Characterization of the β-glucosidase NpaBGS
The cloned NpaBGS cDNA contains 2,331 bp and the deduced amino acid sequence has 776 amino acid residues, with a theoretical molecular mass of 85.1 kDa and isoelectric point of 4.4 (Additional file 1
). The cleavage site for the putative signal peptide is located between residues Ala20 and Ile21. Three potential N
-glycosylation sites were predicted at residues 51, 223 and 532. Conserved domain prediction suggests that it is a β-glucosidase of the glycosyl hydrolase Family 3 (GH3
) carrying a GH3
N-terminal domain (Pfam00933) and a GH3
C-terminal domain (Pfam01915) located at residues 62
270 and 350
Based on amino acid sequences, glucosidases have been classified into several families, with most of the β-glucosidases belonging to either GH1
]. Comparative analyses using BlastX and ExPASy indicated that NpaBGS contains two conserved putative domains of GH3
. The aspartic acid residue Asp251 in the conserved domain (GXVMXD) might be the active-site residue of NpaBGS [21
Purification of NpaBGS expressed inPichia pastoris
The expression of the NpaBGS cDNA in P. pastoris GS115 was conducted under the control of the inducible promoter AOX1 on the pPICZ A vector. A 2-day culture in a medium containing 0.5 % methanol resulted in optimal enzyme production. These conditions were used in subsequent large scale cultures for enzyme purification and the results are summarized in Table . The pattern of fractions containing NpaBGS activities at each purification step was checked by SDS-PAGE (Figure ). A significant mass of the purified enzyme was estimated to be 85 kDa by SDS-PAGE analysis. The enzyme was purified 2.9-fold with a specific activity of 34.5 U/mg against cellobiose as the substrate. In addition, the β-glucosidase activity of the purified NpaBGS was examined by the zymogram assay with 4-methylumbelliferyl-β-D-cellobioside (MUC) staining after electrophoresis on the native PAGE (Figure ).
Summary of stepwise purification of NpaBGS (β-glucosidase)
Figure 1 SDS-PAGE and Zymogram of NpaBGS, which was expressed and purified from aPichia pastorisrecombinant strain. (A) SDS-PAGE of NpaBGS after each purification step. Lane M, protein marker; lane 1, crude extract; lane 2, condensation; lane 3, ammonium sulfate (more ...)
The evidence from enzymatic assay, bioinformatics and homology modeling studies also strongly suggest that NpaBGS is a member of GH3
. In NpaBGS, three putative N
-glycosylation sites were found (Additional file 1
), but no significant difference was observed between the calculated and the apparent molecular weight on SDS–PAGE using the Pichia
expression system (Figure ). Unlike reported cases of glycosylated β-glycosidases [12
], there might be no or little glycosylation on NpaBGS. Furthermore, NpaBGS has also been successfully expressed in Escherichia coli, Bacillus subtilisKluyveromyces marxianus
and K. lactis
(data not shown). The enzyme with no or little glycosylation would be easier to express in both eukaryotic and prokaryotic systems, except for species such as S. cerevisiae
that have a strong glycosylation mechanism. It will be interesting to do site-directed mutagenesis to study whether any of the three sites are actually glycosylated.
The purified NpaBGS and commercial enzyme Novo 188 were used to evaluate the effects of pH and temperature on enzymatic activity using cellobiose as the substrate. The maximum activities of NpaBGS and Novo 188 were observed at 40°C and 60°C, respectively, and NpaNGS was found to display a higher activity than Novo 188 at 40°C, though a lower activity at 60°C (Figure ). Both enzymes showed about 80 % residual activity at 50°C. At pH 6.0, NpaBGS showed a relative activity of 58 % at 30°C and 86 % at 50°C. The effect of pH on the hydrolysis rate was also evaluated for NpaBGS in a reaction system incubated at 40°C for 1 h. High levels of NpaBGS activity were found in a narrow pH range (5.0-6.0), peaking at pH 6.0 (Figure ). At pH values outside the range of 5.0 to 8.0, the relative activities decreased significantly. In contrast, Novo 188 showed an acidity-tolerance at pH 4. The activities of NpaBGS under conditions at different pH values and temperature indicated that NpaBGS prefers weak acidity as the yeast fermentation condition.
