Domain-swapping of mesophilic xylanase with hyper-thermophilic glucanase

Background Domain fusion is limited at enzyme one terminus. The issue was explored by swapping a mesophilic Aspergillus niger GH11 xylanase (Xyn) with a hyper-thermophilic Thermotoga maritima glucanase (Glu) to construct two chimeras, Xyn-Glu and Glu-Xyn, with an intention to create thermostable xylanase containing glucanase activity. Results When expressed in E. coli BL21(DE3), the two chimeras exhibited bi-functional activities of xylanase and glucanase. The Xyn-Glu Xyn moiety had optimal reaction temperature (Topt) at 50 °C and thermal in-activation half-life (t1/2) at 50 °C for 47.6 min, compared to 47 °C and 17.6 min for the Xyn. The Glu-Xyn Xyn moiety had equivalent Topt to and shorter t1/2 (5.2 min) than the Xyn. Both chimera Glu moieties were more thermostable than the Glu, and the three enzyme Topt values were higher than 96 °C. The Glu-Xyn Glu moiety optimal pH was 5.8, compared to 3.8 for the Xyn-Glu Glu moiety and the Glu. Both chimera two moieties cooperated with each other in degrading substrates. Conclusions Domain-swapping created different effects on each moiety properties. Fusing the Glu domain at C-terminus increased the xylanase thermostability, but fusing the Glu domain at N-terminus decreased the xylanase thermostability. Fusing the Xyn domain at either terminus increased the glucanase thermostability, and fusing the Xyn domain at C-terminus shifted the glucanase pH property 2 units higher towards alkaline environments. Fusing a domain at C-terminus contributes more to enzyme catalytic activity; whereas, fusing a bigger domain at N-terminus disturbs enzyme substrate binding affinity.


Background
Enzyme is important in biomass conversion and renewable energy production. However, biotechnological condition demands for thermostable enzymes. Thermostability can be enhanced by fusing proper protein fragments, such as, the non-homologous fragments [1], the thermophilic xylanase homologous N-terminus [2,3], the hyper-thermophilic carbohydrate binding-module [4], and the Pseudoalteromonas arctica esterase OsmC domain [5]. Two catalytic domains, such as, xylanase and glucanase, can also be combined to create bi-functional chimeras [6][7][8]. However, the domain fusions were limited at xylanase one terminus. When fused at the other terminus, the Clostridium thermocellum xylanase and Thermotoga maritima glucanase activities were damaged [7][8][9]. Thereby, domains need to be swapped with each other in fusion investigation.
Xylanase (EC3.2.1.8) is widely used in feed, flour, baking industry, pulp bleaching, etc [3,10,11]. An Aspergillus niger xylanase (Xyn) has high catalytic activity (Gen-Bank: EU375728) [12]. The 185 residue enzyme is more suitable for domain fusion investigation, because it is the smallest xylanase in GH11 family. However, the Xyn optimal temperature for activity (T opt ) is 47°C, and its thermal inactivation half-life (t 1/2 ) at 50°C is 17.6 min [4]. To break down complicated biomass [13], the mesophilic fungal Xyn needs enhanced thermostability and synergistic glucanase activity. An ideal partner is the T. maritima glucanase (Glu) (EC3.2.1.4), because it is hyper-thermophilic and 258 residue big (Met 1 -Glu 258 ) [14,15]. Linker-peptide is also important for enzyme thermostability [16,17], because each domain needs necessary space to form to active conformation [4]. A linker-peptide (Pro 692 -Gly 713 ), which was selected from the hyper-thermophilic T. maritima xylanase A [10], was confirmed having enough space for connecting two domains [4]. Thereby, the 22 residue linker-peptide is used to connect the A. niger Xyn with the T. maritima Glu to construct two chimeras, Xyn-Glu and Glu-Xyn ( Figure 1). After expression in E. coli BL21(DE3), the two chimeras exhibit bi-functional activities of xylanase and glucanase. Especially, fusing the Glu domain at Cterminus increases the mesophilic xylanase T opt value and thermostability. The domain-swapping provides more insights into enzyme fusion investigation.

