Secretive expression of heterologous β-glucosidase in Zymomonas mobilis

Abstract

Background

Zymomonas mobilis is an efficient ethanol fermentation strain, but its narrow substrate range limits its fermentation in lignocellulose hydrolysate. As a potential consolidated bioprocessing (CBP) stain for bioethanol production, the ability of cellulose utilization was necessary. In this study, extracellular expression of β-glucosidase on Z. mobilis was studied as the first step for construction of a practical CBP strain to reduce the use of β-glucosidase in the cellulase components.

Results

The heterologous β-glucosidase from Bacillus polymyxa was expressed in the ethanologenic strain Z. mobilis (ZM4) and secreted extracellularly by an endogenous signal peptide and a fusion protein. The signal peptide SP1086 of the endoglucanase gene ZMO1086 from Z. mobilis was identified and facilitated 12 % of the endoglucanase encoded by ZMO1086 from Z. mobilis ZM4 and 16 % of the β-glucosidase encoded by bglB gene secreted out of the membrane of Z. mobilis ZM4. Another method for enhancement of the β-glucosidase secretion is to fuse the β-glucosidase encoded by bglB with the levansucrase encoded by sacB from Z. mobilis ZM4 to achieve the secretive expression. Its expression level was enhanced two times but only showed a 2 % secretion ratio in this situation.

Conclusions

The SP1086 signal peptide showed an obviously secreting capacity of the β-glucosidase protein. The fusion protein with SacB also showed the secretion effect, but it was less efficient.

Background

Consolidated bioprocessing (CBP) integrates cellulase production, cellulose hydrolysis, and ethanol fermentation into one single unit; thus, the expensive cellulase enzyme production step is removed or the usage is reduced [1]. When cellulase enzyme is produced by CBP cells, the cellulase should be in effective contact with a solid cellulosic substrate for the hydrolysis to occur. To achieve this goal, not only the cellulase enzymes are sufficiently expressed but also the enzymes should be secreted extracellularly onto the periplasm space of the cell or directly to the extracellular medium. The practice of CBP concept was carried out on Saccharomyces cerevisiae strains by displaying cellulase enzymes on the surface [2].

Zymomonas mobilis is an efficient ethanol fermentation strain with low cell biomass generation, high-ethanol yield, and the tolerance to the toxicity of the final product, which makes it into a suitable candidate for CBP strain [3–5]. For the CBP process, unlike endoglucanase and exoglucanase acting on a solid cellulose substrate, β-glucosidase catalyzes the hydrolysis of soluble cellobiose, which easily passes through the glucan cell wall and contacts the enzymes on the periplasmic space of the cell [6, 7]. Therefore, either the extracellular expression into the fermentation broth or the trans-membrane expression in periplasmic space of β-glucosidase could work well for hydrolyzing cellobiose into glucose. In this work, the extracellular expression of β-glucosidase on Z. mobilis was tested as the first step for construction of a practical CBP strain to reduce the use of β-glucosidase in the cellulase components.

It has been tried to express cellulase genes in Z. mobilis [8–13], but their secretion was difficult; thus, a cellulase-expressing Z. mobilis without secretion is not able to function as a CBP cell in lignocellulose biorefining processes. Selecting native signal peptides or secretory pathways is a useful vehicle for enhancing the cellulase secretion on Z. mobilis. The signal peptide of the ZMO0130 gene encoding alkaline phosphatase and the ZMO0131 gene encoding hypothetical protein from Z. mobilis had been used to secret endocellulase and carboxymethyl cellulase from Acidothermus cellulolyticus, and the secretion ratio was more than 40 % of the total proteins [9]. To our knowledge, the secretion of heterologous β-glucosidase had not been successfully tested in Z. mobilis.

Several strategies for secretively expressing β-glucosidase in Z. mobilis ZM4 were investigated. First, a putative homologous endoglucanase gene ZMO1086 [14] in Z. mobilis ZM4 was overexpressed to ensure the secretive function of its signal peptide SP1086 by using a strong enolase promoter Peno. Then, the signal peptide SP1086 of ZMO1086 was used for secretively expressing a heterologous β-glucosidase gene bglB from Bacillus polymyxa in Z. mobilis ZM4. Finally, a fusion protein of β-glucosidase with levansucrase (SacB) from Z. mobilis ZM4 was also constructed to utilize its original secretion mechanism of levansucrase. The results showed that these methods significantly improved the secretion of β-glucosidase in Z. mobilis and provided a way for further practical Z. mobilis-based CBP strain.

