Applied Microbiology and Biotechnology

, Volume 76, Issue 2, pp 417–427

Functional analysis of very long-chain fatty acid elongase gene, HpELO2, in the methylotrophic yeast Hansenula polymorpha

  • Phatthanon Prasitchoke
  • Yoshinobu Kaneko
  • Minetaka Sugiyama
  • Takeshi Bamba
  • Eiichiro Fukusaki
  • Akio Kobayashi
  • Satoshi Harashima
Applied Genetics and Molecular Biotechnology

DOI: 10.1007/s00253-007-1012-y

Cite this article as:
Prasitchoke, P., Kaneko, Y., Sugiyama, M. et al. Appl Microbiol Biotechnol (2007) 76: 417. doi:10.1007/s00253-007-1012-y


We describe the cloning and functional characterization of the fatty acid elongase gene HpELO2, a homologue of the HpELO1 gene required for the production of C24:0 in the yeast Hansenula polymorpha. The open reading frame (ORF) of HpELO2 consists of 1,035 bp, encoding 344 amino acids, sharing about 65% identity with that of Saccharomyces cerevisiae Elo2. Expression of HpELO2 rescued the lethality of the S. cerevisiae elo2Δ elo3Δ double disruptant. An accumulation of C18:0 and a significant increase and decrease in the levels of C24:0 and C26:0, respectively, were observed in the Hpelo2Δ disruptant. These results supported an idea that HpELO2 encodes a fatty acid elongase involved in the elongation of C18:0 to very long-chain fatty acids. The Hpelo1Δ Hpelo2Δ double disruption was nonviable, suggesting that HpELO1 and HpELO2 are the only two genes necessary for the biosynthesis in H. polymorpha. Interestingly, transcription of HpELO2 and HpELO1 were found to be transiently up-regulated by exogenous long-chain fatty acids; however, this up-regulation was not observed with HpELO1 and HpELO2 genes driven by the constitutively expressed promoter of the HpACT gene, suggesting that exogenous fatty acids specifically trigger the transcriptional induction of HpELO1 and HpELO2 through their promoter regions.


Methylotrophic yeast Hansenula polymorpha Elongase for very long-chain fatty acids Transient transcriptional induction 


Fatty acids and their derivatives are essential to all eukaryotic cells owing to the crucial roles that they play in the cell. In general, fatty acids provide the cell with a concentrated energy source and serve as an important building block for membrane structure. Several studies have demonstrated that very long-chain fatty acids (VLCFAs, >18 carbons in length) are a prerequisite for membrane construction and stabilization in various eukaryotic cells. For example, VLCFAs are known to be critical components of cuticular waxes and seed oils and provide stability to pollen grains in higher plants (Cassagne et al. 1994). An appropriate level of VLCFAs is necessary to reduce the water permeability of the barrier epithelia (Zeidel 1996). A recent study has indicated that VLCFAs are essential components of detergent-resistant membrane domains (rafts) in the yeast Saccharomyces cerevisiae, where they help to establish raft association of the proton pumping [H+]-ATPase, Pma1p, and to stabilize its surface delivery via the function of sphingolipids (Gaigg et al. 2006). Even though a broad range of evidence supports the vital role of VLCFAs in cellular physiology, little is known about how cells regulate their intracellular level of VLCFA. Thus, understanding the mechanism of VLCFA biosynthesis is of importance in various organisms.

VLCFAs are synthesized by chain elongation of C16 or C18 saturated acyl-CoA, which is either provided by de novo fatty acid synthesis, through the function of fatty acid synthase complex (Fas complex) and acetyl-CoA carboxylase (Acc), or alternatively, is acquired from the medium. In S. cerevisiae, the microsomal Elo proteins play a key role in a condensation step of the fatty acid elongation of VLCFA biosynthesis (Cinti et al. 1992; Leonard et al. 2004). Three genes belonging to the ELO family (ELO1, ELO2, and ELO3) have been identified (Toke and Martin 1996; Oh et al. 1997). ELO1 plays a role in the generation of long-chain fatty acids by converting myristoyl C14-CoA to palmitoyl C16-CoA. Transcription of ELO1 is rapidly induced in wild-type cells by exogenous myristate (C14:0) and is repressed when cells are supplied with palmitate (C16:0) as a feedback responsive mechanism (Toke and Martin 1996). ELO2 and ELO3 genes are involved in the production of VLCFAs. Whereas ELO2 is required for the formation of VLCFAs of up to C24:0 from C18:0 fatty acids, ELO3 has a broader substrate specificity and is needed to synthesize C26:0 (Oh et al. 1997; Paul et al. 2006). Disruptants of either ELO2 or ELO3 are not lethal, but cells lacking both genes are nonviable (Oh et al. 1997). Expression of ELO2 and ELO3 is down-regulated in the stationary phase and under conditions of nitrogen depletion (Gasch et al. 2000). Moreover, it has been found that defects in the synthesis of VLCFAs, especially C26:0, perturbs the synthesis of sphingolipids and ceramides required for constructing many types of yeast cellular membrane (Oh et al. 1997; Stock et al. 2000; Gaigg et al. 2006).

