Applied Microbiology and Biotechnology

, Volume 89, Issue 2, pp 375–385

Heterologous expression of a thermophilic esterase in Kluyveromyces yeasts


  • Saul Nitsche Rocha
    • Department of Chemical EngineeringUniversity of São Paulo
    • Universidade PositivoRua Pedro Viriato Parigot de Souza
  • José Abrahão-Neto
    • School of Pharmaceutical SciencesUniversity of São Paulo
  • María Esperanza Cerdán
    • Departamento de Bioloxía Celular e Molecular, Facultade de CienciasUniversidade da Coruña
  • Andreas Karoly Gombert
    • Department of Chemical EngineeringUniversity of São Paulo
    • Departamento de Bioloxía Celular e Molecular, Facultade de CienciasUniversidade da Coruña
Applied Microbial and Cell Physiology

DOI: 10.1007/s00253-010-2869-8

Cite this article as:
Rocha, S.N., Abrahão-Neto, J., Cerdán, M.E. et al. Appl Microbiol Biotechnol (2011) 89: 375. doi:10.1007/s00253-010-2869-8


In the present work, a thermophilic esterase from Thermus thermophilus HB27 was cloned into Kluyveromyces marxianus and into Kluyveromyces lactis using two different expression systems, yielding four recombinant strains. K. lactis showed the highest esterase expression levels (294 units per gram dry cell weight, with 65% of cell-bound enzyme) using an episomal system with the PGK promoter and terminator from Saccharomyces cerevisiae combined with the K. lactis k1 secretion signal. K. marxianus showed higher secretion efficiency of the heterologous esterase (56.9 units per gram dry cell weight, with 34% of cell-bound enzyme) than K. lactis. Hydrolytic activities for the heterologous esterases were maximum at pH values between 8.0 and 9.0 for both yeast species and at temperatures of 50 °C and 45 °C for K. marxianus and K. lactis, respectively. When compared to previously published data on this same esterase produced in the original host or in S. cerevisiae, our results indicate that Kluyveromyces yeasts can be considered good hosts for the heterologous secretion of thermophilic esterases, which have a potential application in biodiesel production or in resolving racemates.


Thermophilic esteraseKluyveromyces marxianusKluyveromyces lactisHeterologous expression


Lipases and esterases (EC 3.1.1.x) are enzymes catalyzing reactions of ester bonds. They may either hydrolyze such bonds or form new ones, depending on the enzyme and the conditions. Among other potential applications, these enzymes might be employed industrially in producing biodiesel, in resolving racemates, and upgrading edible vegetable oils (Du et al. 2008; Bornscheuer 2002; Gupta et al. 2003). Thermophilic lipases/esterases would also have the particular benefit of greater stability under harsh environmental conditions, which are not infrequent in industrial reality (Demirjian et al. 2001).

There has been much debate in the literature about the distinction between lipases and esterases. However, in general terms, esterases (EC hydrolyze solutions of short water-soluble acyl chain esters, whereas lipases (EC act on long-chain water-insoluble triacylglycerols (Chahinian and Sarda 2009). Fifty-five entries can be found under EC number and 80 under in the BRENDA enzyme database (Chang et al. 2009).

The expression of thermophilic enzymes in mesophilic hosts has advantages. First, the mild environmental conditions necessary for growth are usually much simpler (in terms of both nutrient requirements and lower temperature) than the extreme conditions required by thermophilic organisms. Second, as in other heterologous systems, it is possible to drive secretion of the enzyme into the culture supernatant, facilitating downstream operations. Third, most mesophilic expression hosts have already been granted GRAS status. Within this perspective, yeasts belonging to the Kluyveromyces taxon are good platforms for such purposes. There have been several successful examples of heterologous protein expression in K. lactis, and a few in K. marxianus (van Ooyen et al. 2006; Rocha et al. 2010).

Müller et al. (1998) reported the over-expression of a lipase from Thermomyces lanuginosus in K. lactis, using the pKD1 plasmid and the strong homologous LAC4 promoter. Although not reported quantitatively, the lipase titer obtained in K. lactis was higher than the corresponding value in Saccharomyces cerevisiae, but lower than in Schizosaccharomyces pombe, Yarrowia lipolytica, and Hansenula polymorpha. For thermophilic enzymes, Walsh and colleagues reported the heterologous production of xylanases from Dictyoglomus thermophilum and from Thermotoga maritima in K. lactis, using the same plasmid and promoter indicated above (Walsh and Bergquist 1997; Walsh et al. 1998). Hong et al. (2007) reported the simultaneous over-expression of thermostable endo-β-1,4-glucanase, cellobiohydrolase, and β-glucosidase in K. marxianus, yielding a strain capable of growing in synthetic medium containing cellobiose or carboxymethylcellulose as the single carbon source and producing 43.4 g/L ethanol from 10% cellobiose.

