An Approach for Lactulose Production Using the CotX-Mediated Spore-Displayed β-Galactosidase as a Biocatalyst

Abstract

Currently, enzymatic synthesis of lactulose, a synthetic prebiotic disaccharide, is commonly performed with glycosyl hydrolases. In this work, a new type of lactulose-producing biocatalyst was developed by displaying β-galactosidase from Bacillus stearothermophilus IAM11001 (Bs-β-Gal) on the surface of Bacillus subtilis 168 spores. Localization of β-Gal on the spore surface as a fusion to CotX was verified by western blot analysis, immunofluorescence microscopy, and flow cytometry. The optimum pH and temperature for the resulting spore-displayed β-Gal was 6.0 and 75℃, respectively. Under optimal conditions, it showed maximum activity of 0.42 U/mg spores (dry weight). Moreover, the spore-displayed CotX-β-Gal was employed as a whole cell biocatalyst to produce lactulose, yielding 8.8 g/l from 200 g/l lactose and 100 g/l fructose. Reusability tests showed that the spore-displayed CotX-β-Gal retained around 30.3% of its initial activity after eight successive conversion cycles. These results suggest that the CotX-mediated spore-displayed β-Gal may provide a promising strategy for lactulose production.

Introduction

Lactulose [β-ᴅ-galactosyl-(1 → 4)-fructose] is an artificial disaccharide with numerous health advantages. Currently, commercial lactulose production is generated by chemical isomerization of lactose in an alkaline medium [2]. The chemical process, however, suffers from serious drawbacks such as high energy consumption, high cost, and removal of byproduct and chemical waste. Therefore, research efforts are increasing to find alternative methods for converting lactose into lactulose. In this context, the enzymatic method shows great potential for lactulose production by using biological catalysts owing to its highly safe and eco-friendly features. The lactulose production was generally carried out with purified enzymes (β-galactosidase (β-Gal), glucose isomerase, or cellobiose 2-epimerase) [36] or permeabilized cells containing enzymes [39]. However, enzyme isolation and purification is time-consuming and expensive. Furthermore, purified enzymes can be very unstable and quickly lose their activities in harsh environments. Some of the problems associated with purified enzymes can be avoided by using permeabilized whole cells instead, but this approach is still hampered by the mass transfer limitation caused by the microbial cell membrane [6]. Therefore, the development of new cost-efficient biocatalysts and biocatalytic processes with industrial practicability for enhanced lactulose production is nowadays more and more preferred.

Recently, bacterial surface display has emerged as a powerful technique with a wide range of potential applications in many fields, such as drug delivery vehicles, whole-cell biocatalysts, biosensors, environmental remediation, and biomolecule screening platforms [20,33]. Owing to the enhanced stability and high safety, spore display has been developed as a unique molecular display platform. In comparison with other Bacillus species used, B. subtilis provides more benefits for the development of spore display systems because of the detailed knowledge about its spore structure [5,9]. The spore coat of B. subtilis is a multilayered structure, which consists of four distinct layers: the basement layer, inner coat, outer coat, and crust [14,25]. Increasing knowledge of the spore coat structure helps to find new anchoring motifs for use in developing spore surface display systems. So far, four outer coat proteins (CotB [15], CotC [24], CotG [17], CotE [12]), and one crust protein (CotZ [11]) have been employed as carriers for successful display of antigens or proteins of biological interest on the surface of B. subtilis spores. Considering both the safety and low cost, B. subtilis spore display has valuable application in the development of whole-cell biocatalysts such as Neu5Ac aldolase [7], β-Gal [19], ʟ-arabinose isomerase [23], and meta-cleavage product hydrolases [29].

Previously, it was shown that urease subunit A (UreA) from Helicobacter acinonychis could give different levels of display efficiency when fused with different carriers (CotB, CotC, and CotG) [10]. Similar results were observed for the FliD protein of Clostridium difficile [26] as well as α-amylase Q and GFPuv [27]. In the case of UreA, CotB gave the highest display efficiency, whereas the CotG-UreA fusion protein was not displayed on the spore but also appeared to be partially processed. In addition, another recent study reported that incorporation of the E. coli CotE-LacZ β-Gal fusion protein resulted in the loss of the recombinant B. subtilis DB104 spore wall integrity [12]. These results clearly indicated that not all coat proteins could be employed as efficient carriers for the development of a successful B. subtilis spore display.

