Identification and characterization of lipases from Malassezia restricta, a causative agent of dandruff

Transcription

1 FEMS Yeast Research, 15, 2015, fov078 doi: /femsyr/fov078 Advance Access Publication Date: 22 August 2015 Research article RESEARCH ARTICLE Identification and characterization of lipases from Malassezia restricta, a causative agent of dandruff Bettina Sommer 1,2,DavidP.Overy 1,2,3 and Russell G. Kerr 1,2,4, 1 Department of Chemistry, University of Prince Edward Island, 550 University Ave., Charlottetown, PEI C1A 4P3, Canada, 2 Nautilus Biosciences Canada, Duffy Research Center, 550 University Ave., Charlottetown, PEI C1A 4P3, Canada, 3 Department of Pathology and Microbiology, Atlantic Veterinary College, University of Prince Edward Island, 550 University Ave., Charlottetown, PEI C1A 4P3, Canada and 4 Department of Biomedical Sciences, Atlantic Veterinary College, University of Prince Edward Island, 550 University Ave., Charlottetown, PEI C1A 4P3, Canada Corresponding author: Department of Chemistry, University of Prince Edward Island, 550 University Ave., Charlottetown, PEI C1A 4P3, Canada. Tel: +(902) ; Fax: +(902) ; rkerr@upei.ca One sentence summary: Genome mining, annotation, recombinant expression and proof of activity of new Malassezia restricta lipases in the context of dandruff. Editor: Guenther Daum ABSTRACT Dandruff, a skin disorder affecting 50% of the world population, is linked with proliferation of lipophilic yeasts of the genus Malassezia (particularly Malassezia globosa and M. restricta). Most Malassezia species show a unique lipid dependency and require external lipids for growth. Genome mining of the incomplete M. restricta genome led to the identification of eight lipase sequences. Sequences representing the class 3 and LIP lipase families were used to clone the lipases MrLip1, MrLip2 and MrLip3, recombinantly expressed in Pichia pastoris, and tested for their activity using mono-, di- and triacylglycerol substrates. Hydrolysis by the M. restricta lipase MrLip1 and MrLip2 (family class 3) was limited to the mono- and diacylglycerol, while MrLip3 (family LIP) hydrolyzed all three substrates. This result confirms that Malassezia family LIP lipases are responsible for the hydrolysis of triacylglycerols, the main component of human sebum. Furthermore, the information regarding lipases from M. restricta presented here might aid in the search for anti-dandruff agents. Keywords: dandruff; Malassezia restricta; lipases; diacylglycerol-like lipase INTRODUCTION Dandruff (D), a mild form of seborrheic dermatitis (SD), is a skin disorder affecting 50% of the world population at some time during their lifespan regardless of gender or ethnicity (Hay and Graham-Brown 1997). Malassezia spp. (particularly Malassezia globosa and M. restricta) are considered as an etiopathological factor of D/SD as well as other dermatological disorders such as pityriasis versicolor and atopic dermatitis (Kieffer et al. 1990; Guehoet al. 1998; Ashbee and Evans 2002; Cafarchia and Otranto 2008; Guillot, Hadina and Gueho 2008). In these conditions, proliferation of the commensal Malassezia population is strongly correlated with increased sebaceous gland activity (Gupta et al. 2001) and symptom generation (such as scalp inflammation, irritation and flaking) is linked with desquamation impairment of the strata corneum (Gupta et al. 2004). Although human sebum is a complex mixture of triacylglycerides, fatty acids, wax esters, sterol esters, cholesterol, cholesterol esters and squalene (Downing et al. 1983; Wheatley 1986), sebum primarily consists of free fatty acids, triacylglycerides and wax esters, which are readily broken down by the skin microbiome into Received: 16 June 2015; Accepted: 13 August 2015 C FEMS All rights reserved. For permissions, please journals.permissions@oup.com 1