Methods

Strains and plasmids

The strains used in this study are shown in Table 1. Escherichia coli DH5α was used in gene clone and plasmids construction. Z. mobilis ZM4 (ATCC 31821) was the host strain for the recombinant strain. Plasmid pHW20a is the shuttle vector with different gene expression [15].

Table 1 Strains and plasmids used in this study

Reagents and chemicals

The restriction enzymes were purchased from Fermentas (Vilnius, Lithuania). The T4 DNA ligase and PrimerSTAR HS DNA polymerase used in PCR were purchased from Takara (Dalian, China). The Tryptone and yeast extract were from Oxoid (Cambridge, UK). The Bacterial DNA kit was from Omega Bio-Tek (Norcross, GA, USA). The PCR purification kit, plasmid mini kit, and gel extraction kit were from Sangong Biotech (Shanghai, China). All other chemicals used in this study were purchased from Sinopharm Chemical Reagent (Shanghai, China).

Gene cloning

The primers used in this study are shown in Table 2. The construction of recombinant plasmids is shown in Fig. 1. The bglB gene fragment was amplified from the genome of B. polymyxa CGMCC 1.794. The ZMO1086 gene fragment was amplified from the genome of Z. mobilis ATCC 31821. NprB signal peptide fragments were amplified from the genome of Clostridium thermocellum DSM 1237.

Table 2 Primers used in this study

Fig. 1

Construction of plasmids. Construction of plasmids pNprB-bglB (express β-glucosidase gene bglB with promoter Peno and signal peptide NprB), pbglB (express β-glucosidase gene bglB with promoter Peno), pZM1086 (express ZMO1086 gene with promoter Peno), pSP1086-bglB (express β-glucosidase gene bglB with promoter Peno and signal peptide SP1086), and pSacB-bglB (express fusion protein SacB-bglB with promoter Peno)

The enolase promoter (Peno) PCR product was cleaved, gel purified, and ligated into pHW20a by the restriction enzyme site EcoRI and BamHI and then constructed the plasmid pHW20a-Peno. The bglB- and ZMO1086-gene-amplified fragment was located at the downstream of the promoter Peno by the same method and got the plasmid pbglB and pZM1086 (Fig. 1). Fragment Peno-SP1086 was amplified from pZM1086, and the PCR product was cleaved, gel purified, and ligated into pHW20a by the restriction enzyme site EcoRI and BamHI to get the plasmid pHW20a-Peno-SP1086. Then, the bglB fragment was inserted and located at the downstream of the promoter Peno and signal peptide by the same method and got the plasmid pSP1086-bglB (Fig. 1). The Peno-NprB fragment was amplified by PCR and ligated into pHW20a by the restriction enzyme site EcoRI and BamHI and then constructed the plasmid pHW20a-Peno-NprB. The bglB fragment was located at the downstream of the promoter Peno and signal peptide NprB and got the plasmid pNprB-bglB (Fig. 1).

The primers p1 and p2 were used to get the sacB fragment and p3 and p4 to get the bglB fragment for the overlap PCR. Then, p1 and p4 were used to get the production sacB-bglB fragment by the overlap PCR. Then, the overlap fragment was located at the downstream of the promoter Peno to get the plasmid pSac-bglB (Fig. 1).

Enzyme activity assay

The recombinant strain was 1 % inoculated and cultivated in 20 mL RM medium with 20 μg/mL tetracycline and 30 μg/mL nalidixic acid, stationary culture at 30 °C for 24 h. The cells were centrifuged at 10,000×g for 5 min at 4 °C, and the supernatant was collected. The proteins in the supernatant were precipitated using ammonium sulfate at 60 % (w/v) saturation and redissolved in 1 mL 50 mM buffer (pH 6.0) to give the extracellular enzyme solution. The enzyme extract in periplasm and cytoplasm was obtained by osmotic shock method [16]. The cell was suspended in STE buffer (20 % sucrose, 30 mM Tris–HCl (pH 8.0), 1 mM EDTA) and incubated on ice for 10 min and then centrifuged at 8000 rpm for 10 min. The supernatant was discarded and the cell resuspended by water. And then, the cell was incubated on ice for 10 min again and centrifuged at 8000 rpm for 10 min to obtain the periplasmic fraction in supernatant. The cell pellet was crushed to make the cytoplasmic fraction.