The low efficiency of B. subtilis spore display and the high interest in industrial applications stress the strong need to improve technologies that explore more repertoires of B. subtilis spore coat proteins, especially crust proteins, as carriers. With increased understanding of the structure of the spore coat and advances in biotechnology engineering, CgeA, CotY, CotZ, and CotX have recently been recognized as crust proteins [14]. The surface location of these proteins and their accessibility to antibodies make them promising candidates for the development of B. subtilis spore displays. Very recently, our group demonstrated that a tetrameric β-Gal from Bacillus stearothermophilus IAM11001 (Bs-β-Gal) could be actively presented on the surface of B. subtilis 168 spores by fusing it with either CotY or CotZ [37]. Herein, we report the successful use of CotX as a carrier for correctly targeting Bs-β-Gal to the spore surface. The properties of the spore-displayed and purified β-Gals were further compared. We also show that lactulose production can be achieved with the spore-displayed β-Gal.

 

Materials and Methods

Chemicals

Lactose, fructose, o-nitrophenol (ONP), and o-nitrophenyl-β-ᴅ-galactopyranoside (ONPG), supplied by Sinopharm Chemical Reagent (China), were of analytical grade and directly used without additional purification. All the HPLC standards were provided by Sigma-Aldrich (China). Other chemicals used for this study were of analytical grade and available commercially.

Bacterial Strains, Plasmids, and Transformation

The bacterial strains and plasmids used are described in Table S1. In this work, B. subtilis 168 (trpC2) (1A1, Bacillus Genetic Stock Center) was used as a host for displaying β-Gal. B. stearothermophilus IAM11001 was used as a source of DNA for PCR amplification of the β-Gal gene (bgaB). The integrative plasmid pJS700a, a kind gift from Dr. Degang Ning (Jiangsu University, China) was used as a display vector [22]. Plasmids were transformed into bacterial strains using standard procedures: CaCl2-mediated transformation of E. coli competent cells [30] and two-step transformation of B. subtilis [28].

Plasmid Construction and Chromosomal Integration

All the primers used are listed in Table S2. The cotX gene was amplified from genomic DNA of B. subtilis 168 by PCR using a pair of specific primers cotX-F and cotX-R. The PCR product containing the CotX promoter and coding region was inserted into the XbaI/KpnI sites on plasmid pJS700a to form pJS-cotX. The gene bgaB encoding B. stearothermophilus IAM11001 β-Gal was amplified from its genomic DNA with the primer pair bgaB-F and bgaB-R. After digestion with KpnI/EcoRI, the fragments were ligated into the corresponding sites in plasmid pJS-cotX to generate pJS-cotX-bgaB, carrying a 3.1 kb DNA fragment encoding for a CotX-β-Gal fusion protein under control of the native cotX promoter.

The cotB, cotC, or cotG genes were amplified by PCR using genomic DNA from B. subtilis 168 as the template and the corresponding primers. The plasmids pJS-cotB, pJS-cotC, pJS-cotG, pJS-cotB-bgaB, pJS-cotC-bgaB, and pJS-cotG-bgaB were constructed in the same way as pJS-cotX and pJS-cotX-bgaB. All the integrative plasmids mentioned above were separately linearized by BglII digestion and then transformed into B. subtilis 168 (trpC2) competent cells. After selection for erythromycin (10 μg/ml) resistant clones on agar plates, the recombinant strains were identified and designated PX102 (expressing the CotX protein), PB102 (CotB), PC102 (CotC), PG102 (CotG), PXB103 (CotX-β-Gal), PBB103 (CotB-β-Gal), PCB103 (CotC-β-Gal), and PGB103 (CotG-β-Gal), respectively.

Preparation of Spores

The spores were prepared by following the method described by Wang et al. [37]. Extraction of coat proteins from the suspensions of B. subtilis strain (PB102, PC102, PG102, PX102, PBB103, PCB103, PGB103, and PXB103) spores (>1 × 1010 spores/ml) was performed by SDS-DTT treatment, as previously reported [4].