Whereas S. cerevisiae synthesizes only monounsaturated fatty acids (MUFAs), the methylotrophic yeast Hansenula polymorpha, a nonconventional yeast with biotechnological potential, is able to produce polyunsaturated fatty acids (PUFAs), such as linoleic acid (C18:2-Δ 9,12) and linolenic acid (C18:3-Δ 9,12,15), in addition to MUFAs (Wijeyaratne et al. 1986; Anamnart et al. 1998). Interestingly, the production of γ-linolenic acid (C18:3-Δ 6,9,12) and octadecatetraenoic acid (C18:4-Δ 6,9,12,15) has been accomplished by heterologous expression of the Mucor rouxii Δ6-desaturase gene in H. polymorpha cells coupled with its own endogenous fatty acid substrates (Laoteng et al. 2005). Therefore, this yeast has been recently considered to be an advanced eukaryotic model organism that is appropriate for studying the mechanism of fatty acid synthesis, as the wide spectrum of genetic, molecular, and biochemical techniques could be applicable. For application of H. polymorpha to production of various fatty acids, our laboratory have attempted to identify and to characterize genes involved in the fatty acid biosynthetic pathway of this yeast. Two genes encoding proteins associated with the production of saturated and unsaturated long-chain fatty acids—namely, fatty acid synthase β-polypeptide (FAS1) and fatty acid Δ9-desaturase (OLE1)—have been cloned and characterized (Anamnart et al. 1997; Lu et al. 2000; Kaneko et al. 2003). Cell harboring a fas1Δ allele cannot grow without supplementation of long-chain saturated fatty acids, and disruption of OLE1 leads to cells that are unsaturated fatty acid auxotrophs. Regarding the biosynthesis of VLCFA in H. polymorpha, we previously cloned and characterized the HpELO1 gene, a homologue of S. cerevisiae ELO3, which is indispensable for the production of C24:0 (Prasitchoke et al. 2007). Cells harboring a Hpelo1Δ disruption showed hypersensitivity to Zymolyase and were more flocculent than wild-type cells, an observation that may result from a defect in sphingolipid biosynthesis, as in the case of the S. cerevisiaeelo3Δ disruptant.

Here, we describe cloning and characterization of another VLCFA elongase gene in H. polymorpha, HpELO2, a homologue of S. cerevisiae ELO2. Heterologous expression of HpELO2 gene in the S. cerevisiae elo2Δ elo3Δ double mutant and disruption of the HpELO2 gene revealed that HpELO2, in addition to HpELO1, is involved in the elongation of VLCFAs. Interestingly, transcription of HpELO1 and HpELO2 was found to be transiently induced by exogenously added long-chain fatty acids.

Materials and methods

Strains, media, and oligonucleotides

The H. polymorpha strains and S. cerevisiae strains used in this study are listed in Table 1 and were cultivated at 37 and 30°C, respectively. All mutant strains of H. polymorpha are derived from BY21400 (synonymous to CBS1976). Yeast cells were grown in rich medium (YPAD) or selective minimal medium (SD or SGal), and sporulation was carried out as described previously (Prasitchoke et al. 2007). Where required, 100 μg/ml Zeocin™ (Invitrogen, Carlbad, CA) was added to yeast media. For plate culture, 2% agar was added. For media supplemented with exogenous fatty acids, 1 mM myristate (C14:0), palmitate (C16:0), or stearate (C18:0) (Wako Pure Chemical, Osaka, Japan) was dispersed in the medium along with 1% (w/v) Brij 58 (Sigma-Aldrich, St. Louis, MO). Escherichia coli strain DH5α (Sambrook et al. 1989) was cultivated in Luria–Bertani (LB) medium, and transformants were selected on LB medium containing 100 μg/ml ampicillin or 25 μg/ml Zeocin™ antibiotics. The oligonucleotides used in this study are shown in Table 2.
Table 1

Yeast strains used in this study




Hansenula polymorpha





leu1-1 ura3-1

Prasitchoke et al. 2007



This study


ZeoR transformant of SH4329

This study


ura3-1 ade11-1 Hpelo1Δ::URA3

This study


leu1-1 Hpelo2Δ::Zeo

This study


ura3-1 leu1-1 Hpelo2Δ::Zeo

This study


ura3-1 Hpelo1Δ::URA3 ade11-1::ADE11-PHpACT-HpELO1

This study


ura3-1 Hpelo1Δ::URA3 ade11-1::ADE11-PHpACT-HpELO2

This study


ura3-1 Hpelo1Δ::URA3 Hpelo2Δ::Zeo



This study


ura3-1 leu1-1 Hpelo1Δ::URA3 Hpelo2Δ::Zeo

This study



Saccharomyces cerevisiae


MATαura3 leu2 his3 ade2-1 trp1-1 elo2Δ::KanMX elo3Δ::KanMX [pEL115s2]

This study

aYGRC is the Yeast Genetic Resource Center, Japan (

Table 2

Oligonucleotides used in this study























aComplementary nucleotides to ZeoR gene are in italics.

Plasmids and genetic manipulations

Open reading frame (ORF) of the HpELO2 was amplified by polymerase chain reaction (PCR) with primers HpELO2ORF-F and HpELO2ORF-R using the genomic DNA of H. polymorpha strain BY21400 as a template; the PCR product was then ligated into the pGEM-T easy vector (Promega Corporation, Medison, WI) to create pHpELO2ORF. PCR amplification was conducted by TaKaRa Ex Taq™ Polymerase (TAKARA BIO, Shiga, Japan). An EcoRI fragment containing the HpELO2 ORF from pHpELO2ORF was inserted after the TDH3 promoter on the multicopy vector p520 (TRP1 as selective marker) (Anamnart et al. 1997) to create pEL205sf.

