Enzyme and Microbial Technology 57 (2014) 48–54

Contents lists available at ScienceDirect

Enzyme and Microbial Technology journal homepage: www.elsevier.com/locate/emt

Characterization of a heat-active archaeal ␤-glucosidase from a hydrothermal spring metagenome Carola Schröder, Skander Elleuche, Saskia Blank, Garabed Antranikian ∗ Institute of Technical Microbiology, Hamburg University of Technology, 21073 Hamburg, Germany

a r t i c l e

i n f o

Article history: Received 4 November 2013 Received in revised form 14 January 2014 Accepted 27 January 2014 Available online 3 February 2014 Keywords: Metagenomic library Pressure Thermostability Aluminum chloride Hyperthermophile

a b s t r a c t Thermostable enzymes are required for application in a wide range of harsh industrial processes. High stability and activity at elevated temperatures, as well as high tolerances toward various reagents and solvents, are needed. In this work, a glycoside hydrolase family 1 ␤-glucosidase (Bgl1) of archaeal origin was isolated from a hydrothermal spring metagenome. The enzyme showed a broad substrate spectrum with activity toward cellobiose, cellotriose and lactose. Compared to most enzymes, extremely high specific activity with 3195 U/mg was observed at 90 ◦ C and pH 6.5. Bgl1 was completely stable at pH 4.5–9.5 for 48 h at 4 ◦ C. More than 40% of activity was measured at 105 ◦ C. A thermal activation was observed at 90 ◦ C after 30 min. Enzyme stability was enhanced (5- and 7-fold) after applying pressure of 100 and 200 bar at 90 ◦ C for 2 h, respectively. The affinity of the ␤-glucosidase to its substrate was significantly increased in the presence of AlCl3 . The Ki value for glucose was 150 mM. These distinctive characteristics distinguish Bgl1 from other enzymes described so far and make this enzyme suitable for application in numerous processes that run at high temperatures. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Microorganisms exhibiting high growth temperature are considered as producer of extreme thermostable enzymes (extremozymes). The majority of these enzymes are more resistant to chemical denaturation compared to their counterparts produced from mesophiles [1]. Of special interest for industrial application are hypertherophilic Archaea since they grow optimally between 80 and 110 ◦ C and produce unique thermostable enzymes [2]. Industrial processes that run at elevated thermal conditions have many advantages. The solubility of complex substrates increases due to lower viscosity and higher diffusion rate of the substrate. Furthermore, the risk of contamination is significantly reduced [3]. In the last decades, many biocatalysts from Archaea exhibiting remarkable characteristics, such as optimal function at 100–110 ◦ C, were discovered [4]. Especially glycoside hydrolases are suitable for a wide range of biotechnological applications. For the efficient bioconversion of lignocellulose to biofuels and chemicals novel glycoside hydrolases that act synergistically are needed [5]. This will lead to the development of biorefinery of the 2nd generation that is competitive to the petroleum-based plants [6]. Endoglucanases (EC 3.2.1.4) cleave within the amorphous regions of long glucose

∗ Corresponding author. Tel.: +49 040 428783117. E-mail address: [email protected] (G. Antranikian). 0141-0229/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.enzmictec.2014.01.010

chains. Cellobiohydrolases (EC 3.2.1.91) act on reducing or nonreducing ends of accessible polysaccharides. Cellobiose, cellotriose or longer oligosaccharides that are produced from this reaction represent the substrates for ␤-glucosidases (EC 3.2.1.21). These enzymes hydrolyze terminal, non-reducing ␤-d-glucose residues and produce glucose monomers [7]. The limiting factor in the conversion of cellulose to glucose is the inhibition of endoglucanase activity by oligosaccharides [8]. Other applications of ␤-glucosidases include the cleavage of phenolic and phytoestrogen glucosides from fruits and vegetables for medical purposes or to enhance the quality of beverages. Furthermore, the enzyme is used to hydrolyze naringin resulting in the reduction of fruit bitterness or gellan hydrolysis resulting in the reduction of viscosity [8,9]. Since most microorganisms, particularly extremophiles, are not cultivable under laboratory conditions, the metagenomic approach was introduced. Metagenomic libraries from high temperature habitats offer a great potential for the identification of novel heat-stable enzymes [10]. Using function-based or sequence-based methods putative genes, e.g. ␤-glucosidase-encoding genes can be identified [11,12]. A number of novel biocatalysts with remarkable properties have been already detected from unknown microorganisms in such metagenomes [13]. In this work, a metagenome was recovered from a hot spring at the island São Miguel (Azores, Portugal). The prevalent elevated temperatures of up to 70 ◦ C favored the presence of thermophilic Bacteria and Archaea. A novel ␤-glucosidase was purified and characterized.