Western and Dot-Blot Analyses

To confirm the expression of β-Gal protein on the surface of B. subtilis spores, the extracted spore coat proteins of each strain (PB102, PC102, PG102, PX102, PBB103, PCB103, PGB103, and PXB103) were fractionated by SDS-PAGE and then electrically transferred to a nitrocellulose membrane (Millipore Corporation, MA, USA), and then probed with rabbit polyclonal anti-β-Gal antibody. After incubation with HRP-conjugated goat anti-rabbit IgG antibody (Invitrogen Life Technologies, NY, USA), signals were visualized using the enhanced HRP-DAB chromogenic substrate kit (Tiangen Biotech Co., Beijing, China) according to the manufacturer’s guidelines. A quantitative measurement of β-Gal enzyme molecules expressed on B. subtilis spores was obtained by dot-blot experiment [37].

Immunofluorescence Microscopy

B. subtilis 168 vegetative cells and the purified spores surface-expressing the CotX-β-Gal fusion protein were assessed by immunofluorescence microscopy as previously reported [37] using rabbit polyclonal anti-β-Gal antibody.

Flow Cytometric Analysis

For flow cytometric analysis, about 1 ml of the spore suspensions of each strain (PX102 and PXB103) was washed three times with PBS and incubated for 1 h at 37℃ with anti-β-Gal antibody diluted in PBS containing 1% BSA. After four washes in PBS, Alexa Fluor 488-conjugated goat anti-rabbit IgG antibody was added and the incubation was continued for another 1 h at room temperature. Samples were then washed three times with PBS and resuspended in 1 ml of PBS. Fluorescently labeled spores were analyzed with a BD FACSCalibur flow cytometer (Becton Dickinson Biosciences, USA). Data were processed using the CellQuest Pro software.

Comparison of Four Different Spore-Displayed β-Gals toward the Hydrolysis of ONPG

In order to evaluate the efficiency of β-Gals displayed on the spores by using four different spore coat proteins as carriers, another three engineered strains, PBB103, PCB103, and PGB103, were constructed as described above. After induction in DSM, the cultures of B. subtilis strains (PBB103, PCB103, PGB103, and PXB103) were collected by centrifugation and then resuspended in 20 mM phosphate buffer (pH 6.0). Finally, the prepared spore suspension was tested for its activity to hydrolyze ONPG.

Purification of β-Gal and Antibody Production

The bgaB gene-specific primers bgaB-1 and bgaB-2 were designed to amplify a 2,019 bp bgaB gene from B. stearothermophilus. The obtained PCR product was TA-cloned initially into the pMD18-T vector to give pMD18-T-bgaB. After digestion with NdeI and XhoI, the fragment was ligated into the same sites of pET-28a to produce pET-28a-bgaB. Following transformation of E. coli BL21(DE3) competent cells (Novagen Inc., USA) with pET-28a-bgaB, expression of recombinant β-Gal was induced by the addition of 0.8 mM IPTG at 30℃ for 10 h. After the induced cells were harvested, cell pellets were lysed using a Vibra Cell 72405 sonicator (Sonics and Materials Inc., USA). The soluble His-tagged protein was purified on a Ni2+ chelating Sepharose HP chromatography column (GE Healthcare Bio-sciences AB, Sweden). For antibody production, the purified His-tagged β-Gal was used to immunize New Zealand white rabbits following the established protocol of Sambrook and Russell [30] using Freund’s adjuvant. All animal experiments were approved by the Ethical Committee of Jiangnan University (JN No. 20120106-0120).

Activity Assay of Spore-Displayed β-Gal

The measurement of spore-displayed β-Gal activity was conducted according to the method of a previous report [37]. The spores of wild-type B. subtilis 168 served as a negative control. One unit of spore-displayed β-Gal was considered as the enzyme quantity that liberated 1 μmol of ONP per minute in the defined conditions.

Biochemical Characterization of Spore-Displayed β-Gal (CotX-Based β-Gal)

The influence of pH on the activity and stability of spore-displayed β-Gal was investigated by determining relative activity using different pH buffers at a concentration of 0.1 M (phosphate citrate buffer for pH 5.0-7.0; potassium phosphate buffer for pH 7.0-8.0; and Tris-HCl buffer for pH 8.0-9.0). For pH stability, the sample was pre-incubated in buffers of different pH values for 2 h at 4℃, and the residual activity was measured using the standard assay procedure. The optimum temperature of spore-displayed β-Gal was determined at different temperatures in the range of 50-80℃. The thermostability of spore-displayed β-Gal was not tested at temperature higher than 75℃, because our preliminary experiments showed that most of the enzyme activity was lost only after 10 min of incubation. To determine the thermostability, the purified spore suspension was pre-incubated at different temperatures (60℃, 65℃, and 70℃) for 2.5 h and aliquots were withdrawn at different time intervals for determination of the residual enzyme activity.

