Screening, Gene Cloning, and Characterizations of an Acid-Stable α-Amylase

Abstract

Based on its α-amylase activity at pH 5.0 and optimal pH of the crude enzyme, a strain (named B-5) with acid α-amylase production was screened. The B-5 strain was identified as Bacillus amyloliquefaciens through morphological, physiological, and biochemical characteristics analysis, as well as 16S rDNA phylogenetic analysis. Its α-amylase gene of GenBank Accession No. GU318401 was cloned and expressed in Escherichia coli. The purified recombinant α-amylase AMY-Ba showed the optimal pH of 5.0, and was stable at a pH range of 4.0-6.0. When hydrolyzing soluble starch, amylose, and amylopectin, AMY-Ba released glucose and maltose as major end products. The α-amylase AMY-Ba in this work was different from the well-investigated J01542-type α-amylase which also came from B. amyloliquefaciens. AMY-Ba exhibited notable adsorption and hydrolysis ability towards various raw starches. Structure analysis of AMY-Ba suggested the presence of a new starch-binding domain at its C-terminal region.

Introduction

α-Amylase (E.C. 3.2.1.1), an endo-type enzyme randomly cleaving α-1,4-glycosidic linkages in starch, is widely used in many commercial fields such as starch saccharification, and textile, baking, paper, and detergent industries [20]. Each application of α-amylase may require unique properties of the enzyme in terms of its substrate specificity, stability, and temperature and pH dependence due to specific processing conditions. As for starch saccharification, the main field demanding α-amylase [11], the acid-stable α-amylase, which keeps active at pH ≤5.0, is more desirable than currently used acid-labile α-amylase, which is maximally active at pH 6.0. The reason is obvious, as the starch slurry has a natural pH of approximate 4.5, using acid-stable α-amylase will save the operating cost for pH adjustment to 5.8-6.5 for liquefaction, as well as the cost for pH re-adjustment back to 4.2-4.5 for subsequent saccharification, and those costs definitely occur in the acid-labile α-amylase case [4].

As there is a huge need for acid-stable α-amylase, a considerable amount of research has been conducted over several decades, obtaining valuable acid-stable α-amylases from various microorganisms such as Bacillus acidocaldarius [10], Aspergillus kawachii [16], Bacillus licheniformis [25], Bacillus acidicola [23], Bacillus sp. Terdowsicous [2], and Bacillus HUTBS62 [1]. Attempts also have been made to engineer a neutral α-amylase towards a good performance at lower pH by methods of mutagenesis and directed evolution [15,24,27]. Despite all the above progress achieved in acid-stable α-amylase, the search for new amylases with more suitable properties is continuously encouraged.

In this work, a bacterial strain capable of producing an acid-stable α-amylase was isolated and identified. Its α-amylase gene was cloned, and investigations on the characteristics and structure of the recombinant enzyme were carried out.

 

Materials and Methods

Culture Media

The agar plate medium for bacteria isolation was as follows: soluble starch 1%, (NH4)2SO4 0.25%, MgSO4 0.02%, KH2PO4 0.3%, FeSO4·7H2O 0.0025%, and CaCl2 ·6H2O 0.025%. The shake flask medium for α-amylase production was the same as above except that 1% tryptone was supplemented. The initial pH of both media was 4.5. All chemicals used in the present work were of reagent grade, unless otherwise noted.

Scanning Electron Microscopy (SEM) of Hydrolyzed Raw Corn Starch

First, 1.5 ml of 5% (w/v) raw corn starch (washed and suspended in pH 5.0 buffer) was mixed with 0.5 ml of crude enzyme and the mixture was incubated at 55℃ with occasional shaking. At different intervals, a small amount of suspension was taken out and mounted onto the adhesive tape on a SEM stub. After drying naturally, the sample was coated with gold. Scanning electron micrographs were taken using a Hitachi S-3400NII microscope (Hitachi Co., Ltd, Japan). The accelerating voltage and the magnification are shown on the micrographs.