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54 2. Materials and methods 2.1. Starting material Samples were collected at Furnas Valley from different locations of a hot spring named Caldeirão on the island São Miguel (Azores, Portugal, 37◦ 46 N, −25◦ 18 E) in September 2010 [14]. Environmental samples included water, mud and sediment and were taken at locations with temperatures ranging from 60 to 70 ◦ C. The pH varied from pH 6 to 7. The collected material was transported and stored at 4 ◦ C under anoxic conditions by adding Na2 S (2.5%, w/v, pH 7, final concentration 0.01%, w/v). 2.2. DNA isolation and amplification The DNA was isolated through direct lysis [15]. Due to low amounts of recovered DNA, the entire genetic information was amplified by multiple displacement procedure. Therefore, 1 ␮L of the template DNA was denatured for 3 min at 94 ◦ C with 1 ␮L random hexamer primer (500 pmol), 2.5 ␮L phi29 buffer (10×) and 1 ␮L DTT. Subsequently, isothermal amplification was performed by addition of 2 ␮L phi29 buffer (10×), 2 ␮L dNTPs (25 mM each), 10 U phi29 DNA polymerase (Thermo Scientific, Schwerte, Germany) in a volume of 15 ␮L filled up with deionized water for 12–16 h at 30 ◦ C. The resulting hyperbranched structures were dissolved [16]. Subsequently, single-stranded overhangs were digested, emerging gaps were filled with nucleotides and the DNA was finally dephosphorylated. After each step the amplicon was purified using SureClean solution (Bioline, Luckenwalde, Germany). 2.3. Metagenomic library construction and screening Fragments with sizes in the range 4–10 kb were purified from an agarose gel (1%) using the GeneJET Gel Extraction kit (Thermo Scientific, Schwerte, Germany). The addition of a single 3 -deoxyadenosine overhang enabled the ligation into the pCR-XL-TOPO vector. E. coli Top10 cells were transformed with ligation mixture according to the manufacturer‘s instructions (Invitrogen, Karlsruhe, Germany). The cells were plated as recommended. After sweeping off the cells, the plasmids were isolated employing the Qiagen Plasmid Plus Midi Kit (Hilden, Germany). Screening for activity-conferring genes took place by using clones derived from transformation of E. coli DH5␣ cells with 1 ␮L of metagenomic plasmid library. The cells were grown on LB agarplates with 50 ␮g/mL kanamycin. The colonies were replicated by stamping on velvet with a subsequent transfer on a new plate. After growth of colonies, screening top-agarose (50 mM sodium acetate, 2.5 mM calcium chloride x 2 H2 O, 170 mM NaCl, 2.5 mM esculin hydrate, 0.4 mM ammonium ferric citrat, 1.2% agarose) was used to overlay cells. Clones were incubated at 70 ◦ C over night. 2.4. Analysis of activity-conferring clones Plasmids were isolated employing the GeneJET Plasmid Miniprep Kit (Thermo Scientific, Schwerte, Germany). Inserts of pCR-XL-TOPO were sequenced by Eurofins MWG Operon (Ebersberg, Germany). Potential activityconferring open reading frames (ORFs) were identified with frameplot 4.0 beta (http://nocardia.nih.go.jp/fp4/). BLAST (Basic Local Alignment Search Tool) analyses were done to determine protein sequence identities employing GenBank (National Center for Biotechnology Information, http://www.ncbi.nlm.nih.gov/). A hypothetical model was constructed on the basis of the deduced amino acid sequence using the SWISS-MODEL workspace (http://swissmodel.expasy.org/) [17].

49

The identified potential catalytic glutamate residues were mutated by PCR using degenerated primers exhibiting one exchanged base (E209 to G209 – for: GCACAATGAATGGACCAAATGTAAC, rev: GTTACATTTGGTCCATTCATTGTGC; E395 to G395 – for: GATAACTGGAAATGGAGTAGC, rev: GCTACTCCATTTCCAGTTATC; exchanged base is underlined). The constructs were fused using the flanking primers of the gene and the purified fragments as template. 2.6. Biochemical characterization of Bgl1 Unless otherwise noted, the standard activity assay was carried out with 2 mM 4-NP-␤-d-glucopyranoside, 20 mM maleate buffer (pH 6.5) in a volume of 1 mL (with deionized water) at 90 ◦ C in a heating block. After 10 min, the assay was stopped by addition of 100 ␮L 0.1 M Na2 CO3 . The optical density was determined at ␭ = 410 nm. One unit of ␤-glucosidase activity is defined as the amount of protein required to release 1 ␮mol of 4-nitrophenol per min at 90 ◦ C and pH 6.5. Protein concentration was determined using the Bradford assay [19]. When influence of pH values was tested, Britton–Robinson buffer (20 mM) was used for a pH range between 3.0 and 10.0 [20]. Temperatures above 95 ◦ C were adjusted using heated oil. Influence of pressure was directly tested by performing the activity assay at 90 ◦ C in batch reactors [21]. Diluted enzyme was incubated at 90 ◦ C in a reactor to investigate the influence of 100 and 200 bar on enzyme stability. Subsequently, after slow release of pressure, the activity assay was carried out at 90 ◦ C. Influence of metal ions at a final concentration of 1 or 5 mM (AgNO3 , AlCl3 (0.001–3 mM), CaCl2 , CoCl2 , CrCl3 , CuCl2 , FeCl2 , FeCl3 , KCl, MgCl2 , MnCl2 , NaCl, NiCl2 , RbCl, SrCl2 and ZnCl2 ) was measured in 50 mM maleate buffer. Detergents (CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate), CTAB (cetyltrimethyl-ammonium bromide), SDS, Triton X-100, Tween 20 and Tween 80), additives (␤-mercaptoethanol, DTT (dithiothreitol), EDTA (ethylenediaminetetraacetic acid), guanidine hydrochloride, urea, iodoacetic acid, sodium azide and Pefabloc SC® ) were added to the assay to a final concentration of 5 mM and organic solvents (acetone, acetonitrile, DMF (dimethylformamide), DMSO (dimethyl sulfoxide), ethanol, methanol, 1,2-propanediol, 1-propanol and 2-propanol) were supplemented to 10%. Substrates, such as 4-NP-␤-d-glucopyranoside, 4-NP-␣d-glucopyranoside, 4-NP-␤-d-galactopyranoside, 4-NP-␣-d-galactopyranoside, 2NP-␤-d-galacto-pyranoside, 4-NP-␤-d-fucopyranoside, 4-NP-␤-d-xylopyranoside, 4-NP-␤-d-manno-pyranoside, 4-NP-␤-d-cellobioside, 4-NP-␣-l-arabinopyranoside and 4-NP-␤-d-glucuronide were tested in a final concentration of 2 mM. Cellobiose (1%, w/v), cellotriose (0.1%, w/v) and lactose (1%, w/v) were analyzed by HPLC (high-performance liquid chromatography) after incubation at 70 ◦ C for 3 h using a HPX42A column (Bio-Rad, München, Germany). High purity water was filtrated and flushed (helium, 10 min) before being employed as running solvent with a flow rate of 0.6 mL/min. Enzyme stability was investigated by pre-incubation of pure enzyme supplemented with 0.1 mg/mL BSA (bovine serum albumin) followed by an activity assay at 90 ◦ C. Kinetic parameters were determined by plotting reaction rates against substrate concentrations followed by nonlinear regression analysis to determine Km and Vmax [22]. The inhibition constant Ki was investigated by plotting the inverse reaction rate using different substrate concentrations against inhibitor concentration [23]. 2.7. Nucleotide sequence accession number DNA sequence of bgl1 including flanking regions was deposited in GenBank (HG326254).