Genetic and sequencing analyses were performed exactly as described previously (Prasitchoke et al. 2007). Southern and Northern blot hybridization was performed using ECL Random Prime Labeling and Detection systems (GE Healthcare UK, Little Chalfont, Buckinghamshire, UK).

Disruption of HpELO2 gene

The Hpelo2 disruptant was constructed by a double-fusion PCR-mediated gene disruption method (Amberg et al. 2005). In two separate PCR reactions, a 5' end 490-bp fragment and a 3′ end 470-bp fragment of HpELO2 were amplified by using pHpELO2ORF as the template and primers HpELO2ORF-F and HpELO2-7NR and HpELO2ORF-R and HpELO2-CF, respectively. A cassette containing the 1.25-kb Zeocin™ marker DNA was also amplified by using pREMI-Z (Mukaiyama et al. 2002) as the template and primers ZEO1-F and ZEO-1R. The first fusion PCR was then conducted by using the 5′ end 490-bp fragment of HpELO2 and the 1.25-kb Zeocin™ marker DNA as templates and primers HpELO2ORF-F and ZEO-1R, resulting in amplification of a 1.7-kb fusion product. The second fusion PCR was subsequently performed by using the first fusion PCR product and the 3′ end 470-bp fragment of HpELO2 as templates and primers HpELO2ORF-F and HpELO2ORF-R, resulting in a 2.2-kb fusion DNA product containing the Zeocin™ marker DNA cassette flanked by 5′ and 3′ regions of HpELO2. This 2.2-kb PCR fragment was used to transform H. polymorpha strain SH4329 to confer zeocin resistance (ZeoR). Colony PCR analysis of 54 ZeoR transformants using primers E2-5F and HpELO2ORF-R indicated that one candidate, H25-19, gave the expected band at 3.5 kb, instead of the 2.3-kb product corresponding to the wild-type allele (data not shown). Southern blot analysis of H25-19 genomic DNA confirmed that the wild-type HpELO2 locus had been changed to the Hpelo2Δ::Zeo locus (data not shown).

Construction of strains harboring the PHpACT-HpELO1 and PHpACT-HpELO2 expression cassette

To generate a constitutively expressed HpELO1 gene, the BamHI/XbaI fragment of HpELO1 from plasmid pEL115s2 (Prasitchoke et al. 2007) was inserted after the HpACT promoter in plasmid pHEX2, resulting in pEL126h. Plasmid pHEX2 was derived by replacing the AOX promoter of pHIPA4 (Haan et al. 2001) (kindly provided by M. Veenhuis) with a NotI/BamHI fragment carrying the HpACT promoter from pHACT850-HyL (Kang et al. 2001) (kindly provided by H-A. Kang). Plasmid pEL126h was linearized with SfiI, which cut pEL126h at a single site within the ADE11-selective marker and then introduced into Hpelo1Δ ade11-1 cells (H12-9A) by transformation. Transfomants, in which the pEL126h fragment was supposed to be integrated into the ade11-1 locus, were selected on glucose minimal medium (SD) without adenine. To confirm this targeted integration, a transformant, H126h, was crossed with strain KYC638 (leu1-1 ura3-1), and the hybrid cells were allowed to sporulate. All of the 48 spores yielded from this hybrid grew on SD without adenine, suggesting that H126h contained the expected ade11-1::[PHpACT-HpELO1-ADE11] allele. Crossing between H126h and H99-2D (Hpelo2Δ::Zeo; see Table 1) strains yielded a diploid strain, H32. A spore clone (H32-9A) of a Hpelo1Δ Hpelo2Δ double mutant carrying the constitutively expressed HpELO1 gene (PHpACT-HpELO1) was obtained among tetrads of the H32 diploid strain.

To construct a strain harboring PHpACT-HpELO2, the HpELO1 gene in pEL126h was replaced with a BamHI/ClaI fragment containing HpELO2 from pHpELO2ORF (see above), resulting in pEL214h. Plasmid pEL214h was then linearized with SfiI and was introduced into H12-9A. A transformant grown on SD without adenine, H214h, was crossed with H99-2D to yield a diploid strain, H30. The H30 diploid strain was sporulated, and a meiotic segregant with the genotype Hpelo1Δ Hpelo2Δ [PHpACT-HpELO2], H30-2D, was selected for further experiments.

Fatty acid analysis

Fatty acids were extracted from 300 ml of yeast culture by direct saponification with 10% KOH in methanol, and fatty acid methyl esters were prepared by a previously described procedure (Prasitchoke et al. 2007). The fatty acid methyl esters were then dissolved in 250 μl of hexane and analyzed by gas chromatography-mass spectroscopy (GC-MS) using a Trace DSQ system (Thermo electron, Waltham, MA) equipped with a CP-SIL 8 CB low bleed/MS (30 m × 0.25 mm I.D., df = 0.25 μm, Varian, Palo Alto, CA). The conditions and operations of GC-MS were exactly the same as those described previously (Prasitchoke et al. 2007). The identity of each peak in the chromatogram, except those of the hydroxyl derivatives, was confirmed by comparison with the mass spectra of authentic standards (Sigma-Aldrich; Tokyo Kasei Kogyo, Tokyo, Japan). For more discrete separation of unsaturated C18-fatty acid species, an OMEGAWAX 250 column (30 m × 0.25 mm I.D., df = 0.25 μm, Sigma-Aldrich), was used as described previously (Prasitchoke et al. 2007).

