Cyclodextrin enhanced the soluble expression of Bacillus clarkii γ-CGTase in Escherichia coli

Background Cyclodextrin glycosyltransferases (CGTases) catalyze the synthesis of cyclodextrins, which are circular α-(1,4)-linked glucans used in many applications in the industries related to food, pharmaceuticals, cosmetics, chemicals, and agriculture, among others. Economic use of these CGTases, particularly γ-CGTase, requires their efficient production. In this study, the effects of chemical chaperones, temperature and inducers on cell growth and the production of soluble γ-CGTase by Escherichia coli were investigated. Results The yield of soluble γ-CGTase in shake-flask culture approximately doubled when β-cyclodextrin was added to the culture medium as a chemical chaperone. When a modified two-stage feeding strategy incorporating 7.5 mM β-cyclodextrin was used in a 3-L fermenter, a dry cell weight of 70.3 g·L− 1 was achieved. Using this cultivation approach, the total yield of γ-CGTase activity (50.29 U·mL− 1) was 1.71-fold greater than that observed in the absence of β-cyclodextrin (29.33 U·mL− 1). Conclusions Since β-cyclodextrin is inexpensive and nontoxic to microbes, these results suggest its universal application during recombinant protein production. The higher expression of soluble γ-CGTase in a semi-synthetic medium showed the potential of the proposed process for the economical production of many enzymes on an industrial scale. Electronic supplementary material The online version of this article (10.1186/s12896-018-0480-8) contains supplementary material, which is available to authorized users.

Escherichia. coli (E. coli) has often been used for the expression of CGTases [8]. However, inclusion bodies generally accumulate in the crowded cytoplasm or periplasmic milieu of E. coli during the protein overexpression process [9].Several strategies have been exploited to reduce the formation of inclusion bodies in vivo, including co-expression of chaperone genes, addition of "chemical chaperones", and lowering the culture temperature and inducer concentration [9,10]. Chemical chaperones, usually osmolytes, help cells counter unfavorable physiological conditions by stabilizing the native folding state of proteins and destabilizing partially folded states and early aggregates [11]. Cyclodextrins have recently been shown to be efficient chemical chaperones, assisting protein refolding and suppressing the misfolding and aggregation of proteins both in vitro and in vivo [12,13].
In the present study, cyclodextrin was employed for the first time as a chemical chaperone in the expression of recombinant γ-CGTase in E. coli. In addition, an efficient process control strategy was developed to reduce inclusion body formation during γ-CGTase production.

Results and discussion
Effect of chemical chaperones on γ-CGTase production by E. coli In this set of experiments, we assessed the efficacy of Ca 2+ , proline, betaine, sorbitol, and cyclodextrins as chemical chaperones in shake-flask culture. Ca 2+ is a ligand of the CGTase molecule and is necessary for maintaining its structure. Proline, betaine and sorbitol have been reported to be beneficial for protein overexpression [11,14]. No inducer was added to the TB medium, but 0.5% (w/v) glycine was added to enhance the secretion of γ-CGTase. Thus, the control culture contained 0.5% (w/v) glycine in TB medium, while the test cultures contained, in addition to the glycine, 1 mM Ca 2+ , a 5-mM concentration of α-, βor γ-cyclodextrin, or a 20-mM concentration of proline, betaine, or sorbitol. Among these chemical chaperone candidates, the cyclodextrins were remarkably effective, compared with the control, in promoting the production of soluble γ-CGTase (Fig. 1a). The total soluble γ-CGTase activity, which is the sum of the extracellular γ-CGTase activity and the γ-CGTase activity extracted from the periplasmic space, obtained in the presence of cyclodextrins was twice that seen with the control (α-cyclodextrin, 2.06-fold; β-cyclodextrin, 2.03-fold; and γ-cyclodextrin, 1.95-fold). The other candidate chemical chaperones elicited increases of less than 20%, compared with the control.
SDS-PAGE analyses revealed that all of the chemical chaperone candidates were somewhat effective in inhibiting the formation of insoluble inclusion bodies ( The band corresponding to γ-CGTase is seen near the 66.2 kDa mass marker 1b), but the cyclodextrins were particularly effective. As seen in Fig. 1b, the densities of the CGTase bands (near the 66.2 kDa mass marker) in lanes 3-5 (α-, β-, and γ-cyclodextrin, respectively) are much lighter than that in lane 1 (control) of Fig. 1b. The cyclodextrins were also particularly effective at increasing the recovery of soluble (extracellular plus periplasmic) CGTase. As seen in Fig. 1c, the densities of the CGTase bands (near the 66.2 kDa mass marker) in lanes 3-5 (α-, β-, and γ-cyclodextrin, respectively) are much greater than that in lane 1 (control). The inhibition of inclusion body formation was probably related to the interaction between the cyclodextrins and hydrophobic amino acid residues in CGTase, especially aromatic amino acid residues [13]. This sequestration of hydrophobic side chains shifts the thermodynamic equilibrium toward the unfolded state and away from the formation of aggregates, which is initiated by the hydrophobic amino acids. This interaction is clearly reversible, as there was no notable difference in specific activity between the γ-CGTase expressed in E. coli with and without β-cyclodextrin (8.959 versus 8.687 U·mg − 1 , see Additional file 1: Figure S1 and Additional file 2: Table S1).
