Carboxyl ester hydrolase from Penicillium expansum: cloning, characterization and overproduction by Penicillium griseoroseum



Elza Fernandes de Araújo, Departamento de Microbiologia/BIOAGRO, Universidade Federal de Viçosa, CEP 36570-000 Viçosa, MG, Brazil. E-mail:



In this study, a gene that encodes a carboxylesterase (carb) in Penicillium expansum GF was cloned, sequenced and overexpressed by Penicillium griseoroseum PG63, and the enzyme was characterized.

Methods and Results

The recombinant strain, P. griseoroseum T55, obtained upon transformation using the plasmid pAN-52-1-carb, showed integration of the carb gene into at least two heterologous sites of the genome by Southern blotting. Furthermore, the recombinant strain T55 exhibited almost a fourfold increase in carboxylesterase activity compared with PG63 strain when both were cultured without inducers. Based on the secondary structure and multiple sequence alignments with carboxylesterases, cholinesterase and lipase, a three-dimensional model was obtained. The α/β barrel topology, that is typical of esterases and lipases, was indicated for the CARB protein with Ser213-Glu341-His456 as the putative catalytic triad. CARB preferentially hydrolysed acyl chains with eight carbon atoms, and its activity was optimal at a pH of 7·0 and a temperature of 25°C. CARB exhibited stability in alkaline pH, high activity under mesophilic conditions and stability in organic solvents.


The CARB protein is potentially useful in bioremediation, food and chemical/pharmaceutical industries.

Significance and Impact of the Study

This study is the first to report the development of a recombinant strain superproducing a Penicillium sp. carboxylesterase.


Lipases (triacylglycerol hydrolases, E.C. 3·1·1·1) and carboxylesterases (carboxyl ester hydrolases, E.C. 3·1·1·3) are members of the α/β hydrolase family that contain the catalytic triad Ser-Asp/Glu-His, where the catalytic residue, Ser, is located at consensus sequence G-x-S-x-G (Jaeger et al. 1999). Lipases and carboxylesterases differ with respect to the pH range within which they are active, their three-dimensional configuration and their substrate affinity (Bornscheuer 2002). Whereas lipases hydrolyse preferentially water-insoluble long-chain acylglycerols, carboxylesterases act on short-chain acylglycerols, giving rise to alcohol and carboxylate (Chahinian et al. 2002).

Lipolytic enzymes are ubiquitous in nature, and those that have been most thoroughly investigated are derived from bacterial strains (Arpigny and Jaeger 1999). In spite of the many uses of carboxylesterases, including the synthesis of esters and modification of triacylglycerols in the food industry, the resolution of racemic mixtures in the fine chemical and pharmaceutical industries, and the biodegradation of industrial waste and pesticides (Faiz et al. 2007; Lv et al. 2011), the industrial application of these enzymes is still limited (Bornscheuer 2002). Thus, the identification of new sources of carboxylesterases that have the properties to meet industrial demands is necessary.

Studies conducted with the filamentous fungus Penicillium griseoroseum PG63, a mutant with a 122-base pair (bp) deletion in the gene that encodes nitrate reductase (niaD), characterized this fungus as a satisfactory host for the expression of homologous or heterologous proteins. The development of an efficient cotransformation system using the pNPG1 plasmid, which contains the P. griseoroseum niaD gene, and a plasmid that contains the gene of interest cloned in frame with a strong constitutive promoter (gpdA of Aspergillus nidulans), thereby favouring the production of the enzyme of interest in inducer-free culture media, allowed our group to generate recombinant strains that overexpressed proteins of pectinolytic complex (Punt et al. 1988; Queiroz et al. 1998; Pereira et al. 2004; Cardoso et al. 2008; Teixeira et al. 2011, 2013a).