Figure 2 Effects of temperature (A) and pH (B) on the activity of purified NpaBGS and Novo 188. The relative activities are expressed as percentage normalized to the sample with the highest activity in each test. The profile of NpaBGS was shown with a solid line, (more ...)
Substrate specificity of NpaBGS
The NpaBGS substrate specificity was determined using the purified protein to avoid the background β-glucosidase activity of P. pastoris. All substrates were assayed at 40°C and pH 6 for 1 h. As shown in Table , enzyme activities were detected for natural substrates such as amygdalin, arbutin, larminarin, phenyl-β-D-glucoside, β-gentiobiose as well as cellobiose. It showed little or no activity for other sugars such as sinigrin, maltose, lactose and sucrose. For synthetic substrates, MUC and MUD were found to have the best activity, but no activity was detectable for MUG (data not shown). Moreover, when MUC was used, the enzymatic product generated measurable fluorescence, indicating that this enzyme was active on short cello-oligosaccharides (Figure ).
Substrate specificity of purified NpaBGS
Most β-glucosidases can be divided into three groups with respect to their substrate specificity: (a) those that exhibit a high specificity towards aryl β-D-glucosides, (b) those that preferentially hydrolyze cellobiose and cello-oligosaccharides (also known as cellobiases), and (c) those that hydrolyze both types of substrates (i.e., broad-specificity β-glucosidases) [24
]. From our experiments, NpaBGS displayed a higher activity on 4-methylumbelliferyl substrates than on cellobiose, demonstrating that NpaBGS has a very high affinity for methylumbelliferyl substrates (Table ). In addition, NpaBGS hydrolyses CMC (Additional file 2
) and MUC (Table ), suggesting that the enzyme possesses both endo-glucanase and β-glucosidase activities [25
]. Thus, this enzyme showed a strong β-glucosidase activity and might also possess other cellulase functions.
Effects of different elements on the NpaBGS activity
To study the effects of metal ions and reducing agents on the NpaBGS activity, we conducted assays in the presence of metal cations, including Al3+, Ca2+, Cu2+, Fe3+, Mg2+, Mn2+ and Zn2+, and reducing agents, such as DTT and β-mercaptoenthanol, at the concentrations of 1 and 10 mM (Table ). At the concentration of 1 mM, only Mg2+, Mn2+ and Zn2+ showed significant enhancement of β-glucosidase activity compared to the reaction containing the chealator EDTA. When the concentration was increased to 10 mM, Al3+, Cu2+ and Fe2+ all showed significant inhibition of enzymatic activity; Al3+ and Cu2+ almost abolished the function. In contrast, Ca2+, Mg2+, Mn2+ and Zn2+ showed significant enhancement of β-glucosidase activity, especially for Mg2+ and Mn2+, which showed very strong enhancement (Table ). Supplementing either DTT or β-mercaptoenthanol showed no apparent effects on enzymatic activity under the condition tested.
Effect of metal ions and reducing agents on the activity of purified NpaBGS
We examined whether Mg2+ and Mn2+ could enhance pH or temperature tolerance of NpaBGS. We found no effect of these ions on pH tolerance (data not shown), but addition of either cation increased NpaBGS’s activity at higher temperatures (Figure ). Interestingly, significant enhancement of the activity by the supplement of Mn2+ was found at 50°C over those assayed at 40°C, suggesting the potential of in vitro application in digesting cellulose at elevated temperatures by supplementing enhancing cations.
The effects of Mg2+and Mn2+cations, to a final concentration of 1 and 10 mM, on the activity of purified NpaBGS at pH 6.0. The activity of NpaBGS (open square), NpaBGS with Mg2+ (gray square) and NpaBGS with Mn2+ (black square) were compared.