Here, we present the results of over-expressing the gene encoding an esterase (EST) from Thermus thermophilus in the yeasts K. marxianus and K. lactis. Two different plasmids and regulatory sequences were used to over-express the T. thermophilus EST gene in both species, yielding a total of four heterologous expression systems. This strategy allowed K. marxianus to be evaluated as an expression host for thermophilic enzymes. This is of industrial interest, since this yeast is capable of growing at temperatures above 40 °C, in contrast to most yeasts (Fonseca et al. 2008). Furthermore, our study allowed K. marxianus to be compared directly with its close relative K. lactis, which has been more thoroughly investigated (van Ooyen et al. 2006). Finally, we also evaluated the EST produced in Kluyveromyces strains in terms of cellular location, glycosylation and biochemical properties, and we compared these data with those previously reported for the same protein produced in the homologous host T. thermophilus and in S. cerevisiae (Fuciños et al. 2005; López-López et al. 2010).

Materials and methods

Strains and media

Escherichia coli strain DH10B (Invitrogen, USA) was used to construct and propagate the plasmids. Kluyveromyces marxianus CBS 6556 was purchased from Centraal Bureau voor Schimmelcultures (Utrecht, The Netherlands). Kluyveromyces lactis PM5-3C (MATa uraA Rag+) was kindly provided by Dr. Micheline Wesolowski-Louvel (Lyon, France). S. cerevisiae BY4742 (MATa his3D1 leu2D0 lys2D0 ura3D0; Brachmann et al. 1998) was provided by Euroscarf (Frankfurt, Germany). The strain K. marxianus SLC33 is an ura3 mutant from K. marxianus CBS 6556, obtained in previous work by selection on 5-FOA-containing plates; ura-phenotype stability was certified for up to 120 generations (Rocha et al. 2010).

E. coli strains were grown on LB (1% bacto-tryptone, 0.5% yeast extract, 0.5% NaCl, 0.1% glucose) liquid or solid (2% bacto-agar) media at 37 °C. Ampicillin (40 mg/L final concentration) was added for plasmid selection. To select yeast transformants exhibiting URA3 expression, complete medium (CM; Zitomer and Hall 1976) lacking uracyl (CM-URA) was used. The EST-expressing yeasts were initially screened in YPHSM + SUC medium, which contains: peptone (80 g L−1), yeast extract (10 g L−1), glycerol (30 g L−1), and sucrose (10 g L−1). For the EST expression studies, strains were grown on the defined medium (DM) described by Verduyn et al. (1992), and subsequently on YPS complex medium (1% yeast extract, 2% bacto-peptone, 2% sucrose).

Isolation and manipulation of nucleic acids

Plasmids manipulation and yeast transformation was performed according to the procedures described in Rocha et al. (2010). PCR reactions were carried out in a final volume of 25 μL with 20 ng of DNA template, 30 pmol of each primer (Table 1), 5 nmol (for the EST-coding gene) or 1.75 nmol (for the INU1 cassette) of each dNTP, 50 mM KCl, 2.5 mM MgCl2, 1 μL of DMSO (only for EST), 1 U (EST) or 0.25 U (INU1) of Taq DNA polymerase (Roche Diagnostics, Mannheim, Germany) in 10 mM Tris–HCl buffer, pH 8.3. For DNA amplification, initial denaturation was at 95 °C (EST) or 94 °C (INU1), followed by 30 cycles of: denaturation at 95 °C (EST) or 94 °C (INU1) for 30 s, annealing at 55 °C for 30 s (EST) or 53 °C for 60 s (INU1), and extension at 72 °C for 3 min (INU1) or 1 min (EST). Final elongation was performed at 72 °C for 3 min for both. All other DNA manipulations were performed as described by Sambrook and Russel (2001).
Table 1


Primer ID





Underlined sequence added an EcoRI restriction site to the amplified fragment



Underlined sequence added an EcoRI restriction site to the amplified fragment



Underlined extremity is an added BamHI restriction site



Underlined extremity is an added BamHI restriction site



Underlined sequence is homologous to the last bases of K. marxianus INU1 secretion signal sequence



Underlined sequence is homologous to the first bases of K. marxianus INU1 terminator

