For the fusion of EcGAD with CBD, the pEKPM-CBD-Lk-EcGAD-H6 expression vector was constructed (Figure 1). In this vector, EcGAD was fused to CBD and native linker of T. harziaum endoglucanase II [19] at its N-terminus and six histidine at its C-terminus, and the fused gene was placed under the control of a T7 promoter. GAD was expressed in E. coli BL21(DE3) harboring pEKPM-CBD-Lk-EcGAD-H6 through IPTG induction. As shown in Figure 2, two distinct bands were observed by SDS-PAGE, indicating that the fusion protein was partially degraded. N-terminal amino acid sequence analysis confirmed that proteolytic cleavage occurred in linker region (-ATTM↓STTT-, ↓ indicates cleavage site). This phenomenon might be related to the fact that inter-domain linkers are particularly susceptible to proteolysis [12]. The reason is that flexible regions can more easily adopt conformations compatible with the active site of endopeptidase [12].

In order to prevent proteolytic cleavage of the linker peptide, a native linker peptide was replaced with a S₃N₁₀ peptide known to completely resist E. coli endopeptidase [12]. As shown in Figure 3A, the fusion protein with a S₃N₁₀ linker was successfully expressed without being degraded by proteolysis, which was confirmed by N-terminal amino acid sequencing. However, analysis of the cellular fraction from induced cells showed that most of the fusion proteins were expressed as an insoluble form. To increase the solubility, the transformants were cultured at a low temperature (at 18 °C) and gene expression was induced by lactose (Figure 3B).

The soluble CBD-GAD fusion protein was purified using a Ni-NTA column and eluted by a stepwise gradient of imidazole. The imidazole fraction (250 mM) contained a 58 kDa protein corresponding to CBD-GAD with a purity of 84% (Figure 4). The specific activity of the purified CBD-GAD was 24.44 ± 0.77 U/mg, which was lower than the GAD fused to six histidine (35.01 ± 1.03 U/mg) expressed in E. coli in this study. The purified enzyme was used to evaluate the adsorption isotherm of CBD-GAD on Avicel (Figure 5). The cellulose-binding capacity was calculated to be 33 ± 2 nmol_CBD-GAD/g_Avicel; this was relatively low compared to other CBD-fusion proteins (0.082–1.10 μmol/g) reported in previous studies [20–22]. This low binding capacity might have been attributed to be due to steric hindrance by the use of modified linker or the complex nature of CBD-GAD. The reusability of the GAD immobilized on Avicel was determined by comparing its relative activity after several uses. As shown in Figure 6, the immobilized GAD retained about 60% of its initial activity after ten consecutive uses.

E. coli DH5α [F⁻, φ80dlacZΔM15, endA1 hsdR17(rK⁻, mK⁺), supE44, thi-1, λ⁻, recA1, gyrA96] was used as the host strain for cloning and plasmid maintenance. E. coli BL21 [F⁻, hsdS, gal, ompT, rB⁻, mB⁻] (Novagen, USA) harboring a lambda derivative, DE3, was used for the expression of GAD.

For the efficient cloning work and use of kanamycin as a selective marker, DNA region between BglII and NdeI in pET-28b(+) was replaced by the region between same enzyme sites in pET-22b(+), which created a modified pET-28b(+). For the CBD-GAD fusion, the gene encoding CBD and linker of THEG were amplified from pGAL-CBD-Lipase [19] using a forward primer (5′-GAAGGAGATATACATATGCAGCAAACTGTTTGGGGGCAA-3′) and a reverse primer (5′-ACGGAGCTCGAATTCGGATCCAGAGCTAGTTGGCGGAGTAGC-3′) in which the underlined characters were introduced for in-fusion ligation. The polymerase chain reaction (PCR) product was ligated into the modified pET-28b(+) plasmid (Novagen, USA) digested with NdeI and BamHI using the In-fusion™ Advantage PCR cloning kit (Clontech, USA). The resulting plasmid was named pEKPM-CBD. The GAD gene was amplified from E. coli genomic DNA by PCR using a forward primer (5′-ACTCCGCCAACTAGCTCTGGATCCGATAAGAAGCAAGTAACG-3′) and a reverse primer (5′-GTGGTGGTGGTGGTGGTGCTCGAGGGTATGTTTAAAGCT-3′) in which the underlined base pairs were introduced for in-fusion ligation. The PCR product was ligated into the pEKPM-CBD plasmid previously digested with BamHI and XhoI using the In-fusion™ Advantage PCR cloning kit, which created pEKPM-CBD-Lk-EcGAD-H6. To replace the native linker peptide with S₃N₁₀ linker sequence as suggested by Kavoosi et al. [12], CBD with the S₃N₁₀ linker are amplified from pEKPM-CBD-Lk-EcGAD-H6 using the specific primers (forward: 5′-GAAGGAGATATACATATGCAGCAAACTGTTTGGGGG-3′, reverse: 5′-CTTCTTATCGGATCC GTTGTTGTTGTTGTTGTTGT TGTTGTTGTTGCTGCTGCTAATGCATTGAGCGTAGTA-3′ in which the underlined characters were introduced for in-fusion ligation). The PCR products containing CBD and the S₃N₁₀ linker were ligated into pEKPM-CBD-Lk-EcGAD-H6 previously digested with NdeI and BamHI, which created pEKPM-CBD-mLk-EcGAD-H6.