α-Amylase Zymogram Analysis

The crude enzyme of the selected strain was subjected to native PAGE with a 5% polyacrylamide stacking gel and an 8% polyacrylamide separating gel. After electrophoresis, the gel was cut vertically into two halves. One half was stained with Coomassie brilliant blue R-250, while the other was put in 1% (w/v) soluble starch (pH 5.0 citric acid-Na2HPO4 buffer) and incubated at 50 ℃ for 30 min. Then the gel was taken out, washed with distilled water, and soaked in iodine solution. When it turned dark blue, the gel was washed again, and the active band was observed as a clear colorless zone.

Strain Identification

The selected strain was identified and classified primarily through morphological observation, and then further identified by 16S rDNA sequence analysis, as well as by physiological and biochemical tests according to the method described in Bergey’s Manual of Determinative Bacteriology [6]. The bacterial 16S rDNA was amplified using the forward primer 5’-GGTTACCTTGTT ACGACTT-3’ and the reverse primer 5’-AGAGTTGATCCTGGC TCAG-3’. The PCR product was sequenced, and the sequence was subjected to BLAST analysis at http://www.ncbi.nlm.nih.gov/.

Cloning of α-Amylase Gene (amy)

Based on the conserved N- and C-terminal sequences of α-amylase genes from the genus Bacillus, two primers were designed as follows: forward primer 5’-GGAATTCCATATGAT GTTTGAAAAACGATT-3’ (NdeI site underlined), and reverse primer 5’-CCAATTCCTCGAGTTAATGCGGAAGAT-3’ (XhoI site underlined). An assumed amplicon of the amy gene was amplified by PCR using genomic DNA extracted from the selected strain as a template. The PCR product was cloned into vector pMD19-T (Takara Inc., Dalian, China) and then sequenced.

Sequence Annotation

The signal peptide of α-amylase was predicted with the SignalP 3.0 Server. The N-terminal catalytic domains and C-terminal postulated SBD were determined using the structure of 1bagA (PDB ID), Bacillus subtilis α-amylase as a template. To find conserved regions, catalytic residues, and potential Ca2+-binding sites, a set of 10 non-redundant α-amylase sequences belonging to the GH13 family were retrieved from the Uniprot database: P00691 (Bacillus subtilis strain 168), P41131 (Aeromonas hydrophila), P00692 (Bacillus amyloliquefaciens), P06278 (Bacillus licheniformis), P30292 (Aspergillus shirousami), Q02905 (Aspergillus awamori), P0C1B3 (Aspergillus oryzae strain ATCC42149/RIB40), P04745 (Homo sapiens (Human)), P00687 (Mus musculus (Mouse)), and P00690 (Sus scrofa (Pig)). A multiple alignment was then performed with ClustalO online (http://www.uniprot.org/align).

Production and Purification of Recombinant α-Amylase

As the postulated α-amylase was found to have a signal peptide containing the first 33 amino acids, another round of PCR was performed to remove the signal peptide by displacing the former forward primer with the present one, 5’GATCCATATGGAAAC TGCAAACAAATC 3’ (NdeI site underlined). The PCR product after double digestion by NdeI/XhoI was ligated with the NdeI/ XhoI-digested pET21a (Novagen). After being validated by DNA sequencing, the recombinant plasmid was introduced into E. coli BL21 (DE3) (Novagen).

The E. coli transformant harboring recombinant plasmids was grown at 37℃ in Luria-Bertani (LB) medium supplemented with ampicillin (100 µg/ml) for 3-4 h. When the OD600 reached 0.6-0.8, isopropyl-β-D-thiogalactopyranoside was added to a final concentration of 0.25 mM. After 4 h of induction, the cells were harvested, washed with sterile distilled water, and centrifuged. The pellet was resuspended in Tris/acetate buffer (pH 7.0) and sonicated. The lysate was centrifuged at 10,000 ×g at 4℃ for 20min to remove the debris and the supernatant was applied to a nickel resin column (Takara Inc.). Owing to the 6×His tag introduced at the C-terminal of AMY by the pET21a vector, the recombinant α-amylase was specifically bound to resin and then eluted successfully by 100-250 mM imidazole.