List of primers used for DNA constructions

Plasmid construction and yeast transformation

The episomal expression vector pSPEST (Fig. 1a) was constructed as follows. The EST CDS, Locus tag TTC0904 in the Kyoto Encyclopedia of Genes and Genomes database (, without the putative native secretion signal (i.e., from the 17th codon), was amplified by PCR from the chromosomal DNA of T. thermophilus HB27, obtained as described by López-López et al. (2010) using the primers PSPESTF and PSPESTR (Table 1). The amplified 990-bp fragment was subcloned into pSPGK1 by the EcoRI site, locating the CDS, and fused to the k1 (killer toxin alpha subunit prepro-sequence) secretion signal sequence between the ScPGK promoter and terminator. The resulting plasmid (pSPEST, Fig. 1a) was used to transform K. lactis PM5-3C and K. marxianus SLC33 cells. Transformants were grown in CM-URA agar plates and those expressing the highest amounts of EST were selected by screening the cells in YPHSM + SUC medium and measuring extracellular EST activity per milligram of biomass dry weight. The resulting constructions and strains (Km-EP and Kl-EP) are detailed in Table 2 and Fig. 1.
Fig. 1

Expression systems constructed in this work a pSPEST; b pINESTi. p promoter; t terminator; SS secretion signal (k1) sequence

Table 2

Heterologous EST constructions

Recombinant strain

Host organism

Plasmid name



Signal sequence



K. marxianus







K. marxianus







K. lactis







K. lactis






List of the four recombinant strains constructed in this work and their details

The constructions using INU1 for EST expression started with amplification of the INU1 cassette (the promoter, ORF, and terminator of the INU1 gene) from the chromosomal DNA of K. marxianus CBS 6556 using the primers PINU1Fb and PINU1Rb (Table 1). Both primers have a BamHI restriction site in their 5′ extremities. After amplification, the INU1 cassette was ligated to the pMBL1-T vector for propagation. Thereafter, the insert was excised by BamHI restriction and cloned into the yeast-E. coli shuttle vector Yeplac195, and the resulting construct was linearized with Tth111I. In parallel, the T. thermophilus HB27 EST CDS, from the 17th codon, was amplified by PCR from the bacterial chromosomal DNA using the primers PINUESTF and PINUESTR (Table 1). These two primers were capable of adding extremities to the amplified gene, which were homologous to the INU1 secretion signal and terminator sequences. The resulting fragment was cloned into pMBL1-T for propagation. The linearized Yeplac195 and the amplified EST carrying homologous extremities were fused by homologous recombination in S. cerevisiae BY4742 transformed with an equal proportion of both plasmid and insert (500 ng) by the lithium acetate method (Ito et al. 1983). The construction was purified from S. cerevisiae cells and, after propagation in E. coli, digested with BamHI, which released the cassette for EST expression under control of the INU1 promoter. This fragment was subcloned into the BamHI site of pNADFL11, resulting in the integrative construction shown in Fig. 1b.

The expression vectors were used to transform K. marxianus SLC33 and K. lactis PM5-3C by the lithium acetate method (Ito et al. 1983). After transformation of the yeast cells with each construction, a first qualitative selection was performed using a top-agar technique, as follows. Agarose was solubilized in 50 mM Tris–HCl buffer containing 40 mM CaCl2. Then, p-nitrophenyl laurate was added to the chilled agarose (50 °C, to avoid degradation) to a final concentration of 2.5 mM. The agarose was poured over the colonies grown on selective medium (CM-URA). After 1-h incubation at 65 °C, the three colonies showing the largest yellowish halos were selected for the screening on liquid medium. An exception occurred with the construct Kl-IN, in which case only one transformant was obtained, in spite of several transformation attempts. The selected strains were grown in liquid YPHSM + SUC medium for 36 h and EST activity was measured. The clones that exhibited the highest activity for each construct were selected for further studies (the screening results are shown on part A of the Electronic supplementary material (ESM)). The resulting constructions and strains are detailed in Table 2 and Fig. 1. Genetic stability of the strains was verified as described in part B of the ESM.