3. Results 2.5. Gene expression and purification

3.1. Identification and analysis of the ORF bgl1 The gene bgl1 was amplified (for: GGATCCGTAAAGTTCCCTAAAGG, BamHI recognition site is underlined, rev: GTCGACCTAAGTAAGAACGTTTGG, SalI recognition site is underlined) and ligated into subcloning vector pJET (Thermo Scientific, Schwerte, Germany) prior to transformation of E. coli NovaBlue. The plasmids were isolated and bgl1 was recovered by restriction excision. The linear gene bgl1 was ligated into BamHI/SalI linearized pQE-80L resulting in plasmid pQE-80L::bgl1. Subsequently, E. coli C43 cells were transformed. 1 L of E. coli C43 pQE-80L::bgl1 culture was grown to an optical density ( = 600 nm) of 0.6. Gene expression was induced with 0.1 mM IPTG. After 12 h of growth, cells were harvested by centrifugation (7000 × g, 4 ◦ C, 10 min). Resuspension of sedimented cells was achieved with 15 mL lysis buffer (50 mM NaH2 PO4 , 300 mM NaCl, 10 mM imidazole, pH 8). For cell disruption, pressure of 1000 psi was applied five times using a French Pressure Cell Press (SLM-Aminco, Maryland, USA). After centrifugation (13,000 × g, 4 ◦ C, 15 min), Ni-NTA-affinity chromatography was conducted with 4 mL of supernatant using NiNTA Agarose and Ni-NTA Spin Columns according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Pure elution fractions were dialyzed (Visking® dialysis tubing, Serva, Heidelberg, Germany) in 20 mM maleate buffer, pH 6.5 (20 mM maleic acid, 20 mM NaOH). Protein was visualized and the molecular mass was surveyed by SDS-PAGE (12%, sodium dodecyl sulphate-polyacrylamide gel electrophoresis) [18]. The quaternary structure was investigated by native PAGE using the Pro Gel Tris Glycin gel (4–20%) as recommended by the manufacturer (Anamed Elektrophorese, Groß-Bieberau/Rodau, Germany).

The putative activity-conferring open reading frame (ORF) bgl1 (1488 bp) was identified in a metagenome derived from a hot spring in the Azores (Portugal) by activity-based screening of the respective E. coli DH5␣ library using esculin as substrate. The deduced protein sequence exhibited an identity of 53% (98% sequence overlap) to a putative ␤-galactosidase from Thermoproteus uzoniensis (YP 004338090). The insert of the plasmid pCR-XL-TOPO comprised 3755 bp (Fig. 1). Besides bgl1, two further ORFs (glnS1 complete and xylA incomplete) were identified on different reading frames but the insert of the gene library clone E. coli DH5␣ pCRXL-TOPO::bgl1 did not show an operon-like organization of ORFs. The deduced amino acid sequence of glnS1 (1428 bp) exhibited 63% identity in a 98% amino acid overlap to a putative glutamine synthetase from Caldiarchaeum subterraneum (BAJ49060). The deduced protein sequence of the incomplete ORF xylA (276 bp) exhibited 51% identity to a section of a xylose-isomerase from Sphingobacterium sp. (YP 004318167). ORF bgl1 contained a GC content