2 2 FEMS Yeast Research, 2015, Vol. 15, No. 7 diacylglycerides, monoacylglycerides and C16 C18 fatty acids (predominantly stearic, oleic, linoleic, palmitic, sapienic and palmitoleic acid) (Ro and Dawson 2005). With exception of the palms and the soles, sebaceous glands are located over the entire skin surface of the human body (Strauss and Pochi 1968); however, sebum secretion and the associated Malassezia population are highest on the scalp, face, chest and back. All members of the genus Malassezia, with the exception of M. pachydermatis, have a unique lipid dependency (Midgley 2000; Batraet al. 2005). As a result, the fungi produce lipolytic enzymes, to digest triacylglycerides contained in the sebum. Genome sequencing of M. globosa reported in 2007 revealed the absence of a fatty acid synthase explaining the observed lipid dependency (Xu et al. 2007)to assimilate fatty acids from external sources. To compensate for this lipid dependency, the genome of M. globosa encodes for 14 lipases, of which 13 are predicted to be secreted (Xu et al. 2007), the genome of M. sympodialis contains at least nine lipases (Gioti et al. 2013), and four lipase sequences and two lipase sequence fragments are known from M. furfur (Shibata et al. 2006). As it is hypothesized that Malassezia lipases interact with the sebum of the strata cornea to mediate the pathogenic processes (Xu et al. 2007), lipases are therefore thought to be relevant therapeutic targets for the treatment of D/SD. DeAngelis et al. (2007) successfully isolated and expressed the M. globosa lipasemglip1 and demonstrated specific activity against mono- and diacylglycerols but not triacylglycerols, a unique trait that is quite uncommon in the large lipase protein family (EC ). The substrate specificity of MgLip1 does not fit with the assumption that lipases from Malassezia spp. digest triacylglycerides contained in sebum. Subsequently, a second secreted M. globosa lipase, MgLip2, was identified and purified (Juntachai, Oura and Kajiwara 2011); however, substrate specificity with mono-, diand triacylgylcerols was not evaluated. Most recently, a third lipase MgMDL2, also a diacylglycerol lipase, was characterized and described (Xu et al. 2015). As mentioned, M. globosaand M. restricta are most commonly found on the human skin and associated with D/SD, where M. restricta is more prevalent in Asian populations (Lee et al. 2013). Our current understanding of lipase expression in M. restricta is considerably less compared to M. globosa. Unlike M. globosa, the genome of M. restricta is partially sequenced; however, when compared with the M. globosa genome an average amino-acid sequence identity of 77% exists among reciprocal best-matched proteins (Xu et al. 2007). Sequence homology suggests that the M. restricta genome encodes for at least five lipases; however, no lipase sequences have been published (Xu et al. 2007; Boekhout et al. 2010). A recent study investigating the contribution of lipases in the virulence of M. restricta demonstrated that three hypothetical lipases were expressed on the skin of patients with SD (Lee et al. 2013). Described here are the full nucleotide- and amino-acid sequences of four lipases from M. restricta as well as fragments of additional four hypothetical lipases. Three lipases, MrLip1, MrLip2 and MrLip3, which belong to two different lipase class families, were cloned and recombinantly expressed in Pichia pastoris. Additionally, we performed substrate activity assays to establish substrate specificity of the expressed M. restricta lipases to predict a functional role of these lipases as virulence factors in the common hyperproliferative skin disorders D/SD. MATERIAL AND METHODS In silico analysis of protein sequences For protein alignment and homology searches, NCBI Gen- Bank BLAST server P ( was used (Altschul et al. 1997). The protein sequences from M. restricta were analyzed by using the Pfam 27.0 server (pfam. xfam.org) (Finn et al. 2014) from the European Molecular Biology Laboratory. Prediction of the signal peptides was performed using SignalP 4.1 at the CBS prediction server ( services/) (Petersenet al. 2011). Construction of the phylogenetic tree Amino-acid protein sequences were aligned with Clustal Omega ( from the European Molecular Biology Laboratory (McWilliamet al. 2013). The phylogenetic analyses were performed using the software package MEGA6 (Tamura et al. 2013). The evolutionary history was inferred using maximum likelihood based on the Whelan And Goldman model to show the highest log likelihood ( ) (Whelan and Goldman 2001). A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = )). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. Phylogenetic trees of the protein sequences were calculated based on the maximum composite likelihood method with 1000 bootstrap replications (n = 1000). Strains and plasmid Escherichia coli NEB 5α (fhua2 (argf-lacz)u169 phoa glnv44 80 (lacz)m15 gyra96 reca1 rela1 enda1 thi-1 hsdr17) was purchased from New England Biolabs (Whitby, Canada). The P. pastoris strain GS115 (his4) and the shuttle vector ppic3.5 were contained in the Pichia Expression Kit from Life Technologies (Carlsbad, USA). Cloning of Mrlip1, Mrlip2 andmrlip3 Genomic DNA from M. restricta strain ATCC was isolated using the fast DNA extraction kit (FASTDNA SPIN KIT FOR SOIL R, MP Biomedicals) according to the manufacturer s protocol. The 3 primer Mr 3144 (Table 1) was designed to amplify the Table 1. Used primers for primer walk and cloning of Mrlip1, Mrlip2 and Mrlip3. primer name Sequence Mr 3144 AGC GAG TGA CCA ATT ATG GTG ACT CG 5 Mrlip1 BamHI ATA TGG ATC CAA TAA TGT CTC TCT TCA ATC GTA TTG CTC TGT T 3 Mrlip1 NotI ATG CGC GGC CGC TTA TTA ATG ATG ATG ATG ATG ATG CAT GTT GCC AAC CTT GGC 5 Mrlip2 BamHI ATA TGG ATC CAA TAA TGT CTG CTT GCT TCC GTG TGA TC 3 Mrlip2 NotI ATG CGC GGC CGC TTA TTA ATG ATG ATG ATG ATG ATG ATC CTT TCC AAC TGT GGC AG 5 Mrlip3 EcoRI ATA TGA ATT CAT AAT GTC TGT CTC TCT TCT CTG GAA ATT CAC T 3 Mrlip3 NotI ATG CGC GGC CGC TTA TTA ATG ATG ATG ATG ATG ATG CGA AAC TAT TCT CCT GCG ACG

3 Sommer et al. 3 incomplete lipase sequence included in the shot gun M. restricta contig MRE Single primer gradient PCR using Economy Tag R PLUS GREEN 2x master Mix (Lucigen) was performed in an Eppendorf Master Cycler nexus gradient (Eppendorf, Mississauga, Canada) beginning with 94 C treatment for 2 min followed by 30 cycles of 30 s at 94 C, 30 s at Cand1min at 72 C. The PCR was completed with a final extension for 10 min at 72 C. The PCR products were sequenced at Eurofins MWG Operon LLC (Huntswill, USA). From the obtained complete sequence from Mrlip1, Mrlip2 and Mrlip3, primers were designed (Table 1). To ensure an easy purification of the enzymes, the codon sequence for six histidines (shown in bold) was added to the 3 primers. The PCR was performed using KOD Hot Start Master Mix (Merck) and 2 μl genomic DNA as template according to manufacturer s specification. The purified PCR products as well as the vector ppic3.5 were digested using the restriction sites BamHI or EcoRI and NotI (recognition sites are underlined), ligated with T4 Ligase (NEB Biolabs) and transformed in competent E. coli cells (Neb 5α). After successful cloning, the final constructs were sequenced to confirm correctness. Production of recombinantly expressed MrLip1, MrLip2 and MrLip3 MrLip1, MrLip2 and MrLip3 were expressed in P. pastoris strain GS115 (his4) (Life Technologies, Carlsbad, USA) according to the Pichia Expression Kit protocol. Supernatants containing heterologous expressed proteins were concentrated using Macrosep R Advance centrifugal filter devices (VWR, Radnor, USA) with a molecular mass cut-off of 3 kda. Protein purification was performed using a Ni 2+ -NTA column (Thermo Scientific, Waltham, USA; binding buffer: 100 mm Hepes, 25 mm Imidazol, ph 8 and elution buffer 100 mm Hepes, 500 mm Imidazol, ph 7). Desalting of the proteins was performed with PD-10 columns (GE Healthcare, Mississauga, Canada). Protein concentration was measured at 280 nm with unfolded protein in 8 M urea. The extinction coefficients were calculated by the ExPASY ProtParam tool (Gasteiger et al. 2005). Lipase activity assays Lipase activity was determined by HPLC measurements. Assay mixtures contained 20 mm Mes buffer ph 5.5, 2.7 mm solubilized monoolein, diolein (1 2 isomer) (Nu-chek Prep, Inc., Elysian, USA) or triolein (Sigma-Aldrich, Oakville, Ontario, Canada)aswellas24μg ml 1 MrLip1, 64 μg ml 1 MrLip2 and 15 μgml 1 MrLip3. The glycerides were solubilized by sonication using a sonicator 3000 (Mixonix, New York, USA). After 24 h at 30 C, 200 rpm, the assays were frozen to stop the enzyme activity and freeze-dried overnight (VirTis, Freezemobil 25 ES, Warminster, PA, USA). The dried samples were extracted with 1:1 ethyl acetate and n-hexane (VWR), dried using a Gene Vac EZ-2 Plus (New York, USA) and solubilized in 1:1 isopropylalcohol and acetonitrile (VWR). All controls