Determination of esterase activity

Lipolytic activity was determined using p-nitrophenyl laurate as substrate, according to the method described by Fuciños et al. (2005). A mixture of 800 μL of 50 mM Tris–HCl buffer (pH 8 at 65 °C) containing 40 mM CaCl2 and 100 μL of a 25 mM p-nitrophenyl laurate solution (dissolved in ethanol) was incubated in a test tube for 10 min at 65 °C for temperature equilibration. The sample (100 μL) was added, and the hydrolysis reaction proceeded for 10 min at 65 °C and was stopped by adding 250 μL of 2 M Na2CO3. The mixture was kept on ice for 15 min before centrifugation (2,500×g for 15 min). The absorbance of the supernatant was read at 400 nm in a Thermo Genesys 20 Spectrophotometer (Waltham, USA). Under these conditions, the molar extinction coefficient was 17.215 mM−1 cm−1 (Fuciños et al. 2005). One activity unit was defined as the amount of enzyme necessary to release 1 μmol of product per minute under the above conditions.

Analysis of esterase production in submerged cultivations

The culture conditions described by Bergkamp et al. (1993), which resulted in the efficient secretion of a heterologous alpha-galactosidase in K. marxianus, were adopted with slight modifications to evaluate EST expression and secretion in K. marxianus CBS 6556 and K. lactis CBS 2359. Transformants, and untransformed cells as negative controls, were grown at 30 °C in baffled 500 mL shake flasks closed with cotton on an orbital shaker (300 rpm), using 100 mL of medium. First, the cells were grown for 24 h in the DM (using 2% glucose as carbon source). Second, these cells were diluted 1:10 into fresh YPS medium and grown for another 48 h. Subsequently, the heterologous enzyme present in the supernatant, the fraction in the cell wall (retained inside the periplasmic space), and the cell-bound fractions were separated according to the method described by Rouwenhorst et al. (1988). EST activity was determined spectrophotometrically in the three fractions. The protein contents of all fractions were measured by the Bradford method (product B6916, Sigma, USA), with bovine serum albumin as standard, according to the manufacturer’s instructions. The dry cell weight of 5 mL samples of each culture flask was determined using 0.45 μm membrane filters and a microwave oven (180 W, 15 min; Olsson and Nielsen 1997). Results for esterase activity are shown on Tables 3 and 4. Cultivations were carried out in duplicate and esterase activity was measured in two samples of each culture. Thus, values presented are the average and standard deviation of four experimental data points.
Table 3

Esterase activities


Final biomassb

EST activityc


Cell wall















11.6 ± 0.3

5.1 ± 1.1

263 ± 41

22.5 ± 3.5

7.6 ± 0.9

97.8 ± 1.4

8.4 ± 0.4

0.2 ± 0.0

217.4 ± 4.0

18.5 ± 0.7

49.4 ± 4.6


11.4 ± 0.1

6.1 ± 0.3

328 ± 11

28.5 ± 0.7

17.2 ± 0.4

101.3 ± 4.4

8.9 ± 0.4

0.6 ± 0.1

222.0 ± 13.9

19.5 ± 0.7

56.9 ± 1.0


9.9 ± 0.1

11.8 ± 3.5

345 ± 17

35.0 ± 1.4

98.5 ± 6.4

661.2 ± 85.4

62.5 ± 9.2

2.4 ± 0.1

1886.1 ± 95.2

192 ± 8

294 ± 0


9.6 ± 0.1

9.0 ± 0.8

287 ± 12

30.0 ± 1.4

101 ± 7

594 ± 0

62.5 ± 0.7

1.6 ± 0.1

1386.2 ± 267.3

144 ± 28

237 ± 25

Heterologous EST production in the supernatant, cell wall and cell-bound fractions of different K. marxianus and K. lactis recombinant strains

aSee “Materials and methods” section, Table 2 and Fig. 1 for a description of the genetic constructions

bSee “Materials and methods” section for a description of the cultivation conditions

cSubcellular distribution according to Rouwenhorst et al. (1988). Values presented are the average and standard deviation of four experimental data points: cultivations were carried out in duplicate and two samples of each culture were analyzed independently

Table 4

Esterase distribution


Esterase activity per g DW in each fraction relative to the totalb,c


Cell wall



45% ± 3%

17% ± 1%

38% ± 2%


50% ± 0%

16% ± 1%

34% ± 1%


12% ± 0%

23% ± 3%

65% ± 3%


13% ± 2%

27% ± 3%

61% ± 5%

Subcellular distribution of heterologous EST in different K. marxianus and K. lactis recombinant strains

aSee “Materials and methods” section, Table 2 and Fig. 1 for a description of the genetic constructions

bSee “Materials and methods” section for a description of the cultivation conditions

cSubcellular distribution according to Rouwenhorst et al. (1988). Values presented are the average and standard deviation of four experimental data points: cultivations were carried out in duplicate and two samples of each culture were analyzed independently