The transformation system adopted by our group represents an advantage because the expression of the protein of interest does not require the addition of inducers, unlike observed for the overexpression of carboxylesterases of Streptomyces diastatochromogens and Geobacillus sp. ZH1 by Escherichia coli and Geobacillus stearothermophilus CICC 20156 by Aspergillus niger M54 and Pichia pastoris GS115 (Khalameyzer and Bornscheuer 1999; Liu 2011; Yanbing et al. 2012).

In the present study, the gene that encodes a carboxylesterase protein, characterized according to its activity profile within a given pH range, its substrate specificity and its protein topology, was isolated from P. expansum GF. The gene was cloned, sequenced and overexpressed by recombinant strain P. griseoroseum T55. Our work indicates P. griseoroseum PG63 to be an excellent host strain for the production of homologous and heterologous proteins.

Materials and methods

Micro-organisms and culture conditions

The present study employed the P. expansum GF (Ferreira 2000) and P. griseoroseum PG63 nia (Pereira et al. 2004) strains, as well as ultracompetent E. coli DH5α cells, which were used for obtainment of the expression vector (Sambroock and Russel, 2001; Inoue et al. 1990). The fungal strains were maintained in potato dextrose agar (PDA) medium enriched with the following (g l−1): hydrolysed casein, 1·5; yeast extract, 2·0; and peptone, 2·0. In a first screening, the transformant strains were assessed for the production of carboxylesterase/lipase in BYPO medium (g l−1): peptone, 10; yeast extract, 3·0; meat extract, 5·0; NaCl, 5·0; KH2PO4, 7·0; agar, 15; and tributyrin, 10 ml l−1. The hydrolysis zone (cm) developed around the colonies after culture at 25°C for 5 days was measured. To measure the carboxylesterase/lipase activity, a suspension of conidia in 0·1% Tween 80 in sterile Milli-Q water (final concentration of 106 conidia) was inoculated into 150-ml Erlenmeyer flasks containing 50 ml of buffered mineral medium (BMM) containing the following (g l−1): K2HPO4, 6·98; KH2PO4, 5·44; (NH4)2SO4, 1·0; MgSO4, 7·0; H2O, 1·1; and sucrose, 10 (Teixeira et al. 2013b).

Cloning of the carb gene

Total DNA from P. expansum GF was extracted according to Specht et al. (1982). The gene carb was amplified by polymerase chain reaction (PCR) using the following primers: forward CarbSF 5′-CCA TGG ATG ATT GGA TTT GCG ACC-3′ and reverse CarbSR 5′-GGA TCC CTA CCG CCA AAA GGC TCG-3′. The reaction mixture contained 0·25 μl of GoTaq DNA polymerase (Promega®, Madison, WI, USA), 5·0 μl of 10× buffer, 2·5 μl of 25 mmol l−1 MgCl2, 1·0 μl of dNTPs (2·5 mmol l−1), 1·0 μl of each primer and 14·25 μl of water. The PCR conditions were as follows: initial denaturation at 94°C for 1 min, 35 denaturation cycles at 94°C for 30 s each, annealing at 58°C for 2 min, extension at 72°C for 2 min and a final extension at 72°C for 10 min. The PCR product was purified using the Wizard SV PCR kit (Promega®), cloned in the pGEM-T Easy vector (Promega®) and used in the transformation of ultracompetent E. coli DH5α cells. The plasmid was purified using the Wizard® Plus SV Miniprep DNA purification system (Promega®), and the carb gene was sequenced by Macrogen Inc. (Seoul, South Korea).