and each enzyme reaction were performed in triplicate. Monoolein, diolein, triolein and oleic acid were quantified by a Thermo FINNIGAN Surveyor HPLC coupled with an evaporative light-scattering detector (Sedex 55) and equipped with an auto sampler and a degasser (Sedere, Sedex 60LT, Alfortville, France). The separation was performed on a Phenomenex Synergi 4u MAX-RP 80A column (length 150 mm, Torrance, USA) with buffer A (acetonitrile/water (9:1) and 0.1% formic acid) as mobile phase for 5 min and a gradient to 100% buffer B (isopropyl alcohol/acetonitrile (1:1) and 0.1% formic acid) for additional 5 min. Finally, the column was washed with 100% buffer B for 20 min. The flow was 1 ml min 1 and the sample volume 10 μl. RESULTS AND DISCUSSION Sequences of Mrlip1, Mrlip2, Mrlip3 andmrlip4 To investigate lipases from M. restricta, a BLASTN search was performed with all 14 lipase sequences previously reported from M. globosa against the genome shotgun contigs from M. restricta in the NCBI GenBank database resulting in eight hits consisting of three complete lipase sequences and five partial sequences (summarized in Table 2). Within contig MRE 3144, a partial lipase nucleotide sequence (549 bp in length) was found similar to Mglip1. For this partial sequence, a 3 primer was designed and a gradient PCR was performed on gdna from M. restricta. The resulting PCR product was sequenced to elucidate the complete lipase sequence, herein designated as Mrlip1 (total length of 915 bp). Additional lipase sequences were found within contigs MRE 0242 and MRE 0614 (both sharing similarities with MGL 3878) where MRE 0242 contained a 405-bp partial sequence of a hypothetical lipase (Table S1, Supporting Information) and MRE 0614 contained the full sequence of the named lipase Mrlip4. Interestingly, the fragments of MrLip0242 and MrLip4 were remarkably similar, differing only in seven amino-acid residues (E275D, Y277H, S282F, I284T, C285G, Y288H and V300A) and three additional amino acids (Y301, L302, P303) in the last 27 amino acids of MrLip4. Unfortunately, obtaining the full sequence of Mrlip0242 is not possible by primer walking, because of the close similarity to Mrlip4. The near-identical sequence similarity observed may be the result of a gene duplication during the evolution of the M. restricta species, in a similar manner Table 2. Overview over the four complete and the four partial lipase sequences from M. restricta. Lipases M. restricta Accession Nb. Gene Contig Accession Nb. Contig Quality of AS MW [kda] Corresponding M. globosa Gene ID Complete sequence MrLip1 KR MRE 3144 AAXK MgLip1 (MGL 0797) Yes MrLip2 KR MRE 2296 AAXK MGL 1331 Yes MrLip3 KR MRE 2625 AAXK MGL 4052 Yes MrLip4 KR MRE 0614 AAXK MGL 3878 Yes MrLip0242 / MRE 0242 AAXK / MGL 3878 No MrLip0571 / MRE 0571 AAXK / MGL 3507 No MrLip2630 / MRE 2630 AAXK / MGL 1769 No MrLip3873 / MRE 3873 AAXK / MGL 0279 No

4 4 FEMS Yeast Research, 2015, Vol. 15, No. 7 Table 3. Overview, classification and secretion of M. restricta lipases; (n.a. = not applicable). Lipases M. restricta Lipase class 3 Lipase class Lip E-values Predicted to be secreted Length of Signal peptides [AS] MrLip1 x 1.4 e 23 x 19 MrLip2 x 3.6 e 19 x 21 MrLip3 x 7.7 e 65 x 21 MrLip4 x 2.2 e 26 x 16 MrLip0242 x 1.5 e 13 n.a. n.a. MrLip0571 x 2 e 29 n.a. n.a. MrLip2630 x 8.3 e 7 n.a. n.a. MrLip3873 x 7.9 e 13 n.a. n.a. as proposed by Xu et al. for the sequence similarities shared between MgLip1 and MGL 0800 in M. globosa (Boekhout et al. 2010). The remaining two complete lipase nucleotide sequences were found within contigs MRE 2296 and MRE Within contig MRE 2296, a sequence designated as Mrlip2 was found to be similar to MGL 1331, and within contig MRE 2625 a sequence sharing homology with MGL 4052 was identified and designated as Mrlip3. Additional hypothetical lipase sequence fragments (Mrlip0571 (Table S2, Supporting Information), Mrlip2630 (Table S3, Supporting Information) and Mrlip3873 (Table S4, Supporting Information)) were also found within the contigs MRE 0571, MRE 2630 and MRE 3873, having homology with MGL 3507, MGL 1769 and MGL 0279, respectively. However, due to the predicted length of the outstanding sequence ( bp), no additional sequencing efforts were attempted to obtain the complete