Construction of chimeras
After step-extension PCR, the xyn-glu and glu-xyn construct genes appeared on the 1.4 % electrophoresis gel as discrete DNA bands at~1.4 kb, approximately the combined molecular masses of contributing gene fragments ( Figure 1). After the sequence accuracies had been confirmed for the recombinant plasmids (GenBank: JQ793634, JQ793635), transformed cells containing pET20b-xyn-glu and pET20b-glu-xyn were induced to express the chimeric xylanases. The two proteins, Xyn-Glu and Glu-Xyn, appeared on the 10 % SDS-PAGE gel as discrete bands at~53 kDa. As expected, the molecular masses are approximately the sum of two contributing moieties ( Figure 1, Table 1), because the chimeras have residues equaling to the sum of Xyn, Glu, and linker peptide.

Properties of the chimeras
When properties were assayed, the two chimeras exhibited bi-functional activities of xylanase and glucanase. When xylanase activity was determined, both chimera Xyn moieties had optimal reaction pH (pH opt ) at 3.8, equivalent to the Xyn. Thus, fusing the Glu domain at either terminus did not alter the xylanase pH property. Both chimera Xyn moiety activities decrease at pH higher than 3.8 ( Figure 2). As to thermal property, the Xyn-Glu Xyn moiety had optimal reaction temperature, T opt , at 50°C, compared to 47°C for the Xyn ( Figure 3A). The Xyn moiety had t 1/2 at 50°C for 47.6 min,~2.7-times longer than the Xyn ( Figure 3B, Table 1). Fusing the Glu domain at C-terminus increased Figure 1 Construction of the chimeras. The domains Xyn and Glu were connected by overlapping-extension PCR. The genes xyn and glu were amplified using p1/p2 and p3/p4, respectively, and the xyn-glu was created using p1/p4 (Upper). Thereafter, the genes glu and xyn were amplified using p5/p6 and p7/p8, respectively. The glu-xyn was created using p5/p8. The wild Xyn and Glu were amplified using p1/p8 and p5/ p4, respectively. The fusion genes appear on the gel as discrete bands at~1.4 kb, with molecular masses of approximately the sum of the contributing gene fragments (Middle). The two chimeras appear on the SDS-PAGE gel as discrete bands at~53 kDa, with molecular masses of approximately the sum of the Xyn, linker peptide, and Glu (Lower). Linker: linker-peptide, M: marker of DNA and protein.
the xylanase thermo-activity and thermostability. In contrast, the Glu-Xyn Xyn moiety had T opt at 47°C, equivalent to the Xyn ( Figure 3A). However, the Xyn moiety had t 1/2 value at 50°C for 5.2 min, shorter than the wild Xyn. Fusing the Glu domain at N-terminus did not alter the xylanase thermal activity, but decreased its thermostability ( Figure 3B, Table 1).
When glucanase activity was determined, both chimera Glu moiety T opt were higher than 96°C, similar to the Glu ( Figure 4A). The Glu-Xyn Glu moiety had t 1/2 at 95°C for 99.8 min, and the Xyn-Glu Glu moiety had t 1/2 clearly longer than 99.8 min ( Figure 4B, Table 1). The Glu activity increased after incubation at 95°C, because the chimera T opt was higher than 96°C and its thermostability was significantly high. Fusing the Xyn domain at either terminus, especially, at N-terminus, increased the glucanase thermostability. The Glu-Xyn Glu moiety pH opt was 5.8, 2 units higher than the Xyn-Glu Glu moiety and the wild Glu ( Figure 2). Fusing the Xyn domain at C-terminus significantly shifted the glucanase pH property to higher pH environments.
When the kinetics were determined, the xylanase catalytic efficiencies (K cat ) toward birch-wood xylan increased in the order Xyn < Glu-Xyn < Xyn-Glu. The substratebinding affinities (K m ) for birch-wood xylan increased in the order Glu-Xyn < Xyn < Xyn-Glu (Table 1). Fusing the Glu domain at either terminus cooperated with the xylanase to degrade substrate; however, fusing the Glu domain at N-terminus interfered with the xylanase substratebinding affinity. In contrast, the glucanase catalytic efficiencies (K cat ) toward carboxymethyl cellulose sodium (CMC) increased in the order Glu < Xyn-Glu < Glu-Xyn. The substrate-binding affinities (K m ) for CMC increased in the order Xyn-Glu < Glu = Glu-Xyn (Table 1). Fusing the Xyn domain at either terminus cooperated with the glucanase to degrade substrate; however, fusing the Xyn domain at N-terminus interfered with the glucanase substrate-binding affinity. Thus, a domain, especially, a bigger one, is better fused at C-terminus for enzyme getting higher activity.