Production of Lactulose Using the Spore-Displayed β-Gal (CotX-Based β-Gal) as a Whole-Cell Biocatalyst

The biotransformation reaction were carried out in a final volume of 10 ml containing 200 g/l lactose, 100 g/l fructose, and the anchored CotX-linked β-Gal at 75℃ for 12 h. At regular intervals, samples (200 μl) were drawn from the reaction mixture for the measurement of lactulose conversion. In the reusability experiments, the biocatalyst were separated and recycled by centrifugation. The new reactants were added into the repeated-batch reactors containing the separated spore-displayed β-Gal. The sugars lactose, fructose, glucose, galactose, and lactulose in the assay solution were measured by HPLC following the procedure described by Wang et al. [38]. The yield of lactulose was calculated using the following equation:

where Clacto and Clactu represent the initial concentration of lactose and lactulose concentration after the reaction, respectively.

 

Results

Construction and Chromosomal Integration of Gene Fusion

The DNA fragment encoding CotX was amplified by PCR and ligated into the plasmid pJS700a, generating pJS-cotX. Then, the DNA fragment bgaB encoding Bs-β-Gal was subcloned into pJS-cotX, generating pJS-cotX-bgaB (Fig. S1). In an effort to objectively compare the efficiency of β-Gal displayed on the spores via CotX with the other three successful coat proteins CotB, CotC, and CotG, the genes cotB (encoding CotB), cotC (encoding CotC), and cotG (encoding CotG) were first amplified from the genomic DNA of B. subtilis 168 and then cloned into the same plasmid pJS700a, yielding pJS-cotB, pJS-cotC, and pJS-cotG. Next, the recombinant plasmids pJS-cotB-bgaB, pJS-cotC-bgaB, and pJS-cotG-bgaB were constructed in the same way as pJS-cotX-bgaB. Finally, the obtained fusion genes were integrated by a double cross-over event into the amyE site of the B. subtilis 168 chromosome. Correct integration for each transformation was verified by checking the absence of amylase activity and PCR (data not shown).

Surface Display of Fusion Protein on the Recombinant Spores

To confirm that the cotX-bgaB fusion gene was expressed within the B. subtilis spore coat, extracted coat proteins were analyzed by western blot with the primary antibody anti-β-Gal. For the comparison, the extracted coat proteins of each strain (PBB103, PCB103, and PGB103) were also tested for the presence of the respective fusion protein by western blot analysis. As shown in Fig. 1, a main band displaying an apparent molecular mass of about 85 kDa was detected in the coat protein extracts from all four recombinant strains, whereas no bands with the same size in the control strains (PB102, PC102, PG102, and PX102) were observed (Fig. 1, lanes 1-4), indicating that each fusion protein was expressed on the spore coat. The results presented here demonstrated that CotX like the other three carrier proteins is able to anchor β-Gal on B. subtilis spores.

Fig. 1.Detection of the presence of Bs-β-Gal by western blot. The extracted coat proteins were resolved on SDS-PAGE and reacted with anti-β-Gal antibody. Lane M, protein ladder; lanes 1-4, coat proteins extracted from purified spores of B. subtilis strains PB102, PC102, PG102, and PX102, respectively; lanes 5-8, coat proteins extracted from purified spores of B. subtilis strains PBB103, PCB103, PGB103, and PXB103, respectively; lane 9, purified E. coli-expressed recombinant Bs-β-Gal.

Immunofluorescence Microscopy and Flow Cytometric Analyses

Immunofluorescence microscopy was used to assess the possible surface localization of CotX-linked β-Gal fusion protein on B. subtilis 168 spores. As illustrated in Fig. 2A, a significant increase in the fluorescence intensity was detected on the purified spores harboring the fusion protein, whereas no obvious fluorescence was observed with B. subtilis strain PX102 spores, indicating that the CotX-β-Gal fusion protein was successfully presented on the surface of B. subtilis spores via fusion with CotX.

Fig. 2.Evaluation of spore-displayed Bs-β-Gal by immunofluorescence microscopy (A) and flow cytometric analysis (B). Preparation of spores were obtained from cultures of B. subtilis strains PX102 and PXB103. Samples were labeled with rabbit polyclonal anti-β-Gal antibody, followed by Alexa Fluor 488-conjugated goat anti-rabbit IgG antibody. (A) BF and IF, bright-field and immunofluorescence images, respectively. Scale bar, 10 μm. (B) A total of 10,000 spores were detected for each experiment.