Esterase glycosylation

The EST expressed and secreted from the four different constructions of K. marxianus and K. lactis (Table 2) was analyzed with respect to the degree of glycosylation. For this purpose, the supernatant and cell wall fractions were concentrated in 10 kDa cut-off ultrafilters (Millipore, USA). To de-glycosylate the EST, 15 μg of the protein were denaturated by incubation for 30 min at 100 °C in a 10 μL final volume reaction containing 0.5% SDS and 40 mM 2-mercaptoethanol. After cooling, Tergitol NP-40, sodium phosphate buffer pH 7.5, and PNGase F (New England Biolabs, USA) were added to final concentrations of 1%, 50 mM, and 5 U/μL, respectively, in a 20 μL final reaction volume. The mixture was incubated for 2 h at 37 °C. The migration patterns of glycosylated and deglycosylated forms of EST were compared on a 10% SDS-PAGE gel as described by Laemmli (1970). To detect bands exhibiting specific EST activity, the procedure described in Fuciños et al. (2005) was followed. After electrophoresis, the gels were renatured by washing them for 20 min at 65 °C in a solution of 20 mM Tris–HCl buffer (pH 8.0) containing 0.5% (w/v) Triton X-100. EST activity was detected with α-naphthyl acetate as described previously (Schmidt-Dannert et al. 1996), except that the reaction was carried out at 65 °C. After incubation, active ESTs appeared as reddish bands. Following the EST-specific reaction, the gels were subjected to a cycle of staining with Coomassie Blue R-250 in order to stain the other proteins present, as described by Laemmli (1970).

Biochemical characterization of esterase

The enzyme synthesized by the transformant of each species, either K. marxianus or K. lactis, that exhibited the highest biomass specific activity in both fractions (supernatant and cell wall) was characterized biochemically: the pH and temperature for optimal activity were determined, as well as thermal inactivation. EST activity was measured at 65 °C and at the following pH values: 3.0, 4.0 and 5.0 (50 mM acetate buffer containing 40 mM CaCl2); 6.0, 7.0 and 8.0 (50 mM HEPES buffer containing 40 mM CaCl2); 8.6, 9.0, 9.5 and 10.2 (50 mM Tris buffer containing 40 mM CaCl2). Thermal inactivation was assessed by incubating for 4 h at 80 °C, 85 °C and 90 °C. Samples were taken at 0, 1, 2, and 4 h and the enzyme activity was measured immediately (at 65 °C). pH stability of the heterologous esterases was determined by incubating for 2 h at 65 °C in the buffers mentioned above (from pH 3.0 to 10.2). Activity was always measured at pH 8.0. Finally, the temperature for optimum enzyme activity was determined using the standard procedure (at pH 8.0) at different temperatures (40 °C, 45 °C, 50 °C, 55 °C, 60 °C, 65 °C, 75 °C).


Esterase expression levels and subcellular distribution

Table 3 shows that all the expression systems described in this work were able to express and partially secrete the heterologous EST from T. thermophilus. K. marxianus and K. lactis cells, before transformation with either the episomal plasmid or the integrative cassette, were grown under the same conditions as those employed for the recombinant strains and esterase activity was measured in the three selected fractions (supernatant, cell wall, and cell-bound). Since no significant activity was detected in the host strains in comparison with the recombinant strains and using the same enzyme assay (results not shown), it may be concluded that the esterase activity detected in the three fractions of the recombinant strains are indeed due to the expression of the cloned gene from T. thermophilus.

When K. marxianus was used as host, transformants Km-EP and Km-IN grew to final cell concentrations around 11.5 g DW L−1, corresponding to a normal biomass yield on the sugar substrate, since the initial carbon source (sucrose) concentration was around 20 g L−1 (Bellaver et al. 2004; Fonseca et al. 2007). Both recombinant strains of K. lactis, which gave slightly lower biomass yield than K. marxianus, performed similarly (Table 3; Kiers et al. 1998). Comparison of the two recombinant strains, integrative and episomal, of K. marxianus shows that both subcellular distribution and EST expression levels were very similar (Tables 3 and 4). Fully secreted EST (into the culture medium) was in the range 260 (Km-EP) - 325 (Km-IN) U/L. For K. lactis, subcellular distribution and expression levels were also quite similar in both recombinant strains, integrative and episomal. Fully secreted EST was in the range 290 (Kl-IN)–345 (Kl-EP) U/L. To compare, although culture media and conditions are different, in a previous study (López-López et al. 2010) in which the same EST was expressed in S. cerevisiae using the episomal system YEpFLAG-1, higher total EST titers were obtained (about 9,000 U/L after 72 h of culture) than those with the two episomal Kluyveromyces systems in this work (about 600 and 3,000 U/L for Km-EP and Kl-EP, respectively). Although plasmid stability was slightly higher in Km-EP (see ESM), total EST production was notably higher in Kl-EP. However, in S. cerevisiae, more than 80% of the EST was trapped in the periplasmic space, whereas in K. marxianus this value was around 17% and in K. lactis around 25%. Also, percentage of EST secreted into the culture medium reached the highest level in the case of K. marxianus (about 50%) compared with the other two yeasts (about 12%).