Nucleotide sequence

The nucleotide sequence data of HpELO2, including 5′- and 3′-portions (1,693 bp), have been deposited in the DDBJ/EMBL/GenBank nucleotide sequence databases with the accession number AB270614.


Cloning and sequencing of the HpELO2 gene

PCR amplification of H. polymorpha genomic DNA using degenerate primers designed on the basis of conserved regions of the ELO gene family yielded two distinct fragments of the expected length (∼300 bp) named EL106 and EL108 (Prasitchoke et al. 2007). On the one hand, the EL108 fragment has been previously shown to be a part of the HpELO1 gene required for the production of C24:0 (Prasitchoke et al. 2007). On the other hand, the sequence of EL106 suggested the existence of another fatty acid elongase gene in H. polymorpha, which we designated HpELO2. To clone the full-length HpELO2 gene, locus-specific gene tagging was performed. The randomly integrative/mutagenic plasmid pREMI-Z (Mukaiyama et al. 2002) was modified to a locus-specific integrative version by inserting the 0.15-kb MluI fragment of EL106 into the MluI site of pREMI-Z, resulting in pREMI-Z-E2. Plasmid pREMI-Z-E2 was linealized with SacII (unique site in EL106 region) and introduced into H. polymorpha strain SH4329. Among 175 Zeocin™ resistant transformants, only one, H21-28, showed an expected DNA band corresponding to 2.5 kb (Fig. 1a) in colony PCR using primers E2-2F and E2-2R (Table 2). As shown in Fig. 1b, Southern analysis of H21-28 using the EL106 DNA fragment as a probe revealed that two copies of the disruption cassette were integrated into the HpELO2 locus and that some regions of the putative HpELO2 locus were deleted, as indicated in Fig. 1a. Nevertheless, the 5′ and 3′ portions of the tagged integrated pREMI-Z-E2 DNA were recovered by circularization of the BglII-digested fragment (8.8 kb) of H21-28 genomic DNA, resulting in plasmid pH21-28. For sequencing analysis, a 2.3-kb BamHI/PstI fragment and a 2.5-kb BglII/PstI fragment of pH21-28 were subcloned into pBluescript II SK (+), resulting in plasmids pEL201 and pEL203, respectively (Fig. 1a). The E2-5F and E2-5R primers were subsequently designed from the sequences of pEL203 and pEL201, respectively. By using primers E2-5F and E2-5R, a 3.0-kb PCR fragment, H2-3K, containing the whole HpELO2 gene was amplified from the genomic DNA of H. polymorpha wild-type cells (BY21400) (Fig. 1c). Sequence analysis revealed that the H2-3K fragment contained an ORF of 1,035 nucleotides. This ORF contained a region of sequence identical to that of the EL106 DNA fragment; therefore, we concluded that this gene was indeed HpELO2. Chromosome blotting analysis of BY21400 using a DNA fragment of the HpELO2 ORF as a probe showed one signal corresponding to chromosome VI (data not shown) in contrast to chromosome IV, which carries the HpELO1 gene (Prasitchoke et al. 2007).
Fig. 1

Cloning of the HpELO2 gene. a Host (SH4329)-panel represents the whole wild-type allele of the HpELO2 gene. White boxes indicate the EL106 DNA fragment contained in the plasmid pREMI-Z-E2. The arrowslabeled E2-5F, E2-5R, E2-2F, E2-2R indicate the PCR primers used in this cloning. Two lines designated numbers 1 and 2 below the HpELO2 ORF represent the undeleted regions, and dotted lines represent the regions of HpELO2 that are deleted in the H21-28 strain. Transformant(H21-28)-panel represents the HpELO2 allele double-integrated with plasmid pREMI-Z-E2, linearized by SacII, in the transformant H21-28. Two lines designated numbers 1 and 2 represent the undeleted regions of HpELO2. The thin lines indicate integration of the pREMI-Z-E2 plasmid into the chromosome of H21-28. At the bottom, the 8.8-kb BglII fragment indicates plasmid pH21-28. The 2.3-kb BamHI/PstI fragment was inserted into plasmid pEL201. The 2.5-kb BglII/PstI fragment was inserted into plasmid pEL203. P PstI, S SacII, B BglII, Ba BamHI. b Southern analysis of H21-28 genomic DNA digested with the indicated restriction enzymes using the EL106 DNA fragment as a probe. c Restriction map of the HpELO2 locus and flanking region. The whitebox indicates the 380-bp EL106 DNA fragment, which was primarily amplified using degenerate primers (see text). The H2-3Kfragment indicates the resultant 3.0-kb DNA product of PCR amplification obtained with the primers E2-5F and E2-5R that contains the HpELO2 ORF (1035 bp; gray-shaded horizontal arrow). The HpELO2 ORF was amplified by HpELO2ORF-F and HpELO2ORF-R primers. E EcoRI, S SacII, M MluI, B BglII