The effect of cyclodextrin concentration on soluble γ-CGTase production by E. coli In these experiments, we assessed the effect of increasing concentrations (0-10 mM) of β-cyclodextrin on the production of soluble γ-CGTase. β-Cyclodextrin was used because it is cheap and readily available, and it effectively promotes the production of soluble γ-CGTase. As shown in Fig. 2, the total soluble γ-CGTase activity increased as the β-cyclodextrin concentration increased until it reached a plateau at a β-cyclodextrin concentration of approximately 7.5 mM. At this point the total soluble γ-CGTase activity had reached 5.51 U·mL − 1 . In addition, the ratio of extracellular γ-CGTase activity to total γ-CGTase activity decreased from 71.7 to 55.0% when the concentration of β-cyclodextrin increased from 0 to 10 mM.
The effect of glycine concentration on γ-CGTase production by E. coli We have recently shown that the addition of glycine to the culture medium improves the secretion of recombinant proteins in E. coli because it increases the permeability of the E. coli cell membrane [15]. Thus, glycine should also influence the diffusion of cyclodextrins across the E. coli cell membrane, thereby augmenting their ability to influence protein production.
As shown in Fig. 3, both total soluble γ-CGTase activity and cell growth decreased when the concentration of glycine increased from 0 to 10 g·L − 1 . However, the ratio of extracellular γ-CGTase activity to the total soluble γ-CGTase activity increased, in both control cultures (from 71.7 to 97.0%) and cultures supplemented with β-cyclodextrin (from 56.3 to 91.8%). More importantly, the effect of β-cyclodextrin on total soluble γ-CGTase activity, as a multiple of the control value, increased from 1.98-fold (5.61 vs 2.83 U·mL − 1 ) to 2.97-fold (3.17 vs 1.17 U·mL − 1 ) as the glycine concentration increased. Thus, the increase in E. coli cell membrane permeability   by glycine did appear to enhance the diffusion of β-cyclodextrin into these cells, but this effect was second to the detrimental effect of glycine on cell growth and protein production.
The effect of the induction conditions on γ-CGTase production by E. coli Temperature Induction temperature is an important parameter for recombinant protein production in E. coli. In previous studies of the production of recombinant α-CGTase, pullulanase, arginine deiminase and glutamate decarboxylase, we have shown that temperature has important impacts on both cell growth and recombinant enzyme production. In general, induction at lower temperatures increases the solubility of the recombinant proteins by preventing the formation of inclusion bodies. To assess the effect of induction temperature on γ-CGTase production, E. coli cells producing γ-CGTase were cultured at 25, 30, or 37°C. As shown in Fig. 4, soluble γ-CGTase activity decreased as the culture temperature increased. The amount of soluble γ-CGTase produced at 37°C was less than 30% of that produced at 25°C for both the control and the culture supplemented with 7.5 mM β-cyclodextrin. However, it should be noted that no matter which temperature was used during the induction, a greater amount of soluble γ-CGTase was produced in the presence of β-cyclodextrin than in its absence (25°C, 1.98-fold; 30°C, 2.04-fold; and 37°C, 2.11-fold). These results demonstrate that adding β-cyclodextrin does not change the effect of induction temperature on the production of soluble γ-CGTase. Rather, β-cyclodextrin addition approximately doubled the production of total soluble γ-CGTase at each temperature. In addition, the ratios of extracellular γ-CGTase activity to total soluble γ-CGTase activity were (25°C, 71.7%; 30°C, 75.9%; and 37°C, 76.3%) in control cultures and (25°C, 56.3%; 30°C, 70.9%; and 37°C, 70.3%) in cultures supplemented with β-cyclodextrin.

IPTG and lactose concentration
Since pET-24a (+) is a T7 promoter-based expression system, protein production is usually induced by adding IPTG or lactose to the culture medium. The impacts of IPTG concentration on cell growth and soluble γ-CGTase production were investigated at an induction temperature of 25°C. As shown in Fig. 5a, the dry cell weight (DCW) decreased with increasing IPTG concentration. Total soluble γ-CGTase production also decreased. When 0.5 mM IPTG was added, total γ-CGTase production in the absence and presence of 7.5 mM β-cyclodextrin dropped to 10.6 and 13.7% of those seen in the absence of IPTG (2.83 and 5.61 U·mL − 1 ), respectively. In addition, the ratio of extracellular γ-CGTase activity to total soluble γ-CGTase activity increased from 71.7 to 100% in control cultures and from 56.3 to 87.0% in cultures supplemented with β-cyclodextrin. However, this trend could not overcome the decrease in total γ-CGTase production, so we conclude from these experiments that induction with IPTG is detrimental to E. coli cell growth and not suitable for γ-CGTase production.