Enzyme Activity and Protein Concentration Assays

The activity of α-amylase was measured according to Liu and Xu [13] with a few modifications. First, 5 ml of 1% (w/v) soluble starch (pH 5.0 citric acid-Na2HPO4 buffer) was pre-incubated at 55℃ for 5 min, and then 0.5 ml of appropriately diluted enzyme sample was added. After 10 min of reaction, 0.5 ml of the reaction mixture was removed and put into a tube containing 5 ml of 0.1 M HCl. The 0.5 ml resulting solution was then taken out to mix with 5 ml of an iodine solution. The absorbance of the final solution was measured at 660nm. The enzyme inactivated by boiling for 10min was treated as a control sample. One unit of enzyme activity was defined as the amount of enzyme that causes a 1% decrease of the color of the starch/iodine solution in 1 min under the reaction condition. The protein concentration of α-amylase preparation was determined using the method of Bradford [3], with bovine serum albumin as a standard.

SDS-Polyacrylamide Gel Electrophoresis

The SDS-PAGE was performed using the method of Laemmli [12], with 5% of polyacrylamide in the stacking gel and 10% in the separating gel.

Effects of pH and Temperature on Activity and Stability of Purified α-Amylase

The optimum pH for α-amylase activity was determined in the pH range from 3.0 to 7.0 (citric acid-Na2HPO4 buffer) using the above standard assay method. For determining the pH stability, the enzyme was incubated in pH 2.5-7.0 citric acid-Na2HPO4 buffers at 40℃ for 1 h, and the residual activity was measured under the standard assay condition. Temperature dependence of α-amylase activity was measured over a temperature range of 40-90℃ at pH 5.0. Thermal inactivation of the enzyme was done by incubating the enzyme at 60℃ and 55℃ (pH 7.0, the pH of purified enzyme solution). Samples were removed at different time intervals (10-60min) and immediately cooled on ice. Residual activity was measured using the standard assay method. All assays were performed in triplicates and data were means ± standard deviations from three replications.

Effects of Metal Ions and EDTA on Purified α-Amylase Activity

α-Amylase activity was assayed in the presence of 5 mM various metal ions (K+, Ca2+, Mg2+, Mn2+, Fe2+, and Fe3+) and EDTA at the condition of pH 5.0 and 55℃. The liberated reducing sugar from soluble starch was estimated by the dinitrosalicylic acid (DNS) method [17] with glucose as a standard.

Substrate Specificity

Various substrates (amylose from potato, amylopectin from potato, and pullulan, all from Sigma-Aldrich co., USA) along with soluble starch and different raw starches were used to examine the substrate specificity of the purified α-amylase. It is necessary to mention that all sorts of raw starch must be washed 3 times with pH 5.0 buffer before use, so as to remove any kind of soluble sugars or proteins probably existing in starch. The reaction mixture consisted of 0.2 ml of purified α-amylase (0.40 mg/ml) and 1.8 ml of 1% (w/v) of each substrate (pH 5.0 buffer). After 30 min of incubation at 55℃, the liberated reducing sugars were estimated by the DNS method as described above.

Analysis of End-Products of Hydrolysis

To analyze the end-products of hydrolysis of various substrates, the same reaction mixture as used in the above section was incubated at 55℃ for 2 h. The sugars released were assayed by thin layer chromatography (TLC) using a silica gel plate in a solvent system of butanol-acetic acid-water (2:1:1). The migrated products were visualized by heating the plate at 110℃ in an oven, after it was dipped in methanol-sulfuric acid (8:2).

Adsorption Rate on Raw Starches

The washed and pre-cooled different raw starches (pH 5.0) were mixed with purified α-amylase to a final volume of 2 ml containing 5% raw starch and 0.113 mg/ml α-amylase protein. After shaking gently at 4℃ for 2 h, the sample was centrifuged to sediment raw starch granules, and the residual protein in the supernatant was measured. The adsorption rate (AR) was defined by the following equation: AR(%) = [(B-A)/B] × 100, where A indicates the residual protein after adsorption and B represents the original enzyme concentration.