HpELO2 is highly homologous to S. cerevisiae ELO2 gene involved in elongation of VLCFAs and rescues lethality of S. cerevisiae elo2Δ elo3Δ double disruptant

The HpELO2 ORF consists of 1,035 nucleotides encoding a putative protein of 344 amino acids, as shown in Fig. 2. The hydropathy and membrane topology analysis suggested that the HpElo2 protein contains five predicted membrane-spanning regions, which are conserved in all yeast Elo proteins. A highly conserved region in the HpElo2 protein lies between the predicted membrane-spanning regions II and III (Fig. 2). This region contains the consensus amino acids of a histidine box (HxxHH), a motif that is assumed to form an iron-binding domain and has been previously identified in several fatty acid desaturases (Fox et al. 1994; Shanklin et al. 1994) and elongases (Oh et al. 1997), including HpElo1 (Prasitchoke et al. 2007). A comparison of the amino acid sequence of HpElo2 with those of S. cerevisiae Elo1, Elo2, and Elo3 and HpElo1 showed that HpElo2 has the highest identity (65%) to S. cerevisiae Elo2, suggesting that HpELO2 is an ortholog of the S. cerevisiae ELO2 gene that is involved in VLCFA elongation.
Fig. 2

Alignment of the amino acid sequence of HpElo2. The deduced amino acid sequence of HpELO2 was aligned with the sequences of fatty acid elongases in H. polymorpha (HpElo1) and S. cerevisiae (ScElo1-ScELo3) using Clustal X. Presumptive membrane-spanning regions of HpElo2 are underlined and numbered I–V. Conserved regions in the elongase proteins, LHxxHH (histidine-box) and MYxYY (tyrosine-rich box), are indicated in bold. HpElo2 shares 47, 65, 48, and 49% identity with ScElo1, ScElo2, ScElo3 and HpElo1, respectively. Asterisks indicate identical residues. Two dots indicate highly conserved residues; single dots indicate weakly conserved residues

To assess the activity of HpELO2, a complementation test with S. cerevisiae elo mutations was performed. It has been reported that the S. cerevisiae eloelo3Δ double mutant is nonviable because of an inability to produce VLCFAs. We previously demonstrated that viability is restored to the S. cerevisiaeelo2Δ elo3Δ double mutant by expression of the H. polymorpha VLCFA elongase gene HpELO1 (Prasitchoke et al. 2007). First, therefore, we examined whether the HpELO2 gene could also suppress the lethality of the S. cerevisiae elo2Δ elo3Δ double mutant. Plasmid pEL205sf, harboring HpELO2 driven by the TDH3 promoter, was introduced into an S. cerevisiae elo2Δ elo3Δ double mutant harboring plasmid pEL115s2 (S9-17B), a URA3-based vector expressing HpELO1 driven by the GAL1 promoter. S9-17B grows on galactose medium, but cannot grow on glucose or 5′-fluoroorotic acid (5′-FOA) medium because its growth depends on the expression of HpELO1 (Prasitchoke et al. 2007). After transformation, Trp+ transformants were selected on galactose-based minimal medium (SGal) without tryptophan. One of resulting Trp+ colonies was streaked on SGal containing 5′-FOA. As shown in Fig. 3, S9-17B harboring vector p520 (S9-17B [p520]) did not grow, whereas S9-17B harboring pEL205sf (S9-17B [pEL205sf]) grew. Moreover, S9-17B [pEL205sf] grew and the wild-type strain even on glucose, which would not induce the GAL1 promoter for HpELO1 expression, indicating that expression of HpELO2, in addition to that of HpELO1, suppressed the lethality of the S. cerevisiae elo2Δ elo3Δ double mutant. As mentioned above, because S. cerevisiae ELO2 and ELO3 are essential for the fatty acid elongation of C18:0 to VLCFAs, this observation suggests that HpELO2 is involved in the elongation of VLCFAs from C18:0 substrates.
Fig. 3

Expression of HpELO2 suppresses the lethality of the S. cerevisiae elo2Δ elo3Δ double disruptant. Cells were streaked on glucose-minimal medium (SD), galactose-minimal medium (SGal), or SGal containing 5′ FOA (SGal+FOA), and photographs were taken after 2 days of incubation at 30°C

Characterization of Hpelo2Δ disruptant

We further investigated the function of HpELO2 by constructing an Hpelo2Δ disruptant strain. For this purpose, we newly created the Hpelo2Δ disruptant strain by using ZeoR gene as described in the Materials and methods section. Disruption of HpELO2 in H. polymorpha haploid cells did not cause lethality on YPAD medium at 37°C, which suggested that HpELO2 is not an essential gene under this cultivation condition. We previously reported that the Hpelo1Δ disruptant exhibited hypersensitivity to Zymolyase and a flocculent phenotype (Prasitchoke et al. 2007). However, the disruptant of HpELO2 did not display such sensitivity to Zymolyase or cell flocculency (data not shown), suggesting that unlike the Hpelo1Δ disruptant, the cell wall structure of Hpelo2Δ seems to be normal.