50

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54

aa 4-477: GH1 superfamily Bgl1: 495 aa Insert: 3,755 bp glnS: 1,428 bp

bgl1: 1,488 bp

xylA: 276 bp

Fig. 1. Insert organization of E. coli pCR-XL-TOPO::bgl1. The insert of pCR-XL-TOPO::bgl1 contained two complete and one incomplete ORF. The deduced protein sequence of glnS exhibited 63% identity to a putative glutamine synthetase from Caldiarchaeum subterraneum (BAJ49060), bgl1 exhibited 53% identity to a putative ␤-galactosidase from Thermoproteus uzoniensis (YP 004338090) and xylA is 51% identical to a partial putative xylose isomerase from Sphingobacterium sp. (YP 004318167). A conserved domain characteristic for GH1 was detected (pfam00232: GH1 superfamily) within the deduced protein sequence of Bgl1.

of 36.6%. The GC content at the third base position in each triplicate (GC3 content) was shown to be 26.7%. The proportion of purines accounted for 60%. Noticeably, out of six synonymous arginine codons 19 codons were composed of AGA, three of AGG and none of the remaining four. Additionally, 22 isoleucine codons of ATA, four of ATC and eight of ATT were observed in the nucleotide pattern. The deduced 495 amino acid comprising sequence of Bgl1 exhibited a molecular mass of 57.2 kDa and a theoretical pI of 6.3. Similarities to the glycoside hydrolase family 1 (GH1, cl01046) were detected. Conserved domains of GH1 (carbohydrate transport and mechanism, COG2723) and of GH1 superfamily (pfam00232) were detected ranging from amino acid 1 to 484 and from 4 to 477, respectively. A hypothetical tertiary model was constructed on the basis of sequence information using the automated mode of the SWISS-MODEL workspace (data not shown). A characteristic TIM-barrel structure was predicted, wherein the conserved motifs asparagine–glutamate–proline and threonine–glutamate–asparagine–glycine were detected. The glutamate residues (E209 and E395) were identified as potential catalytic residues. Accordingly, the hypothetical model showed these amino acids to be arranged in the center of the protein. To confirm the catalytic function, mutants were constructed by single base substitution resulting in glycine (G) replacing either E209 or E395. The respective mutated E. coli strains exhibited no activity toward esculin (Fig. 2). The putative protein Bgl1 exhibited 50–53% identity to putative glycoside hydrolases from species belonging to the class Thermoprotei of the phylum Crenarchaeota, such as Thermoproteus uzoniensis (YP 004338090), Acidilobus saccharovorans (YP 003816826), Caldivirga maquilingensis (YP 001540243) and Vulcanisaeta moutnovskia (YP 004243760). The transcription initiation motif was found to be typical for archaeal genes with a TATA box-like region, just as described for

Fig. 3. SDS- and native PAGE analysis of purified Bgl1. Purified Bgl1 was verified by SDS PAGE (200 V, 50 min, M: Unstained Protein Molecular Weight Marker, Fermentas and native PAGE (Pro Gel Tris Glycin 4–20%, 12 h, 80 V, 4 ◦ C, M: High Molecular Weight marker, GE Healthcare) was employed for quaternary structure investigation.

eukaryotes, between position −41 and −32 (TTATTAATT). Additionally, two upstream-located adenines at positions −47 and −46 (transcription factor B recognition site) were detected. 3.2. Enzyme purification The ORF bgl1 was cloned into plasmid pQE-80L and overexpressed in E. coli C43. The protein Bgl1 was purified by Ni-NTA affinity chromatography. A purification factor of 2.0 was determined with a yield of 6.2% (Table 1). The predicted monomeric calculated molecular weight of 57.2 kDa was confirmed by SDSPAGE analysis (Fig. 3). The quaternary structure of Bgl1 was predicted to be tetrameric by the program SWISS-MODEL (data not shown). This was confirmed by detection of a single band at 230 kDa by native PAGE analysis (Fig. 3). 3.3. Substrate specificity

Fig. 2. Plate activity test of point-mutated strains. Plated cells of E. coli C43 pQE-80L::bgl1 with single-base mutation in the E209-encoding triplet (G209), the unmutated strain and with single-base mutation in the E395-encoding triplet (G395). The respective glutamates were previously identified as catalytic residues due to sequence comparisons. Activity was verified using esculin-agarose overlay.

The enzyme Bgl1 accepted a broad range of substrates and preferred 4-NP-␤-d-fucopyranoside with a specific activity of 3529.7 U/mg (Table 2). The specific activity toward 4-NP-␤-dglucopyranoside was calculated to be 3194.7 U/mg and toward 4-NP-␤-d-galactopyranoside 2137.7 U/mg. Specific activities of 971.3 U/mg and 288.6 U/mg were obtained using 2-NP-␤-dgalacto-pyranoside and 4-NP-␤-d-xylopyranoside as substrate. Lactose, cellobiose (Fig. 4) and cellotriose (data not shown)

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54

51

Table 1 Purification of Bgl1.

