come up in this screen. Therefore, we took a more careful look at the results for the strains lacking these genes. The results are listed in Table 4.

View this table:

Growth of strains lacking the genes described to be important for anaerobic growth in literature as described in the introduction

Systematic nameGeneAerobic growthAnaerobic growth
YEL050CRML2+/−Not done

Of all the genes found in the literature to be connected to anaerobic metabolism, only two, NPT1 and ARV1, were found to be anaerobically essential. A possible explanation for this apparent discrepancy could be the redundancy of the genes in question (see Discussion). In contrast, we did find 21 genes to be important for anaerobic growth that were not previously implicated. Indeed, analysis of the list shows no logical reason for these genes to be anaerobically essential as the genes do not belong to any pathway or functional group linked to anaerobic growth.

Search for anaerobically important genes in the Kluyveromyces lactis genome

A search for the genes listed in Tables 3 and 4 with the genome of K. lactis resulted in identification of 20 genes for which a homologue could not be found. In this comparison we also included the regulatory genes that we identified in our group and that have an anaerobically upregulating activity. These are listed in Table 5.

View this table:

Genes that have a role in anaerobic growth but for which no homologue could be found in the genome of Kluyveromyces lactis

Systematic nameGeneK. lactis ORFSwiss prot qualificationSimilarity to S. cerevisiaePFAM database qualification
YEL047CFRDSVI4423HighHigh (OSM1)High
YOR011WAUS1IV091HighHigh (PDR5)High
YIL013CPDR11II2419HighHigh (PDR12)High

The 20 genes listed in Table 5 are anaerobically essential genes, genes that were linked to anaerobic growth in the literature, and transcription factors for this function. This suggests that K. lactis has deficits both in the regulation of anaerobic genes and in the presence of these genes itself. Because not all genes found to be missing are active in the same process, it could very well be that the inability of this strain to grow anaerobically has multiple causes.


Most of the genes that are essential for aerobic growth have an equally important role under anaerobic conditions, because only 33 of them are not needed at all and 32 are necessary for optimal growth in YPD supplemented with Tween and ergosterol when oxygen is absent. This figure is much smaller than anticipated, given the large number of genes encoding mitochondrial proteins. However, our data confirm that, apart from respiration, mitochondria have many other metabolic functions even under anaerobic conditions, also illustrated by the presence of (pro-)mitochondria in anaerobically grown cells (Plattner & Schatz et al., 1969). It is also remarkable that the transcription level of none of these genes changes significantly when aerobic versus anaerobic cells are compared (Ter Linde et al., 1999; Piper et al., 2004) or when mutants in the Hap1 or Rox1 anaerobic transcription factors mutants are compared to wild-type strains (Ter Linde & Steensma et al., 2002). The remaining, just over 1000, essential genes probably represent the minimum number of household genes that are necessary for growth under a wide variety of conditions. For this reason, and because of the close relationship between the two species, we have not included them in the comparison of the genomes (Bolotin-Fukuhara et al., 2000). There is a possibility, however, that these genes may have evolved in a different way, leaving them nonfunctional for the anaerobic tasks their counterparts in S. cerevisiae perform.

The number of genes, which we identified as being essential for anaerobic growth in S. cerevisiae is also small. We found 23 genes that are specifically needed for anaerobic growth, of which only two were previously described as important for anaerobic growth. Given the limitations of our screen, this is not unexpected. First, many genes involved in anaerobic growth are present in one or more copies; for instance, UPC2 and ECM22 can partially complement each other. Owing to our stringent criteria for growth, we did not consider small differences in growth to be significant. Similarly, the DAN, PAU and TIR genes are all present in multiple copies. Second, cells were grown in YPD with Tween and ergosterol. Genes involved in the synthesis of components present in the medium thus could not be detected.

The two genes that were previously described in literature as being important for anaerobic growth were NPT1 and ARV1. The NPT1 gene was the only one found in an extensive screen for essential anaerobic genes (Panozzo et al., 2002). However, the authors considered the anaerobic conditions used questionable and thus the screen was termed hypoxic rather then anaerobic. Our study confirms the importance of NPT1 for anaerobic growth.

From the results shown in Table 3, it is clear that the genes are involved in various functions. It is remarkable that two genes of the SAGA complex, HFI1 and SPT20, two of the Mediator complex, PGD1 and SIN4, and several other transcription(-related) factors, namely HPR1, ARG82, CTF4 and SNF1, have come up in our screen. Because these complexes and factors are also present and functional under aerobic conditions, it is not clear why they are essential for anaerobic growth. Possibly the combination of low ATP levels, caused by the lower yield under anaerobiosis, and impaired protein synthesis causes some sort of “synthetic lethality”. For most other genes, GUP1 excepted, it is also unclear why their disruption would lead to the inability to grow under anaerobic conditions.