Thus, although in the work with S. cerevisiae only one genetic construction was employed, the results obtained here indicate that the expression levels and subcellular distribution of EST are highly determined by the host strain employed, and not only by the genetic construction (promoter, secretion signal and terminator). Further work with more yeast strains and more genetic constructions is needed to verify whether this is a general observation for the expression of this esterase.

Biochemical characterization of the heterologous esterase

The ESTs expressed heterologously in K. lactis and in K. marxianus were biochemically characterized in terms of pH and temperature effects on activity (Fig. 2), thermal inactivation (Fig. 3), stability at different pH values, and molecular weight and glycosylation (using denaturing gels; Fig. 4).
Fig. 2

Esterase activity at different pH values and temperatures. a EST activity was measured after enzyme incubation at 65 °C in the following buffers: acetate 3.0, 4.0, 5.0 (50 mM acetate buffer containing 40 mM CaCl2); 6.0, 7.0, 8.0 (50 mM HEPES buffer containing 40 mM CaCl2); 8.6, 9.0, 9.5, 10.2 (50 mM Tris buffer containing 40 mM CaCl2). b EST activity was determined under the standard conditions described in the Methods section at the following temperatures: 40 °C, 45 °C, 50 °C, 55 °C, 60 °C, 65 °C, 75 °C. Each point is the mean of a duplicate assay. Km supernatant and Km cell wall: EST produced by Km-IN construction. Kl supernatant and Kl cell wall: EST produced by Kl-EP construction. Both clones were cultivated as described in the “Materials and methods” section
Fig. 3

Esterase stability at different temperatures, Time course of residual EST activity measured after incubation at a 80 °C; b 85 °C; and c 90 °C. Each point is the mean of a duplicate assay. Km supernatant and Km cell wall: EST produced by Km-IN construction. Kl supernatant and Kl cell wall: EST produced by Kl-EP construction. Both clones were cultivated as described in the “Materials and methods” section
Fig. 4

Glycosylation pattern of the esterase expressed in K. marxianus and K. lactis, aK. marxianus cell wall EST; bK. lactis cell wall EST. Glyc glycosylated enzyme. Deglyc deglycosylated enzyme with PNGase F, as described in the “Materials and methods” section. Construction codes are listed in Table 2. Gels were first stained for EST activity with alpha-naphtyl acetate and subsequently with Coomassie blue, as described in the “Materials and methods” section

The pH for maximal activity lies between 8 and 9 for all cases investigated (Fig. 2a), except for the cell wall fraction of K. marxianus EST (in this case, the optimum pH was 9.6). In the work of Fuciños et al. (2005), two native proteins (34 kDa and 62 kDa) exhibited EST activity in T. thermophilus HB27. The authors concluded that one enzyme had maximum activity at pH between 5 and 7 and the other at pH ≥ 9. Since we over-expressed the smaller of the two esterases, our data indicate that the 34 kDa protein is the one presenting maximum EST activity at alkaline pH values. The same enzyme expressed in S. cerevisiae presented a pH optimum between 7.5 and 8.5. When the ESTs produced by the two Kluyveromyces species studied in this work, including both the cell wall and the supernatant fractions, were incubated at 50 °C for 2 h, their activities (measured under optimal conditions) remained stable throughout the incubation period within the pH range from 4.0 to 9.0 (data not shown).

The activity of the heterologous EST was determined at temperatures ranging from 40 °C to 75 °C (Fig. 2b). The optimum temperature for catalysis was 50 °C for the enzyme expressed in K. marxianus and 45 °C for the enzyme expressed in K. lactis, in both the supernatant and the periplasmic fractions. López-López et al. (2010) showed that the same heterologous EST expressed in S. cerevisiae has an optimum temperature of 40 °C, whereas the optimum temperature for EST activity in the original host, T. thermophilus, is 80 °C according to Fuciños et al. (2005). Thus, the enzymes produced by both Kluyveromyces strains have higher optimum temperatures than the one produced by S. cerevisiae, but lower than the value for the enzyme produced naturally by T. thermophilus HB27.