The CMC cellulolytic activity was determined by using carboxymethyl cellulose (CMC) as the substrate. The activity was measured by incubating 1 mL of enzyme extract with 1 mL of 10 g/L CMC in 50 mM Tris–HCl buffer (pH 6.0) at 30 °C for 1 h. The reducing sugars released were estimated by using the dinitrosalicylic acid (DNS) reagent. The enzyme activity was expressed in U/g dry cell weight (DCW), and one unit (U) was defined as the 1 g/L reducing sugars released per min.

The enzymatic activity of bglB was determined by using p-nitrophenyl-β-d-glucopyranoside (pNPG) as the substrate. The pNPG-hydrolyzing activity was estimated by incubating 1 mL of enzyme extract with 1 mL of 8 mM p-NPG in 50 mM citric acid buffer (pH 6.0) at 37 °C for 15 min. To stop the reaction, 1 mL 0.5 M Na₂CO₃ was added. The optical density at 405 nm (OD_405nm) of the solution was determined to obtain the release of p-nitrophenol (p-NP). The enzyme activity was expressed in U/g DCW, and one unit (U) was defined as the μmol of p-nitrophenol released per min.

Fermentation and media

The RM medium contained 10 g/L yeast extract, 2 g/L KH₂PO₄, and 20 g/L glucose. The glucose can be replaced to some other carbon source in fermentation.

Z. mobilis and recombinant strains were 1 % inoculated and cultivated in 20 mL RM medium with 30 μg/mL nalidixic acid (20 μg/mL tetracycline was used additionally for the recombinant strains), stationary culture at 30 °C in the flask. The component of the fermentation broth was assayed by HPLC (LC-20AD, refractive index detector RID-10A, Shimadzu, Kyoto, Japan) fitted with a Bio-Rad Aminex HPX-87H column (Hercules, CA, USA) at 65 °C. The mobile phase was 5 mM H₂SO₄ at 0.6 mL/min.

Results and discussion

Expression of the homogenous cellulase gene ZMO1086 in Z. mobilis

The ZMO1086 gene is a putative cellulolytic gene with an endogenous signal peptide SP1086 according to the Z. mobilis ZM4 genome annotation, providing the carboxymethyl cellulose (CMC) cellulolytic activity but with very low activity [9–11]. To test the secretive expression performance of SP1086, the ZMO1086 expression was enhanced by replacing its original promoter with Peno, a strong promoter in Z. mobilis ZM4. The recombinant plasmid pZM1086 was constructed to overexpress the ZMO1086 gene in Z. mobilis ZM4 (Fig. 1). The ZMO1086 gene fragment was located at the downstream of the promoter Peno, and the signal peptide SP1086 was included in the CDS of ZMO1086 at the 5′ end. The recombinant plasmid was introduced into Z. mobilis by conjugation method to give a recombinant Z. mobilis ZM4/pZM1086.

The CMC cellulolytic activity of the recombinant Z. mobilis ZM4/pZM1086 was measured by using CMC as the substrate. Figure 2 shows that the whole cell CMC activity of Z. mobilis ZM4/pZM1086 was at 6.50 U/g dry cell weight (DCW), approximately 25 folds greater than that of the control strain which was only 0.27 U/g DCW. Approximately 90 % of the CMC activity was located in the cytoplasm (5.36 U/g DCW), while the activity in the periplasmic space and extracellular medium only accounted for 6.31 % (0.41 U/g DCW) and 5.90 % (0.38 U/g DCW) of the whole cell activity, respectively. The results indicated that the secretive expression of ZMO1086 using the signal peptide in Z. mobilis ZM4 was successful although the secretion ratio was not satisfactory. In the previous CMCase expression works, the absence of a signal peptide in Z. mobilis led to the absence of CMC activity in the extracellular space [11, 12].