Vector construction and PG63 protoplast transformation

The pAN52-1-carb vector (7·4 kb), used in the cotransformation of PG63 strain, was obtained by replacing the gfp gene of Aequorea victoria with carb. The protoplasts were obtained and transformed following the methods of Dias et al. (1997) and Queiroz et al. (1998), respectively. After culturing strain PG63 in complete medium (CM) covered with cellophane paper, approximately 0·8 g of mycelia was collected, resuspended in 5 ml of 0·6 mol l−1 KCl in 10 mmol l−1 phosphate buffer, pH 5·8, containing 30 mg ml−1 of an enzyme mix (Lysing Enzymes; Sigma®, St Louis, MO, USA), and incubated at 30°C and 80 rpm for 3 h. The protoplasts were separated from the hyphal fragments by filtering them through gauze, and the protoplasts were subsequently washed by centrifugation two times in ST buffer: 1 mol l−1 sorbitol and 0·1 mol l−1 Tris–HCl, pH 7·5, and once in STC buffer: ST plus 50 mmol l−1 CaCl2, until a final concentration of 107 protoplasts ml−1 was attained. Then, 200 μl of the protoplast suspension was mixed with 6 μg of the pNPG1, 9 μg of pAN52-1-carb and 50 μl of 50% polyethylene glycol (PEG) 6000. A thermal shock of 42°C was applied to the mixture for 1·5 min. The solution was then incubated on ice for 20 min, and 0·5 ml of the same PEG solution was added. After 20 min at room temperature, the protoplasts were plated using the pour plate method in minimum medium (MM) containing sucrose as the osmotic stabilizer and maintained at 25°C for 5 days.

The resulting transformants were subjected to monosporic purification in which 100 μl of the diluted conidia suspension was sown in MM, and an isolated colony was transferred to another plate containing MM following the growth period. The mitotic stability of the carb transformant strains was assessed using an enzymatic activity assay after 12 months of culture.

The integration of pAN-52-1-carb into the genomes of the transformant strains was confirmed by PCR using the primers gpdAF 5′-TAT TTT CCT GCT CTC CCC ACC-3′ and trpCR 5′-TGC TTC ATC TCG TCT CCC GAA-3′. PCR was performed as described in section 'Cloning of the carb gene' except that the annealing temperature was 57°C.

Determination of carb gene copy number

Total DNA from the P. expansum GF, PG63 and recombinant T55 strains was extracted and cleaved using the restriction enzymes, EcoRI and KpnI, which exhibit no cleavage site and one cleavage site, respectively, in the carb gene sequence. The products of the digestion were separated by electrophoresis in 0·8% agarose gel and transferred to a Duralon-UV membrane (Stratagene®, La Jolla, CA, USA) as described by Southern (1975). Hybridization was performed using the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche®, Madison, WI, USA). The carb gene, which was used as a probe, was labelled in a PCR with the primers CarbSF and CarbSR, using the PCR DIG Probe Synthesis kit (Roche®) and the conditions described in section 'Cloning of the carb gene' (as indicated by the manufacturer).

SDS–polyacrylamide gel electrophoresis

Protein levels were measured according to Bradford (1976). The protein profile of the strains was analysed by electrophoresis in denaturing polyacrylamide gel (SDS-PAGE). The separating gel (12·5% acrylamide) and stacking gel (4% acrylamide) were prepared according to Laemmli (1970). The samples were denatured in buffer (Tris–HCl 60 mmol l−1, pH 6·8; glycerol, 10%; SDS, 2·3%; beta-mercaptoethanol, 2%; bromophenol blue, 1%) at 100°C for 10 min. The gel was subjected to 60 volts for 4 h in buffer (Tris–HCl 72 mmol l−1, pH 8·5; glycine 576 mmol l−1; SDS, 0·24%), stained (methanol, 45%; acetic acid, 9%; Coomassie Brilliant Blue R-250, 0·1%) for 1 h and destained (acetic acid, 7·5%; methanol, 25%) for 2 h. Two millilitres of samples was precipitated. Molecular weight determination was performed by comparison with M – protein molecular weight marker from Sigma® (S8445; 2·0 mg ml−1).