Crude extract Ni-NTA

Protein [mg/mL]

Volume activity [U/mL]

43.0 0.1

4625.7 287.5

Total activity [U] 13,877.0 862.6

Specific activity [U/mg]

Purification factor

Yield [%]

1613.9 3194.7

1.0 2.0

100.0 6.2

Fig. 4. HPLC analyses of lactose and cellobiose degradation. The degradation of lactose and cellobiose by Bgl1 was investigated by conducting the activity assay at 70 ◦ C for 3 h in 20 mM maleate buffer (pH 6.5) using 1% (w/v) of substrate. Subsequently, the samples were subjected to HPLC analyses using a HPX42A column with a flow rate of 0.6 ml/min using high purity water.

Table 2 Substrate specificity of Bgl1.

3.5. Influence of temperature and pH

Substrate (2 mM)

Specific activity [U/mg]

4-NP-␤-d-fucopyranoside 4-NP-␤-d-glucopyranoside 4-NP-␤-d-galactopyranoside 2-NP-␤-d-galactopyranoside 4-NP-␤-d-xylopyranoside

3529.7 3194.7 2137.7 971.3 288.6

conversion was shown by HPLC analysis. After incubation for 3 h at 70 ◦ C, lactose was partially converted to form glucose and galactose. Cellobiose was partially hydrolyzed to glucose. 3.4. Influence of hydrolysis products on Bgl1 The partial degradation of lactose, cellotriose and cellobiose indicated an inhibitory effect of the hydrolysis products. The activity of Bgl1 was measured at different concentrations of 4-NP␤-d-gluco-pyranoside (1, 2 and 3 mM) in the presence of glucose (0.5, 1 and 1.5 M). The resulting point of intersection of the plotted results reflected the inhibition constant (Ki ) at 150 mM glucose (data not shown). Moreover, the intersection point indicated a competitive type of inhibition.

Bgl1 was active over a broad temperature (55–115 ◦ C) and pH (4.5–8.5) range using 4-NP-␤-d-glucopyranoside as substrate. Highest activity was detected at 90 ◦ C and at pH 6.5 (Fig. 5). A residual activity of 62% ± 3 and 40% ± 3 was measured at 100 ◦ C and 105 ◦ C, respectively. To analyze the thermostability, the enzyme was incubated separately at different temperatures. Subsequently, the activity assay was performed at 90 ◦ C (Fig. 6). A heat activation of Bgl1 was measured after 30 min at 90 ◦ C (185% ± 10). Incubation for 1.5 h at 90 ◦ C resulted in a residual activity of 67% ± 9. After an incubation time of 24 h at 50 and 60 ◦ C the residual activity was 78% ± 6 and 68% ± 7, respectively. The enzyme was completely stable at 4 ◦ C and pH 4.5–9.5 for at least two days (data not shown). 3.6. Influence of pressure Activity assays were performed at 1, 100 and 200 bar at 90 ◦ C to investigate the influence of pressure on the activity of Bgl1. Pressure (1–200 bar) did not directly influence the enzymatic activity. However, the preincubation of the enzyme at various pressures had a significant effect on the catalytic performance. After pretreatment at 100 and 200 bar for 60 min at 90 ◦ C, the activity was raised to

Fig. 5. Effect of temperature and pH on activity of Bgl1. (A) The ␤-glucosidase activity of Bgl1 was measured at different temperatures (35–115 ◦ C) in 20 mM maleate buffer (pH 6.5) and (B) at different pH values (3–10) at 90 ◦ C with 2 mM 4-NP-␤-d-glucopyranoside. Temperatures above 95 ◦ C were measured in heated oil (discontinuous line).

52

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54 Table 3 Effects of metal ions on the activity of Bgl1.

Fig. 6. Effect of temperature on stability of Bgl1. The ␤-glucosidase activity of Bgl1 was measured at 90 ◦ C in 20 mM maleate buffer (pH 6.5) with 2 mM 4-NP-␤-dglucopyranoside after incubation of enzyme separately at different temperatures (50–90 ◦ C). The starting activities were set to 100%.

Reagent

Relative activity [%] (1 mM)

Relative activity [%] (5 mM)

None AgNO3 AlCl3 CaCl2 CoCl2 CrCl3 CuCl2 FeCl2 FeCl3 KCl MgCl2 MnCl2 NaCl NiCl2 RbCl SrCl2 ZnCl2

100 ± – 168 ± 105 ± 101 ± 126 ± – 101 ± 112 ± 102 ± 104 ± 72 ± 100 ± 88 ± 102 ± 98 ± –

100 ± – 49 ± 69 ± 90 ± 55 ± – 45 ± 31 ± 101 ± 108 ± 84 ± 99 ± 83 ± 103 ± 93 ± –

5 1 4 5 3 3 5 2 4 6 2 3 1 3

5 5 4 2 7 12 7 2 1 4 1 10 3 2

Table 4 Effects of chemical reagents on the activity of Bgl1. Reagent (5 mM)

Relative activity [%]

None ␤-Mercaptoethanol CHAPS CTAB DTT EDTA Guanine hydrochloride Pefabloc SDS Triton X-100 Tween 20 Tween 80

100 113 81 6 119 126 28 44 46 118 52 41

± ± ± ± ± ± ± ± ± ± ± ±

5 3 6 2 2 8 6 4 7 5 5 4

3.7. Influence of various additives Fig. 7. Effect of pressure on stability of Bgl1. Bgl1 was incubated at atmospheric pressure (1 bar), 100 bar and 200 bar for one and 2 h at 90 ◦ C. Subsequently, the residual activity was measured with 4-NP-␤-d-glucopyranoside (2 mM) and maleate buffer (20 mM, pH 6.5) at 90 ◦ C and 1 bar.