Fig. 2

The CMC cellulolytic activity of Z. mobilis ZM4/pZM1086. The activity was determined by using CMC as the substrate at 30 °C for 1 h. The strain Z. mobilis ZM4/pHW20a was used as the controls

Secretive expression using the signal peptide SP1086

From the expression of ZMO1086, we found that the cellulase can be secretively expressed in a restricted form. The signal peptide SP1086 demonstrated its ability of transporting the cellulase protein from cytoplasm onto the extracellular space of Z. mobilis ZM4. In the next step, SP1086 was used in the secretive expression of the heterologous β-glucosidase gene bglB from B. polymyxa CGMCC 1.794 in Z. mobilis ZM4. The bglB fragment was inserted at the downstream of promoter Peno and the signal peptide SP1086 (Fig. 1), then plasmid was introduced into Z. mobilis ZM4 by conjugation to give a recombinant Z. mobilis ZM4/pSP1086-bglB.

The β-glucosidase activity of the recombinant Z. mobilis ZM4/pSP1086-bglB was measured by using pNPG as the substrate. Figure 3 showed that the β-glucosidase activity of the whole cell was 2.48 U/g DCW, and the control strain was approximately 0.16 U/g DCW, indicating the correct expression of bglB in Z. mobilis ZM4. Approximately 16 % of the β-glucosidase enzyme was secreted. The β-glucosidase activity in the cytoplasm was 1.91 U/g DCW, accounting for more than 80 % of the total protein. The β-glucosidase activity in the periplasm was 0.35 and 0.05 U/g DCW in the extracellular medium and accounted for 14.11 and 2.02 % of total proteins, respectively. Compared to the secretive expression of ZMO1086 using the same promoter and signal peptide, the proportion in the periplasmic space was more than doubled from 6.31 to 14.11 %, but the extracellular proportion of β-glucosidase was only half of ZMO1086.

Fig. 3

β-glucosidase activity of the strains Z. mobilis ZM4/pSP1086-bglB and Z. mobilis ZM4/pSac-bglB. The β-glucosidase activity was determined using pNPG as substrate at 37 °C for 15 min. The strain Z. mobilis ZM4/pHW20a was used as the control

Although the secretion ratio of β-glucosidase was still low, it was a significant improvement in Z. mobilis ZM4 compared to the bglB expression by the signal peptide NprB from C. thermocellum, which was almost no secretion (data not shown) although NprB was effective in other strains [17, 18]. The constructed plasmids pNprB-bglB (which attempted to express bglB with promoter Peno and signal peptide NprB) and pbglB (which attempted to express bglB with promoter Peno) (Fig. 1) were also not successful, except the present signal peptide SP1086. It is speculated that the expression and transportation of cellulase enzymes may affect the metabolism of the Z. mobilis cells due to their rigorous peptidoglycan cell wall under the strong influence of cellulase existence. The signal peptide from Z. mobilis ZM4 may mediate the transportation of cellulase enzyme out of the cytoplasm and distribute the cellulase in a unique way without disturbance or injury to the sensitive cell wall of Z. mobilis cells.

Secretive expression by fusion protein

Levansucrase gene sacB from Z. mobilis ZM4 is responsible for the sucrose hydrolysis into glucose and fructose in Z. mobilis ZM4 strain. Since sucrose is unable to transport through the cell membrane of Z. mobilis [19], the enzyme should be located in a position of the membrane and cell wall in contact with sucrose. From the published genome, no signal peptide for levansucrase was reported or annotated in the genome information of Z. mobilis ZM4. It is hypothesized that the protein structure of levansucrase encoded by sacB behaves with a trans-membrane property thus ensuring contact with the soluble sucrose for the hydrolysis. In this study, a fusion protein SacB-bglB linked by seven glycine groups was constructed by taking advantage of the secretion mechanism of SacB. The plasmid pSac-bglB was constructed and introduced into Z. mobilis ZM4 by conjugation to give the recombinant Z. mobilis ZM4/pSac-bglB.