According to previous studies, supplementing divalent or trivalent cations is one strategy to increase enzyme reaction efficiency, thermostability or termination reaction. Different levels of inhibitory effect on enzyme activities by metal ions have been reported, especially Cu2+
for β-glucosidase [24
]. In a previous study, Ca2+
showed several benefits including protein conformation stabilization, higher affinity for the substrate, and a higher thermostability of an endoglucanase of Clostridium thermocellum
]. However, Mg2+
and especially Mn2+
showed a stronger enhancement than Ca2+
on the β-glucosidase activity of NpaBGS (Table & Figure ). In our tests, Al3+
showed a significant inhibition of NpaBGS activity at 10 mM (Table ). This indicated that the requirement of free sulfhydryl groups in NpaBGS is similar to those observed in other β-glucosidases [33
]. The sequence data also suggested that there are thiol groups presented at the active site that are involved in binding or catalysis, or that such groups are essential for maintaining a proper tertiary structure of the enzyme [34
]. This is consistent with the suggestion that the cysteine residue is involved in the stability and activity of β-glucosidases [37
]. On the other hand, we found the stimulatory effect of ions on NpaBGS (Table ) as reported in other studies [28
]. This stimulation has been regarded as promoting a significant reduction in the binding specificity and/or deactivation of the site [39
Performance of NpaBGS in Simultaneous Saccharification and Fermentation
Since cellobiose digestion by β-glucosidase might be affected by the end-product feedback inhibition, it is not easy to compare the hydrolytic efficiencies of NpaBGS and Novo 188 (data not shown). We therefore employed a SSF process to compare the efficiencies of NpaBGS and Novo 188. We added an equal amount (2 units) of NpaBGS or Novo 188 separately to 10 ml of yeast cultures with 2 % cellobiose as the sole carbon source and compared their effects on yeast growth. The brewers' yeast Saccharomyces cerevisiae
BY4741 and the thermo-stable kefir yeast K. marxianus
] were employed in separate experiments. Both cultures showed a better growth profile in cell density with the addition of purified NpaBGS than with the addition of Novo 188 at 30°C (Figure ). K. marxianus
KY3 was also studied at higher temperatures, i.e., 37°C and 40°C. It grew better at 37°C than at 40°C or 30°C (Figure ).
Figure 4 The performances of NpaBGS and Novo 188 in SSF at different temperatures. Equal units (2 units) of the two enzymes were added separately to 10 ml yeast cultures with 2 % cellobiose as the sole carbon source and their effects on yeast growth were compared. (more ...)
The effects of NpaNGS and Novo 188 on ethanol fermentation in SSF at different temperatures were also studied. When the yeasts were inoculated in SSF at 30°C, the yeast cultured with NpaBGS showed a better performance in ethanol conversion than the yeast cultured with Novo 188; the same results were observed for both yeast hosts KY3 and S. cerevisiae
(Figure ). In addition, because 40°C is the optimal condition for ethanol production by K. marxianus
] and for the enzyme reaction of NpaBGS (Figure ), a higher ethanol productivity was observed at 40°C than at 37°C and 30°C (Figure ). Although the yeasts stopped producing ethanol after 12 h of culturing, more than 50 % cellobiose was converted to ethanol by K. marxianus
KY3 in the SSF process at 37°C and 40°C. The efficiency of cellobiose digestion by either NpaBGS or Novo 188 was significantly lowered after 12 hours of reaction (data not shown), but the digested sugar was still sufficient to support a weak growth of the yeasts (Figure ). These data indicated that removing the feedback inhibitor (glucose) by yeasts could enhance the activity of β-glucosidase, as reflected by the ethanol productivity by both S. cerevisiae
BY4741 and K. marxianus
KY3. Note that the ethanol production rate at 40°C and the growth rate at 37°C of K. marxianus
KY3 with 2 % cellobiose by the NpaBGS treatment was done using 2 % glucose as the sole carbon source (data not shown). These SSF results indicated that NpaBGS had a significantly higher efficiency for SSF ethanol production by both yeast hosts than Novo 188 at all the temperatures tested, probably due to the faster cellobiose-digestion rate of NpaBGS than Novo 188 under the temperatures tested (30-40°C). In summary, our data indicates that purified NpaBGS is active under a wide range of conditions, with the maximum activity at 40°C in the weak acid condition (pH 5.0-6.0). Moreover, with 1 unit of enzyme, NpaBGS showed a good efficiency in completely converting 2 % cellobiose to glucose within 4 hours in the optimal buffer system (Figure ). The time course assay with the two enzymes for cellobiose digestion was also examined at 40°C and pH 6.0, and the data indicated that NpaBGS had a slightly better efficiency than Novo 188.