246% ± 11 and 224% ± 17, respectively (Fig. 7). When the enzyme was subjected to 100 and 200 bar at 90 ◦ C for a longer period (2 h), the activity measured subsequently was raised to 255% ± 7 and 330% ± 7, respectively. Under standard condition (1 bar) a decrease of activity to 45% ± 6 was observed.

180 relative activity [%]

160 140

Addition of 1 mM CaCl2 , FeCl2 and FeCl3 and 1 or 5 mM of CoCl2 , KCl, MgCl, MnCl2 , NaCl, NiCl2 , RbCl and SrCl2 did not significantly influence the activity (Table 3). AgNO3 , CuCl2 and ZnCl2 (1 mM) completely inhibited the enzyme. The presence of 1 mM CrCl3 and AlCl3 led to an increase of activity of up to 126% ± 3 and 168% ± 1, whereas a concentration of 5 mM reduced the activity. To further investigate the influence of aluminum ions on the activity of Bgl1, different concentrations between 0.001 and 3.0 mM of AlCl3 were tested. The enzymatic action was enhanced in presence of 0.1–2 mM (Fig. 8). The addition of the detergents CTAB, guanine hydrochloride, Pefabloc, SDS, Tween 20 and Tween 80 (5 mM) reduced the activity of Bgl1 significantly (Table 4). The disulfide bond reducing agents ␤-mercaptoethanol and DTT increased the activity of Bgl1. The presence of urea, iodacetic acid and sodium azide had no influence (data not shown), whereas EDTA enhanced the activity (Table 4).

120 100

Table 5 Effects of solvents on the activity of Bgl1.

80 60 40 20 0 0

0.001

0.01

0.1

0.2

0.5

1

2

3

aluminum chloride [mM] Fig. 8. Effect of aluminum chloride on activity of Bgl1. The ␤-glucosidase activity of Bgl1 was measured in the presence of different aluminum ion concentrations (0–3 mM) in 50 mM maleate buffer (pH 6.5) and 90 ◦ C with 2 mM 4-NP-␤-dglucopyranoside.

Reagent (10%)

Relative activity [%]

None Acetone Acetonitrile DMF DMSO Ethanol Methanol 1-Propanol 1,2-Propanediol 2-Propanol

100 ± 67 ± 16 ± 41 ± 63 ± 63 ± 81 ± – 52 ± 31 ±

5 23 13 5 2 13 17 1 18

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54

53

Table 6 Kinetic parameters of Bgl1. Substrate (2 mM)

Salt (1 mM)

Vmax [␮mol/min]

Km [mM]

kcat [s−1 ]

kcat /Km [mM/s]

4-NP-␤-d-glucopyranoside 4-NP-␤-d-glucopyranoside 4-NP-␤-d-galactopyranoside 4-NP-␤-d-galactopyranoside

– AlCl3 – AlCl3

0.022 0.021 0.027 0.026

0.8 0.4 6.5 3.8

19,922 19,081 24,451 23,546

26,213 54,517 3779 6262

The addition of solvents (10%), such as acetone, acetonitrile, DMF, DMSO, ethanol, methanol, 2-propanol and 1,2-propanediol resulted in the reduction of activity, while 1-propanol completely inhibited Bgl1 (Table 5). 3.8. Enzyme kinetics The kinetic parameters were determined using 4-NP-␤-d-glucopyranoside and 4-NP-␤-d-galactopyranoside as substrate. Since the addition of 1 mM AlCl3 led to an increase of enzyme activity, kinetic experiments were performed in the presence and absence of Al3+ simultaneously (Table 6). Km amounted to 0.8 mM 4-NP␤-d-glucopyranoside and 6.5 mM 4-NP-␤-d-galactopyranoside. In the presence of aluminum ions the Km values decreased to 0.4 mM and 3.8 mM, respectively, while no change in the maximum reaction rate (Vmax ) was observed. The catalytic efficiency (kcat /Km ) was hence higher in the presence of the inorganic salt (Table 6). Values of 54,517 mM 4-NP-␤-d-glucopyranoside and 6262 mM 4-NP-␤-d-galactopyranoside per second were calculated. 4. Discussion Thermostable ␤-glucosidases are required for various biotechnological applications including food and beverage industry and lignocellulose-utilizing biorefineries. The finding of a novel heatactive ␤-glucosidase will contribute to the exploration of the divers genetic potential of extreme habitats. Several typical characteristics indicate that the gene encoding Bgl1 is of archaeal origin. The open reading frame exhibited transcription initiation regions typical for Archaea. Furthermore, the arginine codon AGA and the isoleucine codon ATA were reported to occur most frequently in archaeal genes [24,25]. Accordingly, a preference for these codons was detected within the codon usage pattern of bgl1. A codon preference within synonymous codons was reported to result from selective pressure within evolution of populations. A specific pattern may help to identify phylogenetic relationships [26]. In addition, the deduced protein Bgl1 exhibited structural similarities (50 to 53%) to annotated, uncharacterized enzymes from the archaeal families Thermoprotacea and Acidilobacea. Recently, the diversity of prokaryotes isolated from a hot spring at Furnas Valley (Azores, Portugal) was investigated by 16S rRNA gene sequence analysis. About 25% of Crenarchaeota in this sample were described to be representatives of Thermoproteacea [14]. Since the gene bgl1 was isolated from the same sampling site, an uncultured crenarchaeon is assumed as originating host. Potential characteristics of nucleotide sequences from thermophiles, such as the presence of a high GC content, a high GC3 content or a high purine proportion were not detected for bgl1 [27,28]. Crenarchaeal genes encoding thermostable proteins were reported to differ from corresponding bacterial genes, especially by their low content of guanine and cytosine [29]. Homology modeling predicted the protein Bgl1 to display typical GH1 characteristics with a compact TIM-barrel structure and two potential catalytic glutamate residues (E209 and E395). The catalytic requirement was confirmed by single base mutations resulting in proteins with G209 or G395 that did not show ␤glucosidase activity toward esculin. Similar mutants constructed