sequences for these lipase nucleotide sequence fragments. Confirmation that all sequences encode for lipases To confirm the determined protein sequences from the M. restricta genome were indeed lipases, a sequence search query against the Pfam Database (Finn et al. 2014) was performed. MrLip1, MrLip2 and MrLip4 returned the best match for protein sequences corresponding with the lipase 3 protein family (lipase class 3, accession number PF01764) with highly significant E-values (Table 3), while the closest match for MrLip3 was the lipase LIP protein family (lipase class LIP, accession number PF03583). Sequence fragments of MrLip0571, MrLip2630 and MrLip3873 were also searched against the Pfam Database and returned closest matches for MrLip2630 and MrLip3873 with the lipase protein family class 3, and the hypothetical lipase fragment of MrLip0571 was classified within the lipase LIP protein family. Phylogenetic analysis was carried out using the newly identified lipases and hypothetical lipases from M. restricta as well as all described lipases from M. globosa (Xu et al. 2007), M. sympodialis (Gioti et al. 2013), M. furfur and M. pachydermatis (Shibata et al. 2006) (Fig. 1). From the lipase sequence alignment, a conserved serine residue within the GXSXG motif typical for lipases (Derewenda and Derewenda 1991; Derewenda 1994) was apparent (Fig. 1). The phylogenetic tree clearly divides the lipases within the genus into four distinct clades, with two predominant clades encompassing the majority of the lipases, one representing lipase family class 3 (containing MrLip1, MrLip2 and MrLip4) and the other representing the LIP lipase family (containing MrLip3 and MrLip0571). Within the lipase family LIP clade, the hypothetical M. restricta lipase MrLip0571 formed a subclade with MfLip3, MGL 3507 and an unnamed protein product from M. sympodialis (CCV ). MGL 3507 is reported to have an extra region rich in serine, lysine and glycine, approximately 110 amino acids longer at the C-terminus than the rest of the LIP class proteins that was speculated to tether the lipase to the cell wall (Boekhout et al. 2010). An associated yet distinct third clade of family class 3 lipases (consisting of MrLip3873, MGL 0279 and a hypothetical lipase from M. sympodialis (CCV )) is clearly separated from the main lipase family class 3 clade (with strong bootstrap statistical support) suggesting that these lipase sequences, although evolutionarily similar, are divergent. The M. globosa lipase MGL 0279 is unique within the M. globosa lipases due to its lengthy nucleotide sequence of 3729 bp and the fact that it is not secreted, suggesting that lipases within this clade may have a specific role in the lipid metabolism, compared to other family 3 lipases. As the M. restricta lipase fragment MrLip3873 is classified within the same clade as MGL 0279 and the M. sympodialis hypothetical lipase fragment (CCV ), we hypothesize that the complete sequence length of MrLip3873 should be comparable and that this lipase is also not secreted. The fourth clade in the phylogenetic tree, consisting of MrLip2630, M. globosa lipase MGL 1769 and an unnamed protein product from M. sympodialis, represents a possible new family of lipases as the protein sequences are not evidently classified to either the lipase family 3 or the LIP family. However, a Pfam search with the MGL 1769 amino-acid sequence indicates that MGL 1769, as well as MrLip2630, is most similar to lipase family 3 but with a low E- valueof5.3e 8. Cloning and heterologous expression of MrLip1, MrLip2 and MrLip3 In silico analysis of the complete sequences of MrLip1, MrLip2, MrLip3 and MrLip4 with Signal P 4.0 (Petersen et al. 2011) identified signal peptides at the N-terminus of all four lipases (Table 3) consistent with secreted lipases from M. globosa (Xu et al. 2007), M. sympodialis (Gioti et al. 2013), M. furfur (Brunke and Hube 2006) and M. pachydermatis (Shibata et al. 2006). To prove the hypothesis that the protein sequences represent secreted lipases, the corresponding nucleotide sequences of MrLip1, MrLip2 and MrLip3 were cloned in the E. coli P. pastoris shuttle vector ppic3.5 for recombinant expression. As the ppic3.5 vector does not provide a sequence for a signal peptide and is used for intracellular expressions of proteins, secretion of the lipases from P. pastoris confirmed that the proteins were secreted due to the presence of their N-terminal signal peptides. Additionally, the 3 primers used for amplification of the genes included the codon sequence for six histidine residues, enabling an easier purification and concentration of the proteins via Ni 2+ - NTA affinity chromatography. The elution fractions for MrLip1