In addition to the 23 genes that we found essential for anaerobic growth, we included genes that have a (potential) regulating role upon anaerobiosis and genes that are known from literature to play a role when cells are growing under anaerobic conditions (Table 4).

The comparison revealed 20 genes that are anaerobically active in S. cerevisiae and missing from K. lactis. Of the 23 genes shown to be essential under anaerobic conditions only, 11 cannot directly be assigned a homologue in K. lactis. Also, the sequences of five genes that act as regulators in the absence of oxygen are not present with high similarity. Three S. cerevisiae genes in Table 5, namely PGD1, CNM67 and ROX1, do seem to have a possible homologue in K. lactis, according to the homology of the structures as predicted by Swissprot (column 4 in Table 5), but the comparison of the sequences in the other columns is not designated ‘high’, so they were included in the list. Three S. cerevisiae genes gave high similarity with K. lactis ORFs, which gave different S. cerevisiae genes when used as probes to search the S. cerevisiae genome. For example, when the FRDS sequence was used against the K. lactis genome, the ORF klact_VI4423 was a significant hit. When this ORF is used to search Swissprot, known yeast annotations, PFAM and KOGG databases, the ORF is identified as a homologue of the OSM1 gene. The FRDS and OSM1 genes in S. cerevisiae are highly homologous. The OSM1 sequence is widely accepted as a gene and annotated as such. The FRDS sequence, however, is not. Therefore, it was not present in the databases used for comparison of the K. lactis ORF. For this particular case, it was clear that K. lactis has only one fumarate reductase enzyme, whereas S. cerevisiae has two highly similar enzymes. The other two cases, AUS1 and PDR11, are less clear. Although AUS1, PDR11, PDR5 and PDR12 all encode members of the ABC transporters, only the first two have been implicated in sterol uptake. At this moment it is not possible to draw a conclusion about the specific function of these K. lactis genes.

Saccharomyces kluyveri is another yeast that can grow under anaerobic conditions (Møller et al., 2004). To validate our data, we have checked the presence or absence in this yeast of the 20 genes in Table 5. In the BLAST search, for which the web page http://www.genetics.wustl.edu/saccharomycesgenomes/ was used, all but three were found to be present in S. kluyveri with a high similarity and a P-value of less than 10–4. The other three, PGD1, CNM67 and VPS65, were also present at a high similarity but their P-values were 0.97, 0.10 and 0.81, respectively. This comparison supports the conclusion that the genes in Table 5 might be a key to understand why K. lactis cannot grow under anaerobic conditions.

Four of the genes for which a K. lactis homologue could not be found, notably ARV1, DAN1, AUS1 and PDR11, are related to sterol uptake. Moreover, three of the missing transcription factors are also involved in sterol uptake, namely SUT1, SUT2 and UPC2. Import of sterols under anaerobic conditions is essential as their biosynthesis requires oxygen. Therefore, we hypothesize that K. lactis cannot import sterols. The lack of sterol import thus would be one factor that contributes to the inability of K. lactis to grow under anaerobic conditions. Because 14 more anaerobic genes are absent in K. lactis, it appears unlikely that sterol uptake is the only factor. For example, the S. cerevisiae gene ARV1, which was described earlier as essential for anaerobic growth and which came up in our screen for such genes, is absent in the K. lactis genome.

In addition, regulation might also play a role. For example, a single functional homologue of the AAC genes is present in K. lactis (named KlAAC), but this gene is downregulated under anaerobic conditions, leaving K. lactis with low levels of a functional ADP/ATP carrier when oxygen is absent (Trezeguet et al., 1999).

As the number of anaerobically important genes missing in K. lactis is extensive, it is probable that several of these genes will be needed to allow K. lactis to grow under anaerobic conditions. Both complementation assays and transcriptome analysis would be needed to explore this issue further. By supplying the cells with the proper genes, either encoding transcription factors or anaerobically essential proteins, K. lactis could become less dependent on the availability of oxygen, if not able to grow under completely anaerobic conditions. Experiments to test this hypothesis are in progress.


This work was supported by a grant from the Netherlands Organization for Scientific Research (NWO/ALW nr. 811.35.004). We also would like to thank Raymond Brandt for the dissection work.


  • Editor: Lex Scheffers


View Abstract