by other authors also reflected the functional importance of the previously sequence-based identified catalytic glutamate residues [30,31]. The probable homotetrameric complex consisted of four 57.2 kDa monomeric units, which is in good agreement to prior reported ␤-glucosidases affiliated to GH family 1 [9]. Several short chain sugar components were degraded, such as lactose, cellobiose and cellotriose. Notable high specific activities (more than 3190 U/mg) were calculated for the conversion of 4-NP␤-d-fucopyranoside and 4-NP-␤-d-glucopyranoside. The majority of characterized ␤-glucosidases were reported to exhibit specific activities below 100 U/mg (http://www.brenda-enzymes.org). Moreover, Bgl1 showed a tolerance against glucose with Ki for competitive inhibition amounting to 150 mM, which is comparable to other enzymes, such as the thermostable ␤-glucosidase from Fervidobacterium islandicum with a Ki value of 211 mM glucose [32]. The thermostability of the enzyme at 90 ◦ C was enhanced 5- or 7-fold after 2 h when pressure (100 and 200 bar) was applied. Activities higher than 300% were observed. Usually, heat causes enzymes to unfold and aggregate [33]. It can be assumed that pressure counteracts denaturation of Bgl1 at 90 ◦ C by promoting the compact structure with an increase in packing [34]. Especially enzymes from thermophiles, particularly archaeal proteins, were described to be stabilized at high temperatures and simultaneous application of pressure [35,36]. For instance, the thermostability of a protease produced by Methanococcus jannaschii increased almost 3-fold at 125 ◦ C and 500 bar [37]. Noticeably, aluminum chloride influenced the enzyme affinity rather than the maximum reaction rate. An enhanced activity in the presence of Al3+ was recently described for the ␤-glucosidase from Thermotoga thermarum [38]. Crystal structures might shed light on the interaction of aluminum with the enzyme. The good performance of the ␤-glucosidase Bgl1 at temperatures up to 105 ◦ C and under moderate pressure as well as the good tolerance toward hydrolysis products, various reagents and solvents make this enzyme an ideal candidate for application in such processes. Acknowledgments This work was supported by the German Federal Ministry of Education and Research (BMBF) in the funding cluster “Biorefinery2021”. Thanks are also due to CLARIANT, formerly Süd-Chemie, for financial support and interest. Further thanks are due to Vera Haye, Milton da Costa, Christian Kirsch and Irina Smirnova for support, advice and supply of equipment. References [1] Egorova K, Antranikian G. Industrial relevance of thermophilic Archaea. Curr Opin Microbiol 2005;8:649–55. [2] Adams MW. Enzymes and proteins from organisms that grow near and above 100 degrees C. Annu Rev Microbiol 1993;47:627–58. [3] Krahe M, Antranikian G, Markl H. Fermentation of extremophilic microorganisms. FEMS Microbiol Rev 1996;18:271–85. [4] Eichler J. Biotechnological uses of archaeal extremozymes. Biotechnol Adv 2001;19:261–78. [5] Rizk M, Antranikian G, Elleuche S. End-to-end gene fusions and their impact on the production of multifunctional biomass degrading enzymes. Biochem Biophys Res Commun 2012;428:1–5.