5 Sommer et al. 5 Figure 1. Phylogenetic tree of all known lipases from M. restricta, M. globosa, M. sympodialis, M. furfur and M. pachydermatis based on maximum likelihood. In this paper, described lipases are shown in bold. Next to the phylogenetic tree an alignment shows the conserved amino acid motif GXSXG typical for lipases. and MrLip3 showed a distinct band of approximately 34 and 52 kda in agreement with their calculated molecular weight (Fig. 2). However, purified MrLip2 had an apparent molecular mass of 49 kda, which is larger than the deducted 35 kda. Most likely, the higher molecular mass is caused by glycosylation, a common post-translational modification of P. pastoris that was previously reported for the expression of MgLip1 and MgMDL2 (Xu et al. 2012, 2015). Activity of the lipases with mono-, di- and triolein An HPLC-based activity assay was used to detect the fatty acids released from the substrates mono-, di- and triolein at ph 5.5 (Fig. 2). The HPLC results demonstrated that MrLip1, MrLip2 and MrLip3 were able to generate free oleic acid from the substrates mono- and diolein; however, only MrLip3 was able to hydrolyze triolein in significant amounts (Fig. 3). Additionally, MrLip1 and MrLip2 were able to cleave the oleic acid from both the 1 and the 2 position of the glycerol backbone as 1,2 isomers of diolein were converted into oleic acid and glycerol (no monoolein was detected). Monoolein and diolein assays involving MrLip3 resulted in the detection of small amounts of oleic acid; interestingly, most substrate was converted to triolein. A similar result was observed for MrLip1 and MrLip2; monoolein was converted to diolein affirming that all three M. restricta lipases were able

6 6 FEMS Yeast Research, 2015, Vol. 15, No. 7 Figure 2. Schematic overview showing the activity of MrLip1, MrLip2 and MrLip3 with monoolein, diolein and triolein. Additionally, the SDS polyacrylamide gel electrophoresis of MrLip1, MrLip2 and MrLip3 is shown with the molecular weight markers indicated. Figure 3. MrLip1, MrLip2 and MrLip3 were active against the substrates mono- (A) and diolein (B), but only MrLip3 was active against triolein (C). After 24 h incubation at 30 C, the assays were analyzed by HPLC and the detected products (oleic acid, monoolein, diolein and triolein) are shown in %. The pure substrate served as control. to perform the reverse reaction to build diacyl- and triacylglycerols (Fig. 3). The substrate specificity observed for MrLip1 and MrLip2 is comparable to previous reports of the limited substrate range of the M. globosa lipase MgLip1 and MgMDL2 (DeAngelis et al. 2007; Xu et al. 2015). Substrate limitations observed for MgLip1 result from a novel lid in loop conformation and two bulky hydrophobic residues adjacent to the catalytic site and the N-terminal hinge region of the lid which hamper triacylglycerols from entering the enzyme active site (Xu et al. 2012). As the concentration of di- and monoacylglycerols is comparatively low in newly expressed and accumulated human sebum compared to triacylglycerols (Stewart et al. 1978; Wheatley 1986; Dawson 2007), the inability of MrLip1, MrLip2, MgLip1 and MgMDL2 to degrade triacylglycerols raises a question regarding theroleofmalassezia family class 3 lipases in the pathology of D/SD. Our substrate activity results showed that the family LIP lipase MrLip3 can convert triacylglycerols into glycerol and free fatty acids and suggest that similar Malassezia lipases classified within the family LIP clade are virulence factors associated with the breakdown of triaclyglycerols during the onset of D/SD. In turn, family class 3 lipases may play an active role in degrading mono- and diacylglycerols liberated from triacylglycerols by Malassezia family LIP lipases (or other commensal skin associated microbes).