The β-glucosidase activity of the recombinant ZM4/ pSac-bglB was measured by using pNPG as the substrate. Figure 3 showed that the whole expression of β-glucosidase was 5.26 U/g DCW, approximately two folds greater than that of Z. mobilis ZM4/pSP1086-bglB. However, the secretive proportion was not obviously promoted. The β-glucosidase activity was 0.05 and 0.06 U/g DCW in the periplasm space and in the extracellular medium, accounting for 1.0 and 1.1 % of the whole cell activity, respectively. The result indicated that the fusion protein strategy was not feasible for BglB secretion in Z. mobilis ZM4.

Conclusively, Z. mobilis-ZM4/pSac-bglB-expressing fusion protein showed the highest β-glucosidase expression level in the cell, while Z. mobilis ZM4/pSP1086-bglB secreted most of the enzyme proteins into the periplasm up to 0.35 U/g DCW. The extracellular activities of two recombinants were similar, near to 0.05 U/g DCW. The use of the signal peptide SP1086 showed the better effect in secretive expression, and the fusion protein showed the highest expressive level of BglB but with less enzyme proteins secreted into periplasmic space.

The substrate-utilizing ability of the recombinants Z. mobilis ZM4/pSP1086-bglB and Z. mobilis ZM4/pSac-bglB was tested by using 10 g/L cellobiose and 10 g/L glucose as the carbon source with Z. mobilis ZM4 harboring the empty pHW20a used as the control strain. Both the recombinants utilized glucose at a similar rate, and all the glucose was consumed in the fermentation, and the growth of Z. mobilis ZM4/pSP1086-bglB was faster with a slightly higher DCW at the end of fermentation (data not shown). Figure 4 showed that the control strain Z. mobilis ZM4/pHW20a did not use any cellobiose, but the recombinants Z. mobilis ZM4/pSP1086-bglB and Z. mobilis ZM4/pSac-bglB also utilized only 0.4 and 0.2 g/L of cellobiose, respectively, while ethanol titer did not show obvious difference; the results indicated that, although the extracellular β-glucosidase activity was detected in the constructed recombinants, it was still not sufficient to give the effective CBP fermentation performance. Further enhancement of the β-glucosidase activity is required to give the practical CBP capacity.

Fig. 4

Utilization of cellobiose by Z. mobilis ZM4/pSacB-bglB and Z. mobilis ZM4/pSP1086-bglB. Z. mobilis ZM4/pHW20a was used as the control strain. White circle for Z. mobilis ZM4/pHW20a, gray for Z. mobilis ZM4/pSP1086-bglB and black for Z. mobilis ZM4/ pSacB-bglB. 1 % of the strain culture were inoculated and cultivated in 20 mL RM medium with 1 with 30 μg/mL nalidixic acid and 20 μg/mL tetracycline, stationary culture at 30 °C in the flask

The exogenous signal peptide SP1086 realized the secretive expression of a heterologous β-glucosidase in Z. mobilis ZM4 and showed the more efficient secretive proportion than the fusion protein, which caused a higher total expression level with reducing secretion. As a potential CBP strain, both the recombinant strains Z. mobilis ZM4/pSP1086-bglB and Z. mobilis ZM4/pSac-bglB did not utilize cellobiose well, and some further improvement should be done to meet the CBP fermentation.

Conclusions

In this study, we overexpressed ZMO1086 in Z. mobilis and confirmed its signal peptide SP1086 could work for the secretion in Z. mobilis. On this basis, we realized the secretion of heterologous β-glucosidase gene bglB in Z. mobilis ZM4/pSP1086-bglB in which about 16 % of the total protein was secreted. Meanwhile, the recombinant Z. mobilis ZM4/pSac-bglB which expressed a fusion-protein-combined β-glucosidase with levansucrase (SacB) also showed extracellular β-glucosidase activity with 2 % secretion ratio. In the fermentation by using RM media with 10 g/L glucose and 10 g/L cellobiose as the carbon source, Z. mobilis ZM4/pSP1086-bglB utilized 0.4 g/L cellobiose and Z. mobilis ZM4/pSac-bglB utilized 0.2 g/L cellobiose during a 72-h fermentation. It was not an efficient result, and the secretion of β-glucosidase in Z. mobilis ZM4 needs improvement in further research.