The Xyn-Glu Xyn moiety T opt was 3°C higher than the Xyn. This is consistent with the Xyn T opt 3°C increase pH opt : optimal reaction pH, T opt : optimal reaction temperature, t 1/2 (min): thermal inactivation half-life, the residual xylanase activity was assayed after incubation at 50°C at 10-min intervals from 0 -30 min, and the residual glucanase activity was assayed after incubation at 95°C at 20-min intervals from 0 -60 min. K m and K cat were determined for birch-wood xylan and carboxymethyl cellulose sodium (CMC). after fusing the hyper-thermophilic carbohydrate-binding module [4]. Previous domains were mainly fused at enzyme one terminus. For example, the Bacillus subtilis xylanase was fused at the Bacillus amyloliquefaciens glucanase C-terminus [6]. The T. maritima xylanase was fused at the T. maritima glucanase C-terminus [8]. The T. maritima β-glucosidase was fused at the T. maritima cellulase C-terminus [9]. When the domains were fused at the other terminus, the chimeras lost both moiety activities [6,8]. There are so many xylanase domains fused at glucanase C-terminus. The situation probably indicates that xylanase locates naturally at downstream of glucanase. It is probably a trace of natural evolution, because the site location order is shown in the glucanase-xylanase bi-functional protein selected from meta-genome library [13]. The domain-swapping created different effects on each moiety properties. The N-terminal Glu domain decreased the xylanase thermostability; however, the Cterminal Glu domain increased the xylanase thermostability. The Glu has 258 residues,~1.4-times bigger than the 185 residue Xyn. Probably, the C-terminal Glu protects the Xyn from thermal inactivation; whereas, the N-  terminal Glu interferes with the Xyn folding and confirmation-forming, therefore, interferes with its thermostability. In contrast, fusing the Xyn at either terminus, especially, at N-terminus, increased the glucanase thermostability. In addition, fusing the Xyn at C-terminus shifted the glucanase pH opt 2 units higher to alkaline environment. The result probably reflects that glucanase cooperates with other catalytic domains on both sides, because two glucanases, CelA and CelB, are shown connecting together in the T. maritima genome (EMBL-Bank: Z69341.1) [14]. Each domain has a natural conformation to perform catalytic activity. Therefore, each domain needs a necessary space to form to an active conformation [4,[17][18][19], that is the reason why a suitable linker-peptide is so important. The decreased chimera T opt values in previous investigations are attributed to linker-peptide interference. For example, fusing the B. amyloliquefaciens βglucanase decreased the B. subtilis xylanase T opt value and thermostability [6]. Fusing the T. maritima glucanase also decreased the xylanase T opt value. When the domains were fused at other terminus, many enzyme activities were damaged [6,8]. The phenomenon is attributed to lack of linker-peptide. For example, the glucanases lost activities when fused at the other terminus [8,9]. To exclude the linker-peptide disturbance, we used the natural 22 residue linker-peptide to connect the Glu and Xyn. In addition, boundary determination is important for each domain to have complete residues. Only a domain has complete residues, can it form to an active confirmation. In the meantime, we have to eliminate the extra residues encoded by unnecessary endo-nuclease sites in expression vectors. Otherwise, the extra residues would disturb enzyme properties [20]. Thereby, NdeI and XhoI were used to delete the unnecessary endo-nuclease sites within the pET20b(+).