Because the S. cerevisiaeelo2Δ elo3Δ double disruptant is lethal (Oh et al. 1997), we also examined whether the double disruption of HpELO1 and HpELO2 would cause lethality in H. polymorpha. A hybrid between Hpelo1Δ (H12-9A: ura3-1 ade11-1 Hpelo1Δ::URA3) and Hpelo2Δ (H99-2D: ura3-1 leu1-1 Hpelo2Δ::Zeo) disruptants was constructed and subjected to tetrad analysis. Among 34 tetrads tested so far, the ratio of tetrad distribution on spore viability was 5:19:10 in 4 viable: 3 viable: 2 viable and no viable Ura+ Zeo+ spore clone was obtained, while other combinations of spore clones were observed. This result suggests that Hpelo1Δ Hpelo2Δ double disruption is lethal in H. polymorpha as in the case of elo2Δ elo3Δ double disruption in S. cerevisiae. This observation also indicates that HpELO1 and HpELO2 are the only two genes necessary for the biosynthesis of VLCFAs in H. polymorpha.

GC-MS analysis of the fatty acid composition of the Hpelo2Δ disruptant showed that the amount of C18:0 was increased threefold compared with the wild-type strain (Table 3), suggesting that HpELO2 is involved in elongation of C18:0 to VLCFAs. This observation is consistent with the results that expression of HpELO2 rescued the lethality of the S. cerevisiae elo2Δ elo3Δ double disruptant, which is not able to elongate C18:0 to VLCFAs (Oh et al. 1997). A marked increase in the amount of C24:0 (tenfold higher than in the wild type) and a significant decrease in C26:0 (threefold) were also observed in the Hpelo2Δ disruptant (Table 3). Moreover, in our previous work (Prasitchoke et al. 2007), the Hpelo1Δ disruptant, which expresses only HpELO2, showed a moderate and large accumulation of C20:0 and C22:0, respectively, coupled with a significant decrease in C24:0 and C26:0 to undetectable levels, suggesting that HpELO2 is not involved in conversion of C22:0 to C24:0. Taken together, these results suggest that HpELO2 is involved in the elongation of C18:0 to C22:0 and is also required, together with the HpELO1, for maintaining normal production levels of C26:0.
Table 3

Fatty acid composition of wild type H. polymorpha and the Hpelo2Δ mutant
















Wild type

0.5 ± 0.1

17 ± 2.1

2.3 ± 0.6

2.9 ± 0.7

44 ± 3.0

29 ± 3.0

0.02 ± 0.0

0.03 ± 0.0

0.1 ± 0.1

0.03 ± 0.0


0.02 ± 0.0

0.03 ± 0.0
















0.4 ± 0.1

16 ± 2.0

0.7 ± 0.1

10 ± 1.8

34 ± 6.9

37 ± 3.4

0.02 ± 0.0

0.4 ± 0.6

0.06 ± 0.0

0.04 ± 0.0


0.2 ± 0.1

0.01 ± 0.0

*Fatty acid species with significant difference between wild-type and Hpelo2Δ disruptant were marked with asterisk along with p value by a Student’s paired t test (p < 0.05).

**p = 0.054

***p = 0.097

aFatty acid composition in wild-type H. polymorpha and the Hpelo2Δ mutant grown on glucose-rich medium for 24 h. Cells with the same OD (OD660) were subjected to extract fatty acids. Fatty acyl species were identified by GC-MS as described in the Materials and methods section.

bData were cited from Prasitchoke et al. (2007).

cUD Undetectable (<0.0001 mol%).

Transcription of HpELO2 is transiently induced by long-chain fatty acids myristate (C14:0), palmitate (C16:0), and stearate (C18:0)

Several reports have shown that exogenously added long-chain fatty acids mediate the transcriptional control of genes involved in fatty acid biosynthesis (Kamiryo et al. 1976; Chirala 1992; Mcdonough et al. 1992; Toke and Martin 1996). We therefore examined the effect of long-chain fatty acids on transcription of HpELO2 gene by determining time course of HpELO2 transcription in H. polymorpha grown on medium containing myristate (C14:0), palmitate (C16:0), or stearate (C18:0). Wild-type strain was grown to exponential phase in YPAD and then shifted to SD medium supplemented with appropriate amino acids and 1 mM myristate, palmitate, stearate, or without fatty acids. Total RNA was prepared from cells collected at appropriate intervals after medium shift and subjected to Northern analysis. Transcription of HpELO1 or HpELO2 in the wild-type cells was low and did not change after the cells were exposed to medium without fatty acid supplementation (data not shown). Interestingly, transcription of HpELO2 in wild-type cells had increased markedly 30 min after the cells were transferred to medium supplemented with myristate, palmitate (Fig. 4a), or stearate (data not shown) but had decreased back to the original levels at 4 h. Moreover, this transient increase of HpELO2 transcription was also observed in Hpelo1Δ disruptant cells grown on medium supplemented with 1 mM myristate (Fig. 4b) or palmitate (data not shown). It is noted that transcription of HpELO1 in wild-type or Hpelo2Δ strain also displayed the transient increase after the cells were transferred to medium supplemented with myristate (Fig. 4a, b), palmitate (Fig. 4a), or stearate (data not shown). These observations suggested that transcription of HpELO1 and HpELO2 were transiently stimulated by exogenously added long-chain fatty acids. To examine whether long-chain fatty acid exerts the effect through promoter region of HpELO1 and HpELO2, we replaced the native promoter of either HpELO1 or HpELO2 with the constitutively expressing promoter of the HpACT gene (PHpACT). The Hpelo1Δ disruptant containing the constitutively expressed PHpACT-HpELO2 gene, H214h, was crossed with the Hpelo2Δ disruptant (H99-2D) and the resultant diploid was sporulated. A spore clone (H30-2D, see Materials and methods section) harboring the Hpelo1ΔHpelo2Δ double disruption with the PHpACT-HpELO2 cassette integrated into the ade11-1 locus was selected and subjected to Northern analysis. Transcription of PHpACT-HpELO2 in H30-2D cells was constitutively high and did not increase even when the cells were exposed to medium supplemented with 3 mM fatty acid (Fig. 4c). This was also true in the case of PHpACT-HpELO1 expression in the Hpelo1Δ Hpelo2Δ double disruptant (H32-9A) harboring the PHpACT-HpELO1 gene (Fig. 4c). These results indicate that exogenously supplemented long-chain fatty acids trigger transient induction of HpELO1 and HpELO2 at a transcriptional level through their promoter regions, but not at a posttranscriptional level.
Fig. 4