In order to further substantiate the findings obtained from immunofluorescence microscopy, flow cytometry was employed for analysis of the surface expression of the CotX-β-Gal fusion protein on the spores. As shown in Fig. 2B, under flow cytometry, a significant increase in fluorescence was observed for the recombinant spores expressing the CotX-β-Gal fusion as compared with the control strain, evidencing the presence of β-Gal on the B. subtilis spore surface.

Efficiency of Expression

An investigation into the possibility of CotX as a carrier for displaying Bs-β-Gal on the spores generated the expected results. Next, the efficiency of β-Gal displayed on the spores via fusion with CotB, CotC, CotG, and CotX was further examined by comparing their activities toward the hydrolysis of ONPG. As expected, all four recombinant spores surface-expressing β-Gal displayed activity ranging from 0.14 to 0.46 U/mg spores (Fig. 3). Interestingly, the CotC-based spore-displayed β-Gal exhibited the lowest activity, whereas the CotG-based recombinant protein showed the highest activity. The activities of spore-displayed β-Gals from PBB103 and PXB103 were found to be 0.30 and 0.42 U/mg spores, respectively (Fig. 3). These results clearly demonstrated that both CotX and CotG may be appropriate carriers for the surface display of Bs-β-Gal on the spore.

Fig. 3.Comparison of the activities of four spore-displayed Bs-β-Gals toward the hydrolysis of ONPG. WT, wild-type B. subtilis 168 spores. Each assay was repeated three times, and the results are expressed as the mean ± SD.

The amount of Bs-β-Gal expressed on the B. subtilis spore coat via fusion with different carriers was quantified by dot-blot analysis. Based on this quantification for all examined strains, CotX-β-Gal accounted for 0.26% of total coat proteins extracted, CotB-β-Gal for 0.19%, CotC-β-Gal for 0.06%, and CotG-β-Gal for 0.29% (Table 1). The concentrations of extracted protein from spores of the four recombinant strains were 4.9 ± 0.15 mg/ml for PXB103, 4.6 ± 0.21 mg/ml for PBB103, 4.4 ± 0.17 mg/ml for PCB103, and 5.1 ± 0.12 mg/ml for PGB103. Hence, the estimated numbers of fusion protein molecules per spore were 1.6 × 103 for PXB103, 8.1 × 102 for PBB103, 3.8 × 102 for PCB103, and 1.7 × 103 for PGB103 (Table 1).

Table 1.Quantification of Bs-β-Gal expression by dot-blot analysis.

Effects of pH and Temperature on the Activity of Spore-Displayed β-Gal by Fusion to CotX

Fig. 4 shows the influences of pH on the activities of spore-displayed and purified β-Gals. The optimal pH of spore-displayed β-Gal was shown to be 6.0 using ONPG as a substrate (Fig. 4A), whereas the purified E. coli-expressed Bs-β-Gal exhibited an optimal pH of 6.5 (Fig. 4B). However, as previously reported, Bs-β-Gal has an optimum pH of 7.0 when expressed in B. subtilis WB600 [3]. As indicated in Fig. 4A, the enzyme had good stability over a pH range of 5.5-9.0. The enzyme activity was almost completely lost in potassium phosphate buffer (pH 8.0); on the contrary, nearly 60% of its initial activity was retained in Tris-HCl buffer at the same pH. This result suggested that potassium phosphate buffer may have an inhibitory effect on the enzyme activity.

Fig. 4.Biochemical characterizations of spore-displayed and purified Bs-β-Gals. Effect of pH on the activities and stabilities of spore-displayed (A) and purified (B) Bs-β-Gals. Enzyme activity was assayed at 75℃ in buffers of varying pH values. The relative activity was expressed as a percent of the maximum activity in the range of pH 5.5 to 9.0. For pH stability, the assay was carried out immediately after 2 h incubation at different pH value buffers. (C) Effect of temperature on the activities of spore-displayed and purified Bs-β-Gals. Temperature dependences of the activities of spore-displayed and purified Bs-β-Gals were determined at 55-80℃. The maximal activity obtained was referred to as 100%. (D) Effect of temperature on the stabilities of spore-displayed and purified Bs-β-Gals. The enzyme was stored at 60℃, 65℃, and 70℃ for 2.5 h and residual activity was measured at pH 6.0. Untreated enzyme was used as a control and its activity was taken as 100%.