 

Results

Screening of Acid α-Amylase-Producing Strain

Samples from solid-state fermented media for vinegar production were screened on pH 4.5 agar plates with starch as the sole carbon source. Based on the I2-starch clear zone surrounding the colonies, 20 strains were selected for further submerged fermentation analysis. Three strains with a high level of α-amylase activity were obtained and named B-1, B-5, and B-6. The optimal pH of crude enzyme from each strain was determined as 6.0, 5.0, and ≥7.0, respectively. Therefore, strain B-5 was chosen for further investigations.

To test whether α-amylase from strain B-5 can act on raw starch at acid pH condition, SEM was used to monitor the shape of raw corn starch granules after incubating with crude enzyme at pH 5.0 for 10-30 min. As shown in Fig. 1, a number of small holes appeared on the surface of the granules only after 10 min of digestion, and after 30 min of digestion these holes became deeper and wider. These results indicated that B-5 α-amylase had the ability to degrade raw starch granules, and thereby would be of value in the starch saccharification industry.

Fig. 1.SEM observation of raw corn starch hydrolyzed by B-5 crude α-amylase (5000×). (1) No treatment; (2-4) Incubation for 10, 20, and 30 min, respectively.

The existence of α-amylase of strain B-5 was also demonstrated by zymogram analysis (Fig. 2). Among three detectable extracellular proteins from strain B-5, only one showed α-amylase activity during the zymogram analysis, indicating that strain B-5 only produced one type of α-amylase.

Fig. 2.Zymogram analysis of B-5 strain α-amylase. 1: Total extracellular proteins detected by PAGE; 2: α-amylase zymogram detection.

Identification of Strain B-5

Strain B-5 was a gram-positive, rod-shaped, aerobic bacterium with an endospore forming in the center of the cell. It was motile by peritrichous flagella. When cultured on the agar plate, it formed a thick and wrinkled colony with an undulate edge; when cultured in a liquid medium without shaking, it would form a pellicle layer on the surface of the medium.

To determine the phylogenetic position of strain B-5, the PCR product of 16S rDNA from strain B-5 was sequenced. The size of the amplicon was 1,378 bp. A quick search with BLAST in the NCBI database clearly revealed that 16S rDNA of strain B-5 h ad 99% identity with th at of both B. subtilis and B. amyloliquefaciens. To further identify B-5’s taxonomic group, a number of physiological and biochemical tests were carried out, as shown in Table 1. These characteristics were in good agreement with the description of B. amyloliquefaciens in Bergey’s Manual of Determinative Bacteriology [6]. The strain B-5 was therefore identified to belong to the B. amyloliquefaciens group.

Table 1.Physiological and biochemical characteristics of strain B-5.

Gene Cloning, Expression, and Basic Characteristics of α-Amylase from Strain B-5

The α-amylase gene from strain B-5 (amy-Ba) was cloned directly from the total genomic DNA by PCR using primers annealing to highly conserved N- and C-terminal regions of α-amylase genes from the Bacillus genus. The gene had an ORF containing 1,980 bp nucleotides that encoded 659 amino acids, including a 33 aa signal peptide (Fig. 3). The gene had been submitted to GenBank with the accession number GU318401 (ADB81848.1 for α-amylase AMY-Ba). Conserved domain searches indicated that AMY-Ba belonged to glycoside hydrolase family 13 (GH13) and possessed four highly conserved regions of GH13 amylolytic enzymes with D217 and E2492 as its catalytic residues.

Fig. 3.Amino acid sequence and structure of AMY-Ba. The signal peptide is in gray. The four conserved regions of α-amylase are indicated within boxes. The two catalytic residues D217 and E249 are marked with a triangle ( ▲ ). Residues involved in Ca2+ binding are marked with an asterisk (*). The postulated SBD is underlined.

The BLASTX analysis showed that AMY-Ba was highly homologous with Bacillus sp. BBM1 α-amylase (ADF47479.1) with only one amino acid difference, E622K. The next closest homologs with 96%-99% of similarity were all putative α-amylases such as EYB36172.1, YP_008419732.1, YP_008411312.1, YP_001419958.1, etc., almost all from B. amyloliquefaciens strains. In view of th e fact th at none of them, including Bacillus sp. BBM1 α-amylase, was expressed heterologously and characterized up to now, it would make sense to get the amy-Ba gene in this work expressed in E. coli and characterized. Fig. 4 shows that AMY-Ba without the signal peptide was successfully produced in BL21 (DE3) cells as an approximately 66 kDa protein and the recombinant AMY-Ba was purified to homogeneity by Ni2+-affinity chromatography.