Carboxylesterase/lipase assays

Carboxylesterase activity was measured using p-nitrophenyl butyrate as a substrate. A total of 50 μl of 7·95 mmol l−1 p-nitrophenyl butyrate in isopropanol was added to 400 μl of 50 mmol l−1 NaH2PO4/Na2HPO4 buffer, pH 7·0. A total of 50 μl of the culture supernatant was added to the solution containing the substrate, and the reaction mixture was incubated at 37°C for 10 min. The absorbance was measured at 410 nm using a spectrophotometer. Controls were made by adding water instead culture supernatant to certain flasks. One unit of carboxylesterase activity was defined as the amount of enzyme needed to release 1 μmol l−1 of p-nitrophenol per minute with a molar extinction coefficient of 8·28 × 103 (mol l−1)−1 cm−1 (Boutaiba et al. 2006).

To assess lipase activity, a modification of the procedure by Winkler and Stuckmann (1979) was employed. Culture supernatants were incubated at 37°C for 30 min with 50 μl of 7·95 mmol l−1 p-nitrophenyl palmitate diluted in isopropanol in 400 μl of 50 mmol l−1 NaH2PO4/Na2HPO4 buffer, pH 8·0, containing gum arabic (0·11%) and sodium deoxycholate (0·23%). The reaction was then stopped by adding 1·5 ml of 0·2 mol l−1 Na2CO3 and 40 μl of 0·1 mol l−1 CaCl2 and centrifuged at 10 000× g for 20 min. The absorbance was measured at 410 nm using a spectrophotometer. One unit of lipase activity was defined as the amount of enzyme needed to release 1 μmol l−1 of p-nitrophenol per minute with a molar extinction coefficient of 12·75 × 103 (mol l−1)−1 cm−1 (Boutaiba et al. 2006).

Enzyme characterization

Carboxylesterase production by the T55 recombinant strain was assessed during 120 h of culture in BMM with shaking at 150 rpm and 28°C. The affinity of carboxylesterase for the substrate was assessed using p-nitrophenyl acetate (C2 – N8130 – Sigma®), p-nitrophenyl butyrate (C4 – N9876 – Sigma®), p-nitrophenyl octanoate (C8 – 21742 – Sigma®), p-nitrophenyl decanoate (C10 – N0252 – Sigma®), p-nitrophenyl dodecanoate (C12 – 61716 – Sigma®) and p-nitrophenyl palmitate (C16 – N2752 – Sigma®) as reaction substrates. To assess the residual activity, the reaction was performed as described in section 'Carboxylesterase/lipase assays', except that gum arabic and sodium deoxycholate were not added to the buffer in the case of the C2 and C4 p-nitrophenol esters.

The effect of pH on enzymatic activity was assessed by incubating the culture supernatants with C4 diluted in the following buffer solutions (50 mmol l−1) at 37°C for 10 min: sodium acetate, pH 5·0–6·0; sodium phosphate, pH 7·0; and Tris–HCl, pH 8·0–9·0. The pH stability was determined by incubating the culture supernatants in these same buffer solutions without substrate at room temperature (±25°C) for 1 h. The residual activity was measured as described in section 'Carboxylesterase/lipase assays'.

The effect of temperature on carboxylesterase activity was assessed by incubating the culture supernatants with substrate for 10 min at temperatures ranging from 8 to 60°C. The thermal stability was determined by incubating the culture supernatants for 30 min at temperatures ranging from 8 to 60°C. The residual activity was measured as described in section 'Carboxylesterase/lipase assays'.

The effects of water-soluble solvents on esterase activity were examined by incubating the culture supernatants for 1 h with methanol, ethanol, isopropanol, acetone, acetonitrile, dimethylformamide (DMF), dimethyl sulfoxide (DMSO) and tetrahydrofuran (THF) at concentrations of 10 and 30%. The residual activity was measured as described in section 'Carboxylesterase/lipase assays', and the nontreated sample was considered as 100%.

The enzymatic stability under storage conditions at temperatures of −20, 8 and 25°C was assessed over a period of 60 days. A 0·03% sodium azide solution was used to inhibit bacterial and fungal growth in the culture supernatants stored at 8 and 25°C. All experiments were performed in triplicate.