7 Sommer et al. 7 Our substrate specificity assays were performed at ph 5.5 and 30 C to approximate skin conditions likely to occur on a human head. The physiological human skin temperature ranges 28 C 30 C depending on air temperature and the time spent in the environment (Plasencia, Norlen and Bagatolli 2007). The natural skin ph is on average below 5 (Lambers et al. 2006). Similar conditions were also used for the activity assays of MgLip1 (DeAngelis et al. 2007). Experimental evidence of the expression of lipases from Malassezia spp. on human skin is limited. The M. restricta family class 3 lipases, MrLip1 and MrLip0242 have been confirmed by RT-PCR to be expressed on the human scalp of patients with SD, and MrLip0242 was the most frequently expressed lipase in this disease pattern (Lee et al. 2013). Xu et al. (2007) demonstrated that several family class 3 and family LIP lipases were also expressed from M. globosa on a healthy human scalp. Further in vitro lipase expression efforts are needed to assess the substrate specificity of the numerous lipases encoded within the Malassezia spp. genomes, and additional in vivo studies using RT-PCR and representative primers for the full range of Malassezia lipases are required from both healthy and D/SD damaged skin to further elucidate the etiopathological role these lipases play in associated skin disorders. In summary, in this study we described four complete lipase sequences from M. restricta as well as four hypothetical lipase fragments. We demonstrated that the enzymes were lipases, by confirmation of related sequences, the presence of the lipase active site consensus sequence GXSXG and by cloning and expression of three lipases in P. pastoris. Furthermore, we showed that lipases from the lipase family 3 are only able to digest monoand diacylglycerols, while a lipase from the LIP family was also able to digest triacylglycerols. As Malassezia lipid metabolism is an associated virulence factor of D and SD, our research provides further insights into the substrate specificity of lipases from M. restricta. SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online. ACKNOWLEDGEMENTS The authors thank Kate McQuillan (Nautilus) for technical assistance and gratefully acknowledge financial support from Natural Sciences and Engineering Council of Canada (NSERC), the Canada Research Chair Program, the University of Prince Edward Island, the Atlantic Innovation Fund of Atlantic Canada Opportunities Agency, Nautilus Biosciences Canada Inc. and the Jeanne and Jean-Louis Lévesque Foundation. Conflict of interest. None declared. REFERENCES Altschul SF, Madden TL, Schaffer AA, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25: Ashbee HR, Evans EG. Immunology of diseases associated with Malassezia species. Clin Microbiol Rev 2002;15: Batra R, Boekhout T, Gueho E, et al. Malassezia Baillon, emerging clinical yeasts. FEMS Yeast Res 2005;5: Boekhout T, Guého-Kellermann E, Mayser P, et al. Malassezia and the Skin: Science and Clinical Practice. Springer Science & Business Media. New York:Springer-Verlag, Brunke S, Hube B. MfLIP1, a gene encoding an extracellular lipase of the lipid-dependent fungus Malassezia furfur. Microbiology 2006;152: Cafarchia C, Otranto D. The pathogenesis of Malassezia yeasts. Parassitologia 2008;50:65 7. Dawson TL, Jr. Malassezia globosa and restricta: breakthrough understanding of the etiology and treatment of dandruff and seborrheic dermatitis through whole-genome analysis. In: The Journal of Investigative Dermatology Symposium Proceedings / the Society for Investigative Dermatology, Inc [and] European Society for Dermatological Research, Vol. 12, 2007, pp DeAngelis YM, Saunders CW, Johnstone KR, et al. Isolation and expression of a Malassezia globosa lipase gene, LIP1. J Investig Dermatol 2007;127: Derewenda ZS. Structure and function of lipases. Adv Protein Chem 1994;45:1 52. Derewenda ZS, Derewenda U. Relationships among serine hydrolases: evidence for a common structural motif in triacylglyceride lipases