54

C. Schröder et al. / Enzyme and Microbial Technology 57 (2014) 48–54

[6] Kirsch C, Zetzl C, Smirnova I. Development of an integrated thermal and enzymatic hydrolysis for lignocellulosic biomass in fixed-bed reactors. Holzforschung 2011;65:483–9. [7] Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS. Microbial cellulose utilization: fundamentals and biotechnology. Microbiol Mol Biol Rev 2002;66:506–77. [8] Ketudat Cairns JR, Esen A. beta-Glucosidases. Cell Mol Life Sci 2010;67: 3389–405. [9] Bhatia Y, Mishra S, Bisaria VS. Microbial beta-glucosidases: cloning, properties, and applications. Crit Rev Biotechnol 2002;22:375–407. [10] Rondon MR, August PR, Bettermann AD, Brady SF, Grossman TH, Liles MR, et al. Cloning the soil metagenome: a strategy for accessing the genetic and functional diversity of uncultured microorganisms. Appl Environ Microbiol 2000;66:2541–7. [11] Steele HL, Streit WR. Metagenomics: advances in ecology and biotechnology. FEMS Microbiol Lett 2005;247:105–11. [12] Daniel R. The metagenomics of soil. Nat Rev Microbiol 2005;3:470–8. [13] Antranikian G, Vorgias CE, Bertoldo C. Extreme environments as a resource for microorganisms and novel biocatalysts. Adv Biochem Eng Biotechnol 2005;96:219–62. [14] Sahm K, John P, Nacke H, Wemheuer B, Grote R, Daniel R, et al. High abundance of heterotrophic prokaryotes in hydrothermal springs of the Azores as revealed by a network of 16S rRNA gene-based methods. Extremophiles 2013;17:649–62. [15] Zhou JZ, Bruns MA, Tiedje JM. DNA recovery from soils of diverse composition. Appl Environ Microbiol 1996;62:316–22. [16] Zhang K, Martiny AC, Reppas NB, Barry KW, Malek J, Chisholm SW, et al. Sequencing genomes from single cells by polymerase cloning. Nat Biotechnol 2006;24:680–6. [17] Arnold K, Bordoli L, Kopp J, Schwede T. The SWISS-MODEL workspace: a webbased environment for protein structure homology modelling. Bioinformatics 2006;22:195–201. [18] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. [19] Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem 1976;72:248–54. [20] Britton HTS, Robinson RA. Universal buffer solutions and the dissociation comstant of veronal. J Chem Soc 1931:1456–73. [21] Kirsch C, Wörmeyer K, Zetzl C, Smirnova I. Enzymatic hydrolysis of lignocellulose in a fixed bed reactor. Chem Ing Tech 2011;83:867–73. [22] Michaelis L, Menten ML. Die Kinetik der Invertinwirkung. Biochem Z 1913;49:333–69.

[23] Dixon M. The determination of enzyme inhibitor constants. Biochem J 1953;55:170–1. [24] Noguchi H. MetaGene: prediction of prokaryotic and phage genes in metagenomic sequences. Handbook of molecular microbial ecology I: Metagenomics and complementary approaches, vol. 1. Hoboken, NJ: John Wiley & Sons, Inc.; 2011. p. 433. [25] Brown JW, Daniels CJ, Reeve JN. Gene structure, organization, and expression in Archaebacteria. CRC Crit Rev Microbiol 1989;16:287–338. [26] Andersson SGE, Kurland CG. Codon preferences in free-living microorganisms. Microbiol Rev 1990;54:198–210. [27] Lao PJ, Forsdyke DR. Thermophilic bacteria strictly obey Szybalski’s transcription direction rule and politely purine-load RNAs with both adenine and guanine. Genome Res 2000;10:228–36. [28] Hurst LD, Merchant AR. High guanine-cytosine content is not an adaptation to high temperature: a comparative analysis amongst prokaryotes. Proc Biol Sci 2001;268:493–7. [29] Schleper C, Jurgens G, Jonuscheit M. Genomic studies of uncultivated archaea. Nat Rev Microbiol 2005;3:479–88. [30] Cupples CG, Miller JH. Effects of amino acid substitutions at the active site in Escherichia coli beta-galactosidase. Genetics 1988;120:637–44. [31] Vallmitjana M, Ferrer-Navarro M, Planell R, Abel M, Ausin C, Querol E, et al. Mechanism of the family 1 beta-glucosidase from Streptomyces sp: catalytic residues and kinetic studies. Biochemistry 2001;40:5975–82. [32] Jabbour D, Klippel B, Antranikian G. A novel thermostable and glucose-tolerant beta-glucosidase from Fervidobacterium islandicum. Appl Microbiol Biotechnol 2012;93:1947–56. [33] Winter R, Dzwolak W. Exploring the temperature-pressure configurational landscape of biomolecules: from lipid membranes to proteins. Philos Trans A Math Phys Eng Sci 2005;363:537–62. [34] Kirsch C, Dahms J, Kostko AF, McHugh MA, Smirnova I. Pressure assisted stabilization of biocatalysts at elevated temperatures: characterization by dynamic light scattering. Biotechnol Bioeng 2012. [35] Hei DJ, Clark DS. Pressure stabilization of proteins from extreme thermophiles. Appl Environ Microbiol 1994;60:932–9. [36] Danson MJ, Hough DW. Structure, function and stability of enzymes from the Archaea. Trends Microbiol 1998;6:307–14. [37] Michels PC, Clark DS. Pressure-enhanced activity and stability of a hyperthermophilic protease from a deep-sea methanogen. Appl Environ Microbiol 1997;63:3985–91. [38] Zhao L, Xiea J, Zhanga X, Caob F, Pei J. Overexpression and characterization of a glucose-tolerant ␤-glucosidase from Thermotoga thermarum DSM 5069T with high catalytic efficiency of ginsenoside Rb1 to Rd. J Mol Catal 2013;95:62–9.