Sequence analysis and homology modelling

The deduced amino acid sequence was obtained using DNAMAN version 4·0 software (Tamura et al. 2007). The molecular mass, isoelectric point (pI), signal peptide and putative glycosylation sites were obtained at Multiple alignments were performed using ClustalW2 at (Thompson et al. 1994). The three-dimensional protein model was obtained at (Kelley and Sternberg 2009) and validated as per (Pettitt et al. 2005) with a p-value of 0·00031. The model with possible hydrogen bonds was obtained using PyMOL ( The three-dimensional model was compared with structures available at the Protein Data Bank (PDB) ProFunc server (; Laskowski et al. 2005).


The carb gene has an open reading frame of 1792 bp, three exons and two putative introns (57 and 55 bp) and encoded a protein of 559 amino acids, the first 16 of which corresponded to the signal peptide (Fig. 1). The estimated molecular mass and pI of the deduced protein were 59 kDa and 5·4, respectively, and it exhibited 73% identity and 81% similarity with the Neosartorya fischeri NRRL181 protein (GenBank accession number, XP_001263075.1), which belongs to the carboxylesterase family. No putative glycosylation sites were found in CARB.

Figure 1.

Complete sequence of the Penicillium expansum GF carb gene. The deduced amino acid sequence is shown under the corresponding codons. The arrow points to the cleavage site of the signal peptide. The putative introns are indicated with lowercase letters. The consensus sequence, G-x-S-x-G, is indicated in grey. The asterisk represents the stop codon.

Multiple sequence alignments with carboxylesterases, cholinesterase and lipase predicted that CARB would have the Ser213-Glu340-His456 catalytic triad, and the catalytic residue, Ser, was located at consensus sequence G211-x-S213-x-G215 present in a turn motif known as the ‘nucleophilic elbow’, which connects a β-sheet to an α-helix (Figs 2 and 3). The protein comprised 13 central β-sheets surrounded by 16 α-helixes, and the catalytic triad amino acids were nearby. The resulting three-dimensional model was compared with 584 protein structures available at the PDB, and 99·8% similarity with the Torpedo californica acetylcholinesterase (PDB code: 1gpk) and 85·6% similarity with the Pseudomonas cepacia lipase (PDB code: 3lip) were found. These two enzymes belong to the type B carboxylesterase/lipase subfamily and are thus members of the α/β hydrolase family.

Figure 2.

Multiple sequence alignment of CARB with carboxylesterases from Aspergillus niger ATCC 1015 (EHA23255.1), Neosartorya fischeri NRRL 181 (XP_001263075.1) and P. expansum GF, the lipase from Aspergillus fumigatus Af293 (XP_754236.1), and the cholinesterase from A. kawachii IFO 4308 (GAA84575.1).

Figure 3.

Three-dimensional model of the structure of the Penicillium expansum GF carboxylesterase (a) and detail of catalytic triad (b).

The pAN-52-1-carb plasmid, which included the carb gene cloned in frame with the A. nidulans gpdA promoter and trpC terminator, and pNPG1, which contained the gene that encodes the P. griseoroseum nitrate reductase, were used in the cotransformation of P. griseoroseum PG63 strain. A total of 96 transformants were obtained with a transformation efficiency of 16 transformants per microgram of pNPG1.

The initial screening of strains producing lipases and esterases is conventionally performed in plates containing tributyrin-enriched solid media. Enzyme production is measured by the development of a hydrolysis zone around the colonies (Gupta et al. 2003). The 96 obtained transformants were assessed for the production of carboxylesterase in tributyrin-enriched BYPO medium and compared with PG63. Fourteen transformants exhibited a tributyrin degradation zone approximately 1·2–1·5 times larger than that exhibited by PG63. The DNA of the transformants that exhibited hydrolysis zones larger than that of PG63 was extracted, and PCR was performed with the gpdAF and trpCR primers to determine the integration of pAN-52-1-carb into the transformant strains genomes. Among the 14 transformant strains assessed, only T55 exhibited an amplification band of approximately 2190 bp, corresponding to the final 200 bp of gpdA + 1792 bp carb + the initial 200 bp of trpC (data not shown). This transformant was thus selected for further experiments.