Fig. 4.SDS-PAGE analysis of recombinant α-amylase. 1: Molecular mass markers; 2: Crude extract of E. coli harboring empty vector; 3: Crude extract of E. coli harboring recombinant vector; 4: Purified recombinant α-amylase.

AMY-Ba exhibited the maximum activity at pH 5.0, though the activity between pH 5.0 and 6.2 had very little difference (Fig. 5A). After incubation at different pH buffers (pH 2.5-7.0) at 40℃ for 1 h, the enzyme could retain ≥90% of its activity at a pH range of 4.0-6.0 (Fig. 5B), showing its acid-stable property. The enzyme had an optimal temperature of 70℃ (Fig. 5C), and when incubating at 60℃ (pH 7.0) the enzyme had a time-dependent loss of activity with only 63% activity remained after 30 min incubation, whereas the enzyme was rather stable under 55℃ (Fig. 5D).

Fig. 5.Effects of pH and temperature on the activity and stability of AMY-Ba. (A) Effect of pH on AMY-Ba. (B) pH stability of AMY-Ba. (C) Effect of temperature on AMY-Ba, (D) Thermal stability of AMY-Ba at 60℃ and 55℃.

Effects of Metal Ions and EDTA on AMY-Ba Activity

Among the tested metal ions, K+, Ca2+ , and Mg2+were found to stimulate AMY-Ba activity, whereas Mn2+, Fe2+, and Fe3+ inhibited its activity to different extents (Table 2). The presence of EDTA showed an inhibitory effect on AMY-Ba activity, thus indicating the enzyme was a metalloenzyme. Bioinformatic analysis in Fig. 3 supported this result by presenting six calcium-binding sites in the AMY-Ba amino acid sequence.

Table 2.Effects of metal ions and EDTA on the activity of recombinant AMY-Ba.

Substrate Specificity and End-Product Analysis

Among all tested substrates, AMY-Ba showed the highest activity on amylopectin, followed by soluble starch and amylose (Table 3). Th e enzyme could not act on pullulan, a polysaccharide of α-1,6-bond-linked maltriose, indicating its α-1,4-glycosidic bond cleavage specificity. AMY-Ba also showed hydrolytic ability towards various raw starches. It was reasonable that all sorts of raw starch had the lower digestibility than soluble starch, amylopectin, or amylose, given that no gelatinization step was involved in the preparation of the raw starch suspension.

Table 3.Hydrolytic activity of AMY-Ba on different substrates.

Soluble starch, amylose, and amylopectin were chosen for end-product analysis. TLC (Fig. 6) showed that for hydrolysis of all these three starches, AMY-Ba yielded predominantly glucose and maltose as end-products. Besides glucose and maltose, there was a noticeable amount of one type of other larger saccharide in the hydrolysates of both soluble starch and amylopectin, yet not at all in that of amylose. Presumably, it was related with the limited product of hydrolysis of branched starches.

Fig. 6.TLC analysis of hydrolysates from different substrates. 1, Standard sugars: glucose and maltose; 2, soluble starch; 3, amylose; 4, amylopectin; 5, soluble starch without enzyme; 6, amylose without enzyme; 7, amylopectin without enzyme; and 8, enzyme solution.

Adsorption of AMY-Ba on Raw Starches

To detect the adsorption rate of AMY-Ba onto raw starch, th e mixture of equal amounts of 5% each raw starch (pH 5.0) and 0.113 mg/ml AMY-Ba solution was kept at 4℃ for 2 h. The enzyme was found to be adsorbed onto various raw starches. The adsorption rates of AMY-Ba were 44.1%, 38.5%, 35.6%, and 33.2% for sweet potato starch, potato starch, wheat starch, and corn starch, respectively (Fig. 7). Compared with the hydrolytic activities of AMYBa on different raw starches (Table 3), it was clear that the adsorption rates did not follow the same pattern as for the hydrolytic activities. The highest adsorption rate appeared in sweet potato starch, whereas the highest hydrolytic activity appeared in corn starch, suggesting that a strong adsorption on starch may involve not only biochemical adsorption between the enzyme and its substrate but also nonspecific physical adsorption, the adsorption not contributed to hydrolysis efficiency. The reason for this phenomenon was probably because raw starches from different sources had a different purity due to different starch processings, as evidenced by the SEM observations of sweet potato starch and corn starch tested (data not shown).