Conclusions
Through the natural 22 residue linker-peptide, the mesophilic A. niger Xyn was fused and swapped with the hyper-thermophilic T. maritima Glu. After expression in E. coli, both chimeras exhibited bi-functional activities of xylanase and glucanase. Especially, the Xyn-Glu Xyn moiety had 3°C higher T opt and 2.7-times higher thermostability than the Xyn. In addition to the synergistic glucanase activity, the fungal xylanase was enhanced thermostability. The domain-swapping created different effects on each moiety. Fusing the Glu domain at C-terminus increased the xylanase thermostability, but fusing the Glu domain at N-terminus decreased its thermostability. Fusing the Xyn domain at either terminus increased the glucanase thermostabilities, and fusing the Xyn domain at C-terminus shifted the glucanase pH property 2 units higher towards alkaline environments. From the domain-swapping investigation, we can infer that a domain, especially a larger one, can only be fused at C-terminus to increase enzyme catalytic efficiency.

Construction of chimeras
Through the linker-peptide, the two domains were fused together by overlapping-extension PCR. The standard PCR was carried out using 16.5 μg of pET20b-xyn or pET20b-glu, 1.0 μl of each of two related primers, 5 U of Pfu polymerase, 4.0 μmol of dNTPs, and 1 × polymerase buffer with the following thermal cycling: 4 min denaturation at 95°C, followed by 30 cycles of 1 min denaturation at 94°C, 1 min annealing at 33°C, and 1 min extension at 72°C. The reaction was completed with a 10 min extension at 72°C, unless described otherwise.
The genes xyn and glu were amplified using p1/p2 and p3/p4 and annealing at 24°C (Figure 1). The genes were recovered and served as templates (each 11 μg) to amplify the xyn-glu using p1/p4, annealing at 62.4°C, and extending for 1.5 min. The genes glu and xyn were amplified using p5/p6 and p7/p8, and annealing at 36°C. The amplified genes were recovered and served as templates (each 11 μg) to amplify the glu-xyn using p5/p8, annealing at 65°C, and extending for 1.5 min. The wild Xyn and Glu were amplified using p1/p8 and p5/p4,  Table 2, with italic shown for NdeI/XhoI restriction sites and bold for homologous region. Following PCR amplification, the genes were cloned into pET20b(+) plasmids that had been digested with NdeI/XhoI to delete the redundant endo-nuclease sites. The recombinant plasmids were transformed E. coli BL21(DE3) competent cells, then extracted and sequenced with an ABI 3730 automated DNA sequencer to confirm gene accuracy (Invitrogen Biotechnology, Shanghai, China). Accurately transformed plasmids were grown and induced to produce enzymes according to standard protocols [4]. A C-terminal His 6 tag was included in the chimera sequences to allow the proteins to be purified with Co 2+ -binding resin (Amersham Bioscience). Active fractions were pooled and further purified using sephadex G-25. The enzymes were detected using 12 % polyacrylamide SDS-PAGE, stained with Coomassie brilliant G-250. Protein concentrations were measured according to the Lowry method.

Enzyme properties of the chimeras
Each moiety of both chimeras was assayed in parallel with the Xyn or Glu, respectively. Every data point was determined for three independent reactions, including protein concentration, substrate concentration, optimal reaction temperature (T opt ), optimal reaction pH (pH opt ), residual activity, etc. The pH opt value was determined from pH 2.6 -5.0 in imidazole-biphthalate buffers. The Xyn moiety T opt was determined from 44 -53°C, and residual activity was assayed after incubation at 50°C at 10-min intervals from 0 -40 min. The Glu moiety T opt was determined from 78 to 96°C, and residual activity was assayed after incubation at 95°C at 20-min intervals from 0 -60 min. To indicate thermostability, residual activity was expressed as a ratio relative to the un-incubated enzyme activity, and the thermal inactivation half-life (t 1/2 ) was calculated by fitting the data with the equation y = A * e -kt (Origin, version 8.0).
The kinetics were assayed at T opt and pH opt conditions using birch wood xylans at concentrations from 10 -40 mg/ml for 5 min (Sigma-Aldrich, Shanghai, China). The kinetics for glucanase were assayed at 85°C and pH opt conditions using CMC at concentrations from 2.5 -15 mg/ml for 5 min. The data were fitted with the Hill function to calculate maximal activity (V max ) and K m (Origin, version 8.0). Xylanase activity was determined toward birch-wood xylan and glucanase activity was determined toward CMC using the dinitrosalicylic acid (DNS) method described previously [4]. One unit of activity (U) was defined as the amount of enzyme that produced 1 μmol xylose or glucose per minute.