The pAN-52-1-carb vector promotes enzyme production in the absence of inducers. Following culture in BMM with sucrose as the carbon source for 72 h, T55 exhibited a fourfold increase in carboxylesterase activity compared with PG63 cultured under the same conditions. In addition, the secretion of extracellular proteins was efficient in recombinant strain T55, attaining a level of 6·5 mg l−1. No difference was found in lipase activity between T55 and PG63, suggesting that the carb gene encodes a carboxylesterase. No differences were found in the dry mycelial mass or final culture pH between T55 and PG63 (Table 1).

Table 1. Assessment of carboxylesterase and lipase activities, total protein, dry mycelial mass and final pH of the culture supernatants of Penicillium expansum GF, PG63 and T55 following 72 h of culture in buffered mineral medium
StrainsCE activity (U l−1)Lipase activity (U l−1)Secreted protein (mg l−1)Dry mycelial mass (g l−1)Final pH
Penicillium expansum GF01·1 ± 0·210·5 ± 1·33·7 ± 0·14·4 ± 0·1
PG631·1 ± 0·81·4 ± 0·34·8 ± 0·76·2 ± 0·65·9 ± 0·3
T554·3 ± 0·21·3 ± 0·36·5 ± 0·75·7 ± 0·85·9 ± 0

The protein profile of the culture supernatant of PG63 and recombinant strain T55 grown on MMB was accessed by SDS-PAGE (Fig. 4). One of the most intense bands of T55 visualized in gel showed approximately 55 kDa in comparison with PG63, agreeing with the estimated molecular mass weight for carboxylesterase produced by P. expansum GF (59 kDa). When the copy number of carb gene was evaluated, the recombinant strain T55 showed integration of carb gene into at least two heterologous sites of the genome (Fig. 5).

Figure 4.

Protein profile of the culture supernatant of PG63 and recombinant strain T55. Sigma® molecular size marker (lane 1), PG63 (lane 2), recombinant strain T55 (lane 3).

Figure 5.

Number of copies of the carb gene in Penicillium expansum GF, PG63 and T55 by Southern blot. Penicillium expansum GF – EcoRI (lane 1), PG63 – EcoRI (lane 2), T55 – EcoRI (lane 3), P. expansum GF – KpnI (lane 4), PG63 – KpnI (lane 5), T55 – KpnI (lane 6).

The carboxylesterase activity of recombinant strain T55 was measured over a period of 120 h of culture and was not associated with culture growth. The activity varied from 5·3 to 6·4 U l−1 in the first 72 h, reached a peak of 13·9 ± 0·1 U l−1 following 96 h of culture and began to decrease after 120 h of culture (12·9 ± 2·2 U l−1; data not shown). The enzymatic activity was greater when C8 was used as a substrate, whereas no activity was detected when C2 was used, and no difference in activity was detected when C4, C10 and C12 were used (Table 2).

Table 2. Hydrolysis of p-nitrophenol esters
SubstrateRelative activity (%)
  1. One-hundred per cent of enzymatic activity = 20·1 U l−1.

p-Nitrophenyl acetate (C2)0
p-Nitrophenyl butyrate (C4)28
p-Nitrophenyl octanoate (C8)100 ± 2·0
p-Nitrophenyl decanoate (C10)29 ± 4·0
p-Nitrophenyl dodecanoate (C12)29 ± 6·0
p-Nitrophenyl palmitate (C16) (pH 7·0)0
p-Nitrophenyl palmitate (C16) (pH 8·0)18 ± 3·0