and esterases. Biochem Cell Biol 1991;69: Downing DT, Stewart ME, Wertz PW, et al. Skin lipids. Comp Biochem Physiol B 1983;76: Finn RD, Bateman A, Clements J, et al. Pfam: the protein families database. Nucleic Acids Res 2014;42:D Gasteiger E, Hoogland C, Gattiker A. The Proteomics Protocols Handbook. In: Walker JM, et al. (ed). Protein Identification and Analysis Tools on the ExPASy Server. Totowa: Humana Press, 2005, Gioti A, Nystedt B, Li W, et al. Genomic insights into the atopic eczema-associated skin commensal yeast Malassezia sympodialis. mbio 2013;4:e Gueho E, Boekhout T, Ashbee HR, et al. The role of Malassezia species in the ecology of human skin and as pathogens. Med Mycol 1998;36(Suppl 1): Guillot J, Hadina S, Gueho E. The genus Malassezia: old facts and new concepts. Parassitologia 2008;50:77 9. Gupta AK, Batra R, Bluhm R, et al. Skin diseases associated with Malassezia species. J Am Acad Dermatol 2004;51: Gupta AK, Kohli Y, Summerbell RC, et al. Quantitative culture of Malassezia species from different body sites of individuals with or without dermatoses. Med Mycol 2001;39: Hay RJ, Graham-Brown RA. Dandruff and seborrhoeic dermatitis: causes and management. Clin Exp Dermatol 1997;22:3 6. Juntachai W, Oura T, Kajiwara S. Purification and characterization of a secretory lipolytic enzyme, MgLIP2, from Malassezia globosa. Microbiology 2011;157: Kieffer M, Bergbrant IM, Faergemann J, et al. Immune reactions to Pityrosporum ovale in adult patients with atopic and seborrheic dermatitis. J Am Acad Dermatol 1990;22: Lambers H, Piessens S, Bloem A, et al. Natural skin surface ph is on average below 5, which is beneficial for its resident flora. Int J Cosmet Sci 2006;28: Lee YW, Lee SY, Lee Y, et al. Evaluation of expression of lipases and phospholipases of Malassezia restricta in patients with seborrheic dermatitis. Ann Dermatol 2013;25: McWilliam H, Li W, Uludag M, et al. Analysis Tool Web Services from the EMBL-EBI. Nucleic Acids Res 2013;41:W Midgley G. The lipophilic yeasts: state of the art and prospects. Med Mycol 2000;38(Suppl 1):9 16. Petersen TN, Brunak S, von Heijne G, et al. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods 2011;8:785 6.

8 8 FEMS Yeast Research, 2015, Vol. 15, No. 7 Plasencia I, Norlen L, Bagatolli L. Direct visualization of lipid domains in human skin stratum corneum s lipid membranes: effect of ph and temperature. Biophys J 2007;93: Ro BI, Dawson TL. The role of sebaceous gland activity and scalp microfloral metabolism in the etiology of seborrheic dermatitis and dandruff. In: The Journal of Investigative Dermatology Symposium Proceedings / the Society for Investigative Dermatology, Inc [and] European Society for Dermatological Research, Vol. 10, 2005 pp Shibata N, Okanuma N, Hirai K, et al. Isolation, characterization and molecular cloning of a lipolytic enzyme secreted from Malassezia pachydermatis. FEMS Microbiol Lett 2006;256: Stewart ME, Downing DT, Pochi PE, et al. The fatty acids of human sebaceous gland phosphatidylcholine. Biochim Biophys Acta 1978;529: Strauss JS, Pochi PE. Histology, histochemistry, and electron microscopy of sebaceous glands in man. Handbuch der Haut-und Geschlechtskrankheiten 1968;1:1. Tamura K, Stecher G, Peterson D, et al. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol 2013;30: Wheatley V. The chemistry of sebum. Physiol Pathophysiol Skin 1986;9: Whelan S, Goldman N. A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approach. Mol Biol Evol 2001;18: Xu H, Lan D, Yang B, et al. Biochemical properties and structure analysis of a DAG-Like lipase from Malassezia globosa. Int J Mol Sci 2015;16: Xu J, Saunders CW, Hu P, et al. Dandruff-associated Malassezia genomes reveal convergent and divergent virulence traits shared with plant and human fungal pathogens. P Natl Acad Sci USA 2007;104: Xu T, Liu L, Hou S, et al. Crystal structure of a mono- and diacylglycerol lipase from Malassezia globosa reveals a novel lid conformation and insights into the substrate specificity. J Struct Biol 2012;178:363 9.