Exogenous long-chain fatty acids transiently induce the transcription of HpELO1 and HpELO2. Northern analysis of HpELO1 or HpELO2 in wild-type (KYC638) (a), Hpelo1Δ (H12-9A) and Hpelo2Δ (H99-2D) (b) cells after exposure to medium supplemented with 1mM myristate or palmitate. Each strain was grown at 37°C on YPAD medium to exponential phase (OD660 1.0) and then shifted to SD medium (supplemented with appropriate amino acids) containing 1 mM myristate or palmitate. Total RNA was prepared from the cells immediately before exposure (0 h) and at 0.5, 1, 2, 3, 4, 5, and 6 h after exposure. Equal amounts of RNA (24 μg) were separated by electrophoresis on a 1.5% gel in the presence of formaldehyde, transferred to a nylon filter, and hybridized with a DNA fragment of either the ORF of HpELO1 or HpELO2 as a probe, or the ORF of HpACT as an internal control. The RNA blots were first probed with the HpELO1 or HpELO2 fragment and then reprobed with the HpACT fragment after allowing the luminescence to decrease to a negligible level by extending the exposure time to overnight at room temperature. c Cells of a Hpelo1Δ Hpelo2Δ double disruptant harboring PHpACT-HpELO1 (H32-9A) or PHpACT-HpELO2 (H30-2D) grown on YPAD at 37°C to exponential phase (OD660 1.0) were transferred to SD medium supplemented with 3 mM myristic acid. At each time point, the cells were sampled, total RNA was extracted, and then 24 μg of total RNA from each sample was subjected to Northern analysis as described above. The RNA blot was first probed with the HpELO1 or HpELO2 fragment and then reprobed with the HpACT fragment as an internal control


We previously reported the identification and characterization of the HpELO1 gene involved in elongation of VLCFAs and also provided evidence of another VLCFA elongase in H. polymorpha (Prasitchoke et al. 2007). Here, we analyzed the second elongase gene, termed HpELO2. Sequence analysis of the HpELO2 gene and comparison of its deduced amino acid sequence revealed that the HpELO2 gene product is homologous to S. cerevisiae Elo2, which catalyzes the condensation step in elongation process of VLCFAs (Oh et al. 1997). Expression of HpELO2 rescued the growth lethality of the S. cerevisiae elo2Δ elo3Δ double disruptant, which is known to fail to produce fatty acids longer than C18:0 (Oh et al. 1997), suggesting that HpElo2 elongates the C18:0 fatty acid to VLCFAs. This conclusion is consistent with the fatty acid composition of the Hpelo2Δ disruptant, which showed an accumulation of C18:0 (Table 3). Furthermore, we also showed that the Hpelo1Δ Hpelo2Δ double disruptant, like the S. cerevisiae elo2Δ elo3Δ double disruptant, was nonviable, suggesting that HpELO1 and HpELO2 are the only two genes required for VLCFA production in H. polymorpha cells. Because the Hpelo1Δ disruptant, which expresses the wild-type HpELO2 gene, produces a moderate and large accumulation of C20:0 and C22:0, respectively (Prasitchoke et al. 2007), it is probable that HpElo2 function is required for the elongation of both C18:0 to C20:0 and C20:0 to C22:0. Moreover, the Hpelo2Δ disruptant showed alterations in VLCFA composition including increases in C18:0 (threefold) and C24:0 (tenfold), and a moderate (threefold) decrease in C26:0 (Table 3) compared with the wild type. On the basis of these observations, we conclude that HpELO2 is involved in fatty acid elongation of C18:0 chains to C20:0, C20:0 to C22:0, and C24:0 to C26:0 but is not concerned with conversion of C22:0 to C24:0. Moreover, HpElo2 may have higher activity toward C18:0 substrate than HpElo1 because no accumulation of C18:0 was found in the Hpelo1Δ disruptant (Table 3; Prasitchoke et al. 2007). Unlike in the Hpelo1Δ strain, however, almost all species of VLCFAs found in H. polymorpha wild-type cells were observed in the Hpelo2Δ strain (Table 3). Because Hpelo2Δ cell expresses wild-type HpELO1, this suggests that the substrate specificity of HpElo1 is broader (i.e., C18:0 to C24:0) than that of HpElo2.