The highest carboxylesterase activity was observed at pH 7·0 (100%) followed by pH 6·0 (73%), but no activity was detected at pH 8·0 or 9·0. The enzyme exhibited high stability (approximately 100%) following 1 h of incubation at room temperature at pH 6·0, 7·0, 8·0 or 9·0 (Fig. 6). With respect to temperature, carboxylesterase activity was greatest at 25°C (100%) with approximately 55% of the initial activity remaining at 50°C. No activity was found at 60°C, and this absence was due to the background resulting from the spontaneous hydrolysis of the substrate under these conditions. When thermal stability was determined, approximately 85, 42 and 22% of the initial activity remained following 30 min of incubation at 8, 30 and 50°C, respectively (Fig. 7).

Figure 6.

Optimal pH (●) and stability (○) of CARB overexpressed by Penicillium griseoroseum T55. One-hundred per cent of enzymatic activity was considered to be 5·7 U l−1.

Figure 7.

Optimal temperature (●) and stability (○) of CARB overexpressed by Penicillium griseoroseum T55. One-hundred per cent of enzymatic activity was considered to be 6·7 U l−1.

When the stability in the presence of water miscible solvents was examined, CARB protein was tolerant to low concentration (10%) of acetonitrile, DMF and DMSO. Furthermore, at high concentration (30%), CARB was more active in acetone and lost total activity in methanol (Fig. 8).

Figure 8.

Effect of organic solvents on CARB overexpressed by Penicillium griseoroseum T55. 100% per cent of enzymatic activity was considered to be 7·7 U l−1. image_n/jam12215-gra-0001.png 10%; image_n/jam12215-gra-0002.png 30%.

Upon assessment of the stability under storage conditions at various temperatures, 100% of the carboxylesterase activity was maintained following 60 days of incubation at −20 and 8°C. No carboxylesterase activity was found following 30 days of incubation at room temperature.


The carb gene isolated from P. expansum GF encodes a 559 amino acid protein and was classified as a carboxylesterase because it contains the conserved region, F-[GR]-G-x-(4)-[LIVM]-x-[LIV]-x-G-x-S-[STAG]-G ( These findings were confirmed using the Protein Basic Local Alignment Search Tool (BLASTP; and the Microbial Lipase and Esterase Database (MELDB; After alignment with protein sequences available at the PBD, CARB was shown to exhibit a catalytic triad characterized by the replacement of Asp with Glu, an occurrence that is also found in the acetylcholinesterase from T. californica and the EstA esterase from A. niger. Protein EstA exhibits 538 amino acids, 37% identity with the P. expansum GF carboxylesterase, 12 β-sheets, 13 α-helices and the catalytic triad, Ser210-Glu338-His440 (Bourne et al. 2004). The Ser-Glu/Asp-His catalytic triad is typically found in loops and in enzymes with α/β barrel topology, such as lipases, esterases and serine proteases (Holmquist 2000).

In addition to the pH range of activity and the specificity for defined substrates, lipases differ from esterase in that they have a structure known as a lid, which is rich in hydrophobic amino acids around the active site and controls the phenomenon known as interfacial activation (Bornscheuer 2002). Such structure was not observed in CARB (Fig. 3), indicating that CARB bears the characteristics of carboxylesterases. In addition, the preferential hydrolysis of medium acyl chains by CARB might be due to the presence of the aromatic amino acids, Phe234, Phe345, Phe200, Trp210, Tyr221, Tyr226, in regions surrounding the active site. This type of structure does not favour the interaction of CARB with acyl chains longer than 12 carbon atoms.