Thermal inactivation of the heterologous esterases was determined by incubating the enzymes for 4 h at different temperatures (activity was always measured at 65 °C). The results show that they are relatively stable at temperatures up to 80 °C (Fig. 3). The heterologous EST from K. marxianus remained stable for at least 4 h at 80 °C, whereas during the same period the enzyme from K. lactis lost 20% and nearly 100% of its activity for the cell wall and supernatant fractions, respectively. When the enzymes were incubated at 90 °C (Fig. 3c) they showed similar profiles of activity, decreasing almost to zero within 2 h. The heterologous EST expressed in the Kluyveromyces strains showed a thermostability at 85 °C (Fig. 3b) not far on average from that of the native enzyme (half-life about 2 h), the latter measured in the presence of 1% w/v CHAPS, but lower than that produced in S. cerevisiae, which had a half-life around 4 h at the same temperature (López-López et al. 2010). The thermostability of the recombinant enzymes was measured without adding CHAPS, other detergents or stabilizers.

As the T. thermophilus 34 kDa EST has two putative sites for N-linked glycosylation, the Kluyveromyces strains could attach glycosidic chains to the protein as already observed for S. cerevisiae (López-López et al. 2010). To test this hypothesis, the heterologous ESTs expressed both in K. marxianus and K. lactis were treated with PNGaseF to remove any attached glycosidic chain. Figure 4a and b show that heterologous EST was expressed as three different active proteins in both K. marxianus and K. lactis, since three bands reacted with α-naphthyl acetate. Since the expression system was designed to produce a 34 kDa EST, we suggest that the two heavier bands correspond to ESTs glycosylated at two different levels, resulting in glycoproteins of approximately 38 and 42 kDa. This glycosylation pattern was observed in all four recombinant strains of Kluyveromyces, irrespective of species or expression system. When the heterologous ESTs investigated here were treated with PNGaseF to remove the carbohydrate moiety, the intensity of the band corresponding to the unglycosylated form was higher, and the intensities of the bands corresponding to the two glycosylated forms were lower, than the sample that was not treated with PNGaseF (Fig. 4). However, treatment with PNGaseF did not remove the glycosidic chains completely, although we tried to improve the protocol in several ways, e.g., by varying the EST denaturation time and changing the concentration of the denaturing agent (results not shown).

According to Fig. 4b, K. lactis produced a large protein of approximately 130 kDa with EST activity. This molecular weight is close to that expected for a hypothetical trimer of the most glycosylated monomeric form of EST. It was previously demonstrated that this same esterase can exist as a trimer (Fuciños et al. 2005). According to these authors, trimer disassembly into the monomeric form of the protein is temperature-dependent. This could explain why the trimer disappeared in lanes labeled Deglyc (Fig. 4b), since the samples treated with PNGaseF were boiled for 30 min, whereas those not subjected to deglycosylation were only boiled for 5 min.


We expressed a T. thermophilus HB27 esterase (EST) in K. marxianus and K. lactis using two different expression systems: one episomal and one integrative.

The heterologous proteins produced by the four recombinant yeasts (Table 2) were efficiently secreted and active. The percentage of secretion into the culture medium was higher for K. marxianus, but K. lactis expressed higher total EST levels than K. marxianus, when genetic elements (promoter, terminator) from both S. cerevisiae and K. marxianus were employed (Tables 3 and 4). It was expected that heterologous protein expression would be higher in K. lactis than K. marxianus using expression systems based on S. cerevisiae genetic elements (mainly the promoter), since this had been reported previously for other proteins such as α-galactosidase and glucose oxidase (Bergkamp et al. 1993; Rocha et al. 2010). Nevertheless, it was surprising that the EST titers obtained in K. lactis were higher than those in K. marxianus when the K. marxianus INU1-based expression system was employed, considering that the opposite was found when glucose oxidase was expressed under the same systems (Rocha et al. 2010).

The data presented here and our previous data on the expression of glucose oxidase using the same hosts and similar genetic constructions (Rocha et al. 2010) clearly indicate that the levels of heterologous protein expression are highly protein-dependent and not only strain/expression system-dependent. Thus, when the aim is to optimize heterologous protein expression, reliance on a single favorite host/expression system could be a pitfall.