The temperature profiles of spore-displayed and purified β-Gals were analyzed and compared over the temperature range of 55-80℃. As shown in Fig. 4C, the optimum temperature for activity of spore-displayed β-Gal was 75℃. However, the purified E. coli-expressed Bs-β-Gal displayed optimal activity at 70℃, which was similar to that of B. subtilis-expressed Bs-β-Gal [3]. Thus, the optimal temperature of spore-displayed β-Gal was 5℃ higher than that of the native enzyme. Fig. 4C also shows that the enzyme activity of spore-displayed β-Gal increased slightly with increasing temperature (<75℃). However, a sharp decline in enzyme activity was observed at a higher temperature (80℃). The stability of enzyme is one of the most important factors that limit their industrial application. The spore-displayed β-Gal was incubated at temperatures of 60℃, 65℃, and 70℃ for 2.5 h to examine its thermostability. As illustrated in Fig. 4D, the spore-displayed β-Gal exhibited a slightly higher heat resistance than the purified enzyme under the same condition.

Biocatalytic Production of Lactulose Using the CotX-Based Spore-Displayed β-Gal

We investigated lactulose production using lactose, fructose, and spore-displayed CotX-based β-Gal to explore the potential application of anchored β-Gal as a whole-cell biocatalyst. For lactulose production, 10 ml of 200 g/l lactose and 100 g/l fructose were incubated with 3.5 U/ml of spore-displayed β-Gal. Samples were withdrawn every 1 h for up to 12 h and their sugar compositions were analyzed. As shown in Fig. 5A, the concentration of lactulose increased sharply in the 0 to 2 h phase and peaked at 4 h, and then declined slightly with longer incubation. Moreover, the maximum production of lactulose by spore-displayed β-Gal was 8.8 g/l, which was significantly higher than that of the free enzyme. These results indicate that the spore-displayed β-Gal had high potential as a whole-cell biocatalyst for lactulose production.

Fig. 5.Biocatalytic production of lactulose using the CotX-based spore-displayed β-Gal. (A) Effect of reaction time on lactulose production from lactose and fructose by free and spore-displayed Bs-β-Gals. (B) Reusability of spore-displayed Bs-β-Gal on lactulose synthesis. The lactulose concentration (8.8 g/l) of the first batch was assigned to be 100%.

The reusability of the spore-displayed enzymes is of considerable importance in industrial applications. To evaluate the reusability of the spore-displayed β-Gal, it was subjected to eight successive runs of lactulose production. Fig. 5B shows that the relative catalytic activity of spore-displayed β-Gal decreased gradually with increasing number of reuse times. It was also observed that a relatively high yield of lactulose production could be obtained within four reaction cycles. However, the catalytic efficiency of spore-displayed β-Gal declined rapidly to 30.3% after eight reuses, and only 2.7 g/l lactulose was produced.

 

Discussion

β-Gal is one of the best investigated enzymes for the use in the food industry, and β-Gal from E. coli fused to various carriers has successfully been displayed on the surface of Bacillus species spores [12,19]. In this study, by fusion of a tetrameric Bs-β-Gal to CotX, an active enzyme could be displayed on the spore surface. Furthermore, the biochemical properties of the fusion protein was also investigated. Very recently, such a Bs-β-Gal has been demonstrated with success for construction of B. subtilis spore display when fused with either CotY or CotZ [37], but the highest display efficiency was achieved by fusion to CotX.