A total of 96 transformant strains were generated in the present study. The transformation efficiency was 16 transformants per microgram of pNPG1, an efficiency that was lower than the 40 transformants per microgram of pNPG1 obtained by Cardoso et al. (2008), but twofold greater than the eight transformants per microgram obtained when the pNH24 plasmid, which contains the niaD gene from Fusarium oxysporum, was used in the transformation of PG63 (Pereira et al. 2004). The transformation efficiency increased when the plasmid that carries the homologous niaD gene (pNPG1) was used. Stable transformants were obtained when integrative plasmids are used, as shown by mitotic stability, as well as the continuous activity of carboxylesterase observed after 12 months of culture.

The production of microbial esterases increases in the presence of inducers (Panda and Gowrishankar 2005). The carboxylesterase activity in the culture supernatant of P. expansum GF grown in media containing olive oil (8 ml l−1) as the inducer was 2·9 times greater than that of PG63 cultured under the same conditions (data not shown). The pAN-52-1-carb vector allowed the production of carboxylesterase by recombinant strain T55 in culture media containing sucrose as the single carbon source because the expression of carb gene was controlled by the strong constitutive promoter gpdA of A. nidulans. Under these conditions, the carboxylesterase activity of T55 exhibited almost a fourfold increase compared with host strain PG63 (control). The increased production was also demonstrated by SDS-PAGE results. Although the carboxylesterases from Aspergillus oryzae (Converti et al. 2002), Aspergillus nomius (Ushida et al. 2003), A. niger (Bourne et al. 2004) and A. nidulans (Peña-Montes et al. 2008) have been characterized, this is the first report of overexpression of a carboxylesterase from filamentous fungi.

Although both lipases and carboxylesterases hydrolyse the ester bonds of carboxylic acids, carboxylesterases preferentially act on acyl chains up to 12 carbon atoms long, whereas the maximal activity of lipases is exhibited on acyl chains longer than 12 carbon atoms (Eggert et al. 2002). As the highest activity of CARB overexpressed by T55 was observed on acyl chains comprising eight carbon atoms, it was characterized as a carboxylesterase rather than a true lipase. The same behaviour has been shown by the carboxylesterases produced by Sulfolobus solfataricus P1 (Park et al. 2006) and Geobacillus thermodenitrificans (Charbonneau et al. 2010).

According to Fojan et al. (2000), the maximal activity of lipases occurs at pH levels ≥ 8·0, whereas carboxylesterases perform better at a pH level of approximately 6·0. The optimal pH of most of the already described carboxylesterases varies from 6·0 to 7·0, and this was confirmed in the present study. With regard to temperature, optimal carboxylesterase activity occurs between 15°C (Lemak et al. 2012) and 85°C (Park et al. 2006). CARB was the most active at 25°C and maintained 85% of its initial activity following 1 h of incubation at 8°C, indicating that it could be classified as a cold lipolytic enzyme (Joseph et al. 2008).

The technological utility of enzymes can be enhanced using them in organic solvents rather than their natural reaction media (Klibanov 2001). Lipases and esterases perform reactions of hydrolysis (aqueous media) or synthesis (organic media) of ester linkages (Bornscheuer 2002). The synthesis of esters, applicable as flavour in food industry (Torres et al. 2009), and resolution of racemic mixtures in the pharmaceutical sector are conducted in organic media (Liu et al. 2010). The stability of CARB in organic solvents suggests its potential use in organic synthesis.

To be applied in biotransformation reactions, the ideal esterase must exhibit high activity under mesophilic conditions, a critical factor for the protection of labile substrates in deleterious reactions (Wood et al. 1995). The findings of this work suggest that similar to the EstC esterase of Streptomyces coelicolor A3(2) (Brault et al. 2012), CARB protein might be a good candidate for use in the food industry as a flavour enhancer, in the bioremediation of oil-contaminated soils, as well as for the synthesis of pharmaceutical compounds, as such processes are performed at low temperatures (Araújo et al. 2012).


This research was supported by the Brazilian Foundation for Research Support of Minas Gerais (FAPEMIG) and the National Council for Scientific and Technological Development (CNPq).

Conflict of Interest

The authors declare no conflict of interests.