Optimal pHs for the enzymes produced in yeasts were between 8 and 9 (Fig. 2a), close to the optimal pH of the T. thermophilus native EST (Fuciños et al. 2005). The optimal temperature for EST activity expressed in K. marxianus was 50 °C, 5° higher than the optimal temperature for the enzyme expressed in K. lactis (Fig. 2b) and 30° lower than the optimum for the native EST (Fuciños et al. 2005). The heterologous EST expressed in K. marxianus was stable at temperatures up to 80 °C (Fig. 3a). Thus, the decrease in activity at temperatures above 50 °C (Fig. 2b) is most probably not due to denaturation.

Both yeast species added glycosidic chains at two levels to the heterologous proteins (Fig. 4a and b). Previously, we observed that these same strains of Kluyveromyces also attached glycosides to a heterologously expressed glucose oxidase (Rocha et al., 2010). López-López et al. (2010) described a similar glycosylation pattern of the EST from T. thermophilus HB27 expressed in S. cerevisiae. The same three glycoprotein weights were visualized by SDS-PAGE electrophoresis. Glycosylation of heterologous proteins in yeasts is a very general event (Moehle et al. 1987, De Pourcq et al. 2010). This phenomenon was also observed in other fungi and in mammalian cells, for example, in homologous chitobiosidases produced by Trichoderma harzianum (Harman et al. 1993), in extracellular matrix-associated serine protease inhibitors extracted from human umbilical vein endothelial cells (Rao et al. 1996) or in blue laccase produced by Pleorotus ostreatus (Giardina et al. 1999). The glycosylated forms of the recombinant ESTs are fully active as shown in Fig. 4. However, their industrial scale production and downstream processing could benefit from engineering the glycosylation pathways to increase homogeneity and reduce hyperglycosylation of the expressed proteins. This possibility, recently reviewed by Idiris et al. (2010) and De Pourcq et al. (2010), has so far been explored in other yeasts and fungi but scarcely in K. lactis (knocking out the OCH1 gene coding for a mannosyltransferase was proved to be beneficial) and not explored at all in K. marxianus up to our knowledge.

Finally, it is worth mentioning that the samples loaded on the gels presented in Fig. 4, which represent a pool of the supernatant and the periplasmic fractions obtained after cultivation of the Kluyveromyces strains, were simply concentrated by ultrafiltration and not subjected to any purification procedure such as affinity chromatography. Gels were double stained: first for esterase activity and later using Coomassie blue. Few contaminant proteins can be observed, such as the one corresponding to the band of approximately 110 kDa (glycosylated) in the K. marxianus cultivation. This means that the heterologous esterase was obtained in a quite predominant form. This is interesting from an industrial point of view, since it dramatically facilitates downstream operations.

The comparative analysis of the recombinant ESTs produced by K. marxianus and K. lactis in this work points to a biochemical difference between the two enzymes, mainly affecting optimal temperature of activity and stability. The origin of this difference is uncertain hitherto. In this respect, the results here described combined with those from the same EST produced by S. cerevisiae (López-López et al. 2010) indicate that the size of the glycosydic chain is not the factor responsible for such biochemical difference. This difference, that is more pronounced when the recombinant ESTs are compared with the native enzyme, may be attributed to other strain-dependent post-transcriptional modification. We are tempted to speculate that this could be related to a putative trimer being formed in K. lactis, with esterase activity, whereas in K. marxianus this activity is absent as shown in Fig. 4, although further research is needed to elucidate this point.

To conclude, in spite of the differences found between native and heterologous ESTs, Kluyveromyces yeasts seem a suitable choice for the expression of thermophilic enzymes considering their simple cultivation conditions and GRAS status.


We would like to thank Dr. Nancy da Silva (University of California at Irvine) for kindly providing us the pNADFL11 plasmid, Dr. Hiroshi Fukuhara for providing the pSPGK1 plasmid, and Dr. Wésolowski-Louvel for kindly donating the K. lactis PM5-3C strain. SNR acknowledges grants received from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, São Paulo, Brazil) and from Coordenadoria de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brasília, Brazil), which made possible a 1-year internship for the researcher at the Biochemistry and Molecular Biology Laboratory of Universidade da Coruña (Spain), where the DNA work was carried out. This work was financially supported by FAPESP and by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brasília, Brazil), and by Xunta de Galicia (Proyecto PGIDIT06REM38302PR, Spain). General support to the group of the UDC was funded by “Consolidación” program from C.E.O.U. Xunta de Galicia cofinanced by FEDER.

Supplementary material

253_2010_2869_MOESM1_ESM.pdf (27 kb)
ESM 1(PDF 27 kb)

Copyright information

© Springer-Verlag 2010