There was also a large accumulation (tenfold) of 2-hydroxylated C18:0 (C18:0-2-OH) in the Hpelo2Δ cells (Table 3). This would be explained as well as the accumulation of C22:0-2-OH in the Hpelo1Δ disruptant (Prasitchoke et al. 2007) by an idea that the accumulated C18:0 in the Hpelo2Δ disruptant is unusually incorporated into phytoceramide (Dunn et al. 1998) and consequently hydroxylated on the C-2 position in the sphingolipid synthesis pathway due to the insufficiency of C26:0. It should also be noted that C16:1 significantly decreased (threefold) in the Hpelo2Δ disruptant, although we cannot explain this phenomenon at the present moment.

HpElo1 and HpElo2 must have functional redundancy in the processes required for VLCFA production because either gene is sufficient for sustaining cell growth. Whereas cells of the Hpelo1Δ disruptant exhibited hypersensitivity to Zymolyase and a more flocculent phenotype than wild-type cells (Prasitchoke et al. 2007), however, cells of the Hpelo2Δ disruptant did not (data not shown). This difference can be explained by the idea that a change in cell wall composition occurs in the Hpelo1Δ disruptant but not at least in so far as to cause hypersensitivity to Zymolyase, in the Hpelo2Δ disruptant. We proposed previously that hypersensitivity to Zymolyase and cell flocculency in the Hpelo1Δ disruptant might be caused by a reduction in the glycosphingolipids IPC and M(IP)2C (Schweizer 2004; Prasitchoke et al. 2007). This reduction in IPC and M(IP)2C sphingolipids is not as severe in the S. cerevisiae elo2Δ disruptant as in the elo3Δ disruptant (Oh et al. 1997). If this differential also occur in Hpelo1Δ and Hpelo2Δ cells, it may be possible that the phenotypes of the Hpelo1Δ disruptant are not found in the Hpelo2Δ disruptant. In this manner, it is important to note that whereas C26:0 was not observed in cells of the Hpelo1Δ disruptant (Prasitchoke et al. 2007), it was detected in Hpelo2Δ cells (Table 3). In S. cerevisiae, C26:0 fatty acid plays an important role in the biosynthesis of sphingolipid, which is essential for establishing the association of proteins with rafts at the cell surface (Eisenkolb et al. 2002; Gaigg et al. 2006). Therefore, a defect in the synthesis of C26:0 might lead to cells that are aberrant in sphingolipid biosynthesis in the Hpelo1Δ disruptant, but not in the Hpelo2Δ cell, which produces C26:0.

Long-chain fatty acids are essential to all eukaryotic cells. On the other hand, these molecules are also toxic to cells when present in excess amounts (Oshiro et al. 2003; Tong et al. 2006). Because balancing cellular fatty acid homeostasis is critical, eukaryotic cells have a mechanism that senses exogenous fatty acids and responds by reducing de novo fatty acid biosynthesis (Ohlrogge and Jaworski 1997). In yeast, transcription of the fatty acid synthase genes (FAS1 and FAS2) and the acetyl-CoA carboxylase gene (ACC1) involved in de novo fatty acid synthesis is repressed by exogenously added long-chain fatty acids (Chirala 1992; Kamiryo et al. 1976, 1979). In contrast, our results revealed that transcription of HpELO1 and HpELO2 is transiently up-regulated by exogenously supplemented long-chain fatty acids (Fig. 4a). We have demonstrated previously (Prasitchoke et al. 2007) and in this work that HpELO1 and HpELO2 are required for the elongation of long-chain fatty acid, namely, C18:0, to VLCFAs; in other words, the function of the proteins encoded by HpELO1 and HpELO2 reduces the size of the intracellular long-chain fatty acid pool. Taken together, our finding suggests that cells respond to exogenous long-chain fatty acids not only by decreasing de novo biosynthesis but also by increasing their utilization of fatty acids to balance their cellular fatty acid pools. Moreover, our results revealed that transient induction of HpELO1 and HpELO2 transcription by exogenous fatty acids requires their promoter regions because transcription of the reconstituted genes, PHpACT-HpELO1 and PHpACT-HpELO2 was not up-regulated by exogenously added myristate (Fig. 4c). Although we did not delimit in this work the elements that regulate HpELO1 and HpELO2 expression in response to exogenous long-chain fatty acid signaling, we found a 9-bp element in the upstream regions of HpELO1 and HpELO2 at positions −648/−640 and −553/−545, respectively, having a consensus sequence 5′-GCAGAATCA-3′. The role of this element in HpELO transcription remains to be determined. Finally, we suggest that understanding the mechanism triggering the transcription of the HpELO genes by long-chain fatty acids will be of benefit to control the production of VLCFAs in H. polymorpha.


We would like to thank Y. Sakai for providing plasmid pREMI-Z, M. Veenhuis for providing plasmid pHIPA4, and H. A. Kang for providing plasmid pHACT850-HyL.

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  • Phatthanon Prasitchoke
    • 1
  • Yoshinobu Kaneko
    • 1
  • Minetaka Sugiyama
    • 1
  • Takeshi Bamba
    • 2
  • Eiichiro Fukusaki
    • 1
  • Akio Kobayashi
    • 1
  • Satoshi Harashima
    • 1
  1. 1.Department of Biotechnology, Graduate School of EngineeringOsaka UniversityOsakaJapan
  2. 2.Department of Applied Environmental Biology, Graduate School of Pharmaceutical ScienceOsaka UniversityOsakaJapan

Personalised recommendations