Although all four coat proteins (CotX, CotB, CotC, and CotG) enabled display of Bs-β-Gal on the spore surface, their activities were in the order CotG-β-Gal> CotX-β-Gal> CotB-β-Gal> CotC-β-Gal. The difference in enzyme activities of the four types of spore-displayed β-Gals may possibly be elucidated by assuming the correlation of the enzyme activity and location site of these fusion proteins. It was the observation by Imamura et al. [13] that CotB and CotC are localized approximately center of the outer coat layer, and both are more inside as compared with the CotX protein. As a result, the CotB- and CotC-based β-Gal fusion proteins have to cross through the crust layer and the layer between the outside and the center of the outer coat, which may lead to some loss of enzyme activity of β-Gal. Meanwhile, the location of CotB or CotC within the coat may hinder the substrate to gain access to the active site of the spore-displayed β-Gal. In an attempt to overcome this problem, the CotX-mediated spore display system was applied. Using this approach, a higher level of enzyme activity was obtained with the CotX protein than that of the CotB- or CotC-based β-Gal. To the best of our knowledge, the exact location of the CotG protein within the outer coat is unknown, but it plays a key role in the assembly of the CotB protein [40]. Interestingly, of all anchoring motifs reported to date, the CotG protein has been proven to be a preferred carrier for the development of whole-cell biocatalysts. Based on our current data, the CotG-linked β-Gal showed the highest activity (0.46 U/mg spores), with a slight enhancement (less than 10%) in enzyme activity over the CotX-based β-Gal.

The very first enzyme that has been exposed at the spore surface is the β-Gal of E. coli [19]. It was shown that the enzyme activity of β-Gal displayed on the surface of B. subtilis WB700 spores was 5 U/mg spores (dry weight) when fused to CotG. In addition to E. coli β-Gal, LacA β-Gal from B. subtilis 168 has been expressed as a fusion to CotC and demonstrated to be presented on the surface of B. subtilis RH101 (deleted for the cotC gene) spores. Moreover, the recombinant spores expressing CotC-LacA β-Gal had an estimated activity of 0.18 U/1010 spores [34]. Interestingly, our previous results showed the possibility for improved display efficiency of Bs-β-Gal at the spore surface either through the use of pDG1664, or through pEB03 (high-copy number) as the expression vector. As expected, the enzyme activities of spore-displayed Bs-β-Gals were relatively higher (0.54 U/mg spores for pDG1664 and 1.34 for pEB03) (data not shown). Taken together, these findings further evidenced that one successful spore-surface display example may be influenced by factors such as availability of a novel efficient carrier, the size and properties of the displayed protein, the type of expression vector and host strain, as well as simple and efficient technologies to correctly anchor the displayed protein at the spore coat.

Bioconversion of lactose into lactulose by β-Gal has been gaining a great deal of attention. However, cost-effective production of lactulose from lactose remains a major challenge. Typically, high cost is associated with the large quantity of enzymes required for the complete conversion of lactose into lactulose. A comparison of lactulose production from lactose and fructose by different forms of β-Gals (free, immobilized, or permeabilized cells containing enzyme) from various sources is summarized in Table 2. Enzymatic production of lactulose was first reported in the late 1970s [35], and in subsequent years it has been successfully prepared by various β-Gals, with achieved Ylactu/lacto varying from 5% to 32.5% (see Table 2) [1,8,16,18,21,31,32]. Utilizing Aspergillus oryzae β-Gal is effective in improving the production of lactulose [1,8], but the cost is high because complex purification procedures are required to obtain free enzyme. In pursuit of improving the economic attractiveness of the enzymatic method of lactulose synthesis, Lee et al. [21] developed an alternative approach, which uses ethanol-permeabilized cells of Kluyveromyces lactis ATCC 8585 as a source of β-Gal, and found that a lower lactulose yield (5%) was obtained with 400 g/l of lactose and 200 g/l of fructose. B. subtilis spore display has been shown to be a highly efficient system for surface-expression of various heterologous enzymes. Furthermore, reutilization of the B. subtilis spores enables reuse of the enzymes displayed on their spore surface without reproduction of the B. subtilis spores, which would reduce the cost of B. subtilis spore germination as well as enzyme addition. Thus, B. subtilis spore surface engineering is a promising technology for the production of lactulose. In this study, we developed the CotX-mediated B. subtilis spore-display system for anchoring Bs-β-Gal and investigated the catalytic capability of this system for lactulose production from lactose. As shown in Table 2, the spore-displayed β-Gal produced 8.8 g/l lactulose with a yield of 4.4%. In addition, the thermostability of spore-displayed β-Gal was slightly higher than its native form. The favorable temperature activity of spore-displayed β-Gal and its thermostability are desirable properties for industrial application.

Table 2.Comparison of lactulose production by spore-displayed β-Gal and other free or immobilized β-Gals.

Because surface-displayed enzymes can be prepared inexpensively and easily by simple cultivation, the present study provides, for the first time to our knowledge, an experimental basis for the use of spore-displayed β-Gal as a cost-effective whole-cell biocatalyst system for lactulose production.