For GAD protein without CBD, the GAD gene was amplified from E. coli genomic DNA by PCR using a forward primer (5′-GAAGGAGATATACATATGGATAAGAAGCAAGTAACG-3′) and a reverse primer (5′-GTGGTGGTGGTGGTGGTGCTCGAGGGTATGTTTAAAGCT-3′) in which the underlined base pairs were introduced for in-fusion ligation. The PCR product was ligated into the modified pET-28b(+) plasmid previously digested with BamHI and XhoI using the In-fusion™ Advantage PCR cloning kit, which created pEKPM-GAD-H6.

Transformants were cultured overnight in a test tube containing 2 mL of Luria-Bertani (LB) media with kanamycin (50 μg/mL) at 37 °C. The culture was then transferred to 1 L baffled flasks containing 200 mL of LB media supplemented with kanamycin (50 μg/mL). When cell density reached an optical density (OD) of 0.5 at 600 nm as measured by a spectrophotometer (UVICON, Switzerland), induction was initiated with 0.4 mM IPTG for 3 h. After IPTG induction, the cells were harvested and lysed by sonication for further analysis. Proteins in the lysates were separated by SDS-PAGE according to the method by Laemmli [23] on 10% polyacrylamide gels using a mini-gel apparatus (Hoefer, USA). The separated proteins were stained with Coomassie blue. For culturing at low temperature, the transformants were cultured using the protocol described above at 37 °C to an OD of 0.5 and induced overnight with 0.1% (w/v) lactose.

CBD-GAD-H6 was isolated using SDS-PAGE, and the corresponding bands were transferred to a PVDF (polyvinylidene fluoride) membrane using a semi-dry electroblotter (TE 77 PWR, Amersham Bioscience, UK). The membrane was stained with Ponceau S (AMRESCO, USA) to visualize the GAD bands and N-terminal amino acid sequencing was performed by the Korea Basic Science Institute (Republic of Korea).

The transformants were harvested by centrifugation. The cells were then resuspended in B-PER^®II reagent (bacterial protein extraction reagent, Pierce, USA) to extract soluble proteins. The supernatant was collected for purification; nickel affinity chromatography using Ni-NTA resin (HisTrap, GE Healthcare, UK) was performed to purify CBD-GAD and GAD proteins. The bound proteins were eluted with a three-step gradient of imidazole (12.5, 25, and 250 mM) in 20 mM sodium phosphate buffer with 0.5 M NaCl, pH 7.4. Fractions collected at each step were analyzed by 10% SDS-PAGE. Purified proteins were dialyzed overnight against 20 mM sodium phosphate buffer (pH 7.4) at 4 °C.

Enzymatic activities were measured by a 718 Stat Titrino (Metrohm, Switzerland) with an aqueous solution of 0.1 M HCl. The solution containing MSG of 0.1M and PLP of 0.02 mM was adjusted to final pH of 4.0 and an enzyme was added. The enzyme reaction was performed at 37 °C while recording the titration. Activities were calculated by the slope of the titration curve over 5 min. One unit was defined as the amount of enzyme required for being added 1 μmol H⁺ per minute under assay conditions.

To estimate the binding capacity of CBD-GAD attached to the Avicel, adsorption isotherm measurements were taken at room temperature in 1.5 mL Eppendorf tubes. Samples containing 0–60 nM of CBD-GAD and 0.2 mg of Avicel were mixed with 20 mM sodium acetate buffer (pH 6.0) in a final aqueous volume of 1.2 mL. Each solution was incubated for 1 h to allow the adsorption to saturate. The supernatants were collected by centrifugation and the protein concentration was measured using a BCA protein assay kit (Pierce, USA). This value was then subtracted from the initial amount of protein added to the reaction tube and the difference was considered to be amount of proteins adsorbed onto Avicel. All values were measured in triplicate.

Relative quantification of retained activities after use of immobilized GAD was performed by TLC combined with image analyzer. Samples were prepared as follows; Enzyme was added into the 100 mM MSG and 0.02 mM PLP in 20 mM sodium acetate buffer (pH 4.0). After enzyme reaction at 37 °C for 30 min, immobilized enzyme was recovered by centrifugation and re-used after washing with sodium acetate buffer (pH 4.0). The loss of enzyme molecules due to the recovery process was calculated as 0.3% at each step. The supernatants after enzyme reaction and GABA standard solution were spotted on 10 cm × 10 cm glass TLC plates precoated with 60F254 silica gel (Merck, Germany). The mobile phase was prepared by mixing n-butanol/acetic acid/water (5/3/2, v/v/v). The plates were developed by applying ninhydrin reagent (Fluka, Switzerland) then dried at 80 °C for 5 min. The image analysis of the spot on the plate was processed by TotalLap TL100 Image Analysis Software (Nonlinear Dynamics, UK).