Figure 5 The time course assay for cellobiose digestion by NpaBGS or Novo 188. Both enzymatic assays were conducted by incubating each enzyme at 40°C for different durations in 100 mM in Tris–HCl buffer (pH 6.0) and then measured their β-glucosidase (more ...)
Cellulosic ethanol conversion from napiergrass
In the current process of SSF, the commercial Celluclast 1.5 L is usually employed for converting cellulose to cellobiose or oligosaccharides. Although Celluclast 1.5 L has high exo- and endo-glucanase activities, it has a very low β-glucosidase activity. Therefore, it is necessary to add a β-glucosidase, such as Novo 188, to convert cellobiose to glucose. Instead of adding β-glucosidase, one may transform a β-glucosidase gene into the host. For this purpose, we succeeded in expressing the NpaBGS gene in K. marxianus KY3, and the new strain was employed in SSF using dry napiergrass (2 %) as the feeding stock. While the wild-type KY3 alone could not produce any detectable amount of ethanol, the engineered strain KY3-NpaBGS could produce 1.09 mg/ml ethanol, probably by using the oligosaccharides of napiergrass, such as cellobiose and cellodextrin (Figure ). Furthermore, as a SSF experiment we added 2 ml of Celluclast 1.5 L to 50 ml yeast cultures and compared their effects on KY3 and KY3-NpaBGS alcoholic fermentation. Again, KY3 produced no detectable amount of ethanol, while the strain KY3-NpaBGS could produce 3.32 mg/ml ethanol from napiergrass in one day aerobic culturing at 40°C (Figure ). Although the efficiency of ethanol conversion was low, as a proof of concept the experiment showed that the NpaBGS gene has a potential to be applied in SSF or even in CBP if we also introduce both exo- and endo-glucanase genes into the host. The efficiency of ethanol production of SSF can be improved by using an anaerobic system, increasing the copy number of the NpaGNS genes and immobilizing the enzymes on the cell surface.
Figure 6 The ethanol productivity of the SSF experiment using 2 % dry napiergrass as the solo carbon source. A 2 ml of Celluclast 1.5 L was added to 50 ml yeast cultures and two different strains, KY3 and KY3-NpaBGS, were inoculated in one day aerobic culturing (more ...)
For practical applications such as the separate hydrolysis and fermentation (SHF) process, most of the commercial cellulases show a higher efficiency and thermostability at temperatures higher than 40°C [41
]. In bio-fuel industry, SSF and simultaneous saccharification and co-fermentation (SSCF) are the two major processes currently used for cellulosic-ethanol production. However, most efficient microbes for ethanol production such as brewer’s yeast live below 40°C. For highly efficient SSF or SSCF, the enzymes should have a large capacity to digest the substrates in order to provide carbon sources for the microbes to grow under the same culture condition. A previously characterized rumen cellulase showed an optimal activity at pH from 4.0 to 7.0 and temperature from 35°C to 50°C [18
]. In our experiments, two alcoholic fermentation yeasts were chosen to study the performances of NpaBGS and Novo 188 in SSF at different temperatures. Cell density, glucose concentration and ethanol concentration were used to evaluate the SSF efficiency. Although Novo 188 showed better digestion efficiency and thermostability than NpaBGS at higher temperatures (Figure ), NpaBGS appeared to be superior to Novo 188 at 40°C. Since 40°C is almost the highest temperature for microbe growth and fermentation, NpaBGS has a potential for a direct application in a bio-reactor system and for CBP.
Although the SSF concept has already been published in many previous studies, most of these studies focused on the enzyme cocktail and the enzyme/substrate blending proportion. Indeed, the related enzyme technology focused on enhancing the thermo-stability or increasing the acid tolerance of the cellulolytic enzymes. In this study, we isolated a new beta-glucosidase and showed that it has advantages over the commercial enzyme Novo 188 in SFF applications and that it can be transformed into a host to replace the use of Novo 188. Since NpaBGS was isolated from one of the best nature SSF system, the buffalo rumen, which undergoes microbial fermentation in 38°C, it may have an advantage for constructing an artificial SSF system for yeast ethanol production. We considered a friendly environment for both host growth and the enzyme reaction. This new concept has the potential to reduce the cost as it requires no external supply of beta-glucosidase and it might improve the efficiency of a bio-process, such as SSF or CBP.