Fig. 7.Adsorption rate of AMY-Ba on different raw starches at 4℃ for 2 h.

 

Discussion

We cloned an α-amylase gene from the strain B-5 identified as B. amyloliquefaciens, with an ORF of 1,980 bp registered in GenBank as GU318401. The mature protein AMY-Ba overproduced in E. coli had 626 amino acids showing the approximate apparent molecular mass of 66 kDa. The enzyme had some close homologs with 96%-99% of identity derived from the same species, yet none of them was reported to be expressed heterologously to our knowledge. As for expressed and characterized α-amylase genes so far from B. amyloliquefaciens as reported by Gangadharan et al. [9], Demirkan et al. [5], Priyadharshini et al. [21], and Liu et al. [14], they were found to be either identical or highly similar to the B. amyloliquefaciens strain 1H α-amylase gene (GenBank J01542) [19,26], which was 1,542 bp long and coded for 514 amino acid residues, giving a mature protein of 483 amino acids. There was no significant sequence similarity found between AMY-Ba in this work and J01542, though both enzymes belonged to the GH13 family. Moreover, they differed in some catalytic properties: AMY-Ba showed its optimal pH and optimal temperature at pH 5.0 and 70°C, respectively, whereas J01542-type α-amylase had optimal pH and temperature of pH 5.0 and 50°C according to Gangadharan et al. [8,9], and of pH 6.0 and 55°C according to Demirkan et al. [5], indicating that they fell into two distinct types of α-amylases from B. amyloliquefaciens. Comparing the end-products of soluble starch by the two types of α-amylase mentioned above, it was clear that AMY-Ba released glucose and maltose as major end-products, whereas J01542-type α-amylase released maltose or maltooligosaccharides predominantly, with no or only a minor amount of glucose [5,8], suggesting different action patterns that the two types of α-amylase followed.

Amylase’s degradation ability on raw starch is closely related to the presence of its starch-binding domain (SBD), usually localized at the C-terminal end of the enzyme [22]. Through the SBD, the enzyme can bind to insoluble starch granules, which is thought to increase the local substrate concentration at the enzyme’s active site and hence enhance the hydrolysis ability on raw starch. AMY-Ba was found to adsorb to various raw starches, and therefore the presence of a SBD was suggested in AMY-Ba. Since the homology modeling of AMY-Ba could easily identify its Nterminal structure using 1bagA (PDB ID) as a template, clearly showing three domains (Domains A, B, and C) αamylases generally contain, we guessed its C-terminal extension (472-659 amino acids) might be the SBD. Although no putative conserved domain was found in the C-terminal region by NCBI Protein BLAST, nor any hit to known carbohydrate-binding modules (CBMs) with Hidden Markov Models search [7], a homologous match of 2C3W in PDB derived from Bacillus halodurans C-125 α-amylase’s CBM was finally found to be capable of being a modeling template. As 2C3W was a member of CBM family 25, the C-terminal module of AMY-Ba (i.e., the postulated SBD) was presumed to belong to CBM25. Sequence alignment of this module with the SBD of BAA22082.1, the only case in CBM25 whose SBD function had been demonstrated, showed 41% identity within a short fragment from AMYBa’s 592 to 629 amino acids. Interestingly, in Munoz et al.’s [18] homology model of ADF47479.1, with one amino acid difference to AMY-Ba, the enzyme had two additional domains besides Domains A, B, and C, designated as Domains D and E. Th ey suggested th at both Domains D and E might be the SBDs. Therefore, further research is necessary in order to clarify whether the C-terminal region of AMY-Ba contains one type of SBD or two tandemed SBDs. Furthermore, research on the effects of these modules on raw starch hydrolysis will be of great significance.