Since IPTG, which is a strong, persistent inducer, is not suitable for γ-CGTase production in TB medium, the moderate inducer lactose was assessed as an alternative. As shown in Fig. 5b, increasing lactose concentrations increased the DCW, after a minor decrease at low lactose concentration. In contrast, total soluble γ-CGTase production gradually declined, after a minor increase at low lactose concentration. The greatest total soluble γ-CGTase production was obtained when 2 g·L − 1 lactose was added. These results demonstrate that the addition of lactose is not detrimental to E. coli cell growth and suggest that lactose may be suitable for the induction of γ-CGTase production during large-scale fermentation in a fermenter. In addition, the ratio of extracellular γ-CGTase activity to total soluble γ-CGTase activity changed from 71.7 to 62.4% in control cultures and from 56.3 to 42.5% in cultures supplemented with β-cyclodextrin.
γ-CGTase production in a 3-L fermenter using a twostage, fed-batch strategy with β-cyclodextrin addition Fed-batch culture processes are often used to achieve high cell concentration and improved productivity while minimizing the problems encountered in high-cell-density culture [16]. We have established a fed-batch fermentation strategy that allows high-cell-density E. coli cultivation with high-level recombinant α-CGTase production [17]. In this study, in the pre-induction phase, glycerol feeding was initiated at a flow rate of 6.48 mL·h − 1 (a specific feeding rate of 5.40 mL·L − 1 ·h − 1 ) and then the feeding rate (F t ) was increased exponentially [18] with a controlled specific growth rate (μ set ) of 0.20 h − 1 . Thus, the change in F t with time can be expressed as F t ¼ 6:48e 0:2ðt−t F Þ (mL·h − 1 ). When the DCW reached 30 g·L − 1 , the post-induction phase began and the feeding rate was changed using the gradient-decreasing method [17].
The lactose feeding rate were investigated while using a two-stage induction temperature strategy with and without 7.5 mM β-cyclodextrin addition. As shown in Fig. 6, when the induction was performed in the absence of β-cyclodextrin at lactose feeding rates of 0.15, 0.30 and 0.60 g·L − 1 ·h − 1, total soluble γ-CGTase activities of 22.93, 29.33 and 26.18 U·mL − 1 were achieved, respectively. However, when 7.5 mM β-cyclodextrin was added, total soluble γ-CGTase activities of 36.30, 50.29 and 45.45 U·mL − 1 were achieved. These activities represent 1.58-, 1.71-and 1.74-fold increases over the results obtained in the absence of β-cyclodextrin. The total soluble γ-CGTase activity of the early phase increased with increasing lactose feeding rate. Induction with a lactose feeding rate of 0.30 g·L − 1 ·h − 1 gave the highest total soluble γ-CGTase production with or without β-cyclodextrin addition. The greatest total soluble γ-CGTase activity seen in the presence of β-cyclodextrin (50.29 U·mL − 1 ) was the highest production level seen to date. This improvement in γ-CGTase yield has the potential to reduce the cose of γ-cyclodextrin production, which may accelerate the use of γ-cyclodextrin in novel applications. This enthusiasm, however, must be tempered. Although β-cyclodextrin addition increased total soluble γ-CGTase activity, extracellular γ-CGTase activity was not remarkably increased by the adition of β-cyclodextrin. In other words, secretion efficiency appears to be compromised by the addition of β-cyclodextrin, which may due to the protection of E. coli cells by β-cyclodextrin, as can be inferred from the increase in DCW.
Since specific productivity (units of enzyme activity per gram of cell weight per hour of culture) is an important parameter in the industrial production of enzymes, the parameters used for recombinant γ-CGTase production in shake flasks and in a bioreactor with fed-batch fermentation are compared in Additional file 3: Table S2. In shake-flask culture, the specific productivity in the presence of β-cyclodextrin increased as the culture temperature decreased from 37 to 25°C (from 6.6 to 24.0 U·g cell − 1 ·h − 1 ). These levels observed were 1.93-and 1.80-fold greater, respectively, than the levels observed in the absence of β-cyclodextrin. In the case of fed-batch fermentation at 25°C in a 3-L bioreactor, the specific productivity in the presence of β-cyclodextrin was 1.45-, 1.58-, and 1.55-fold greater than that observed in the absence of β-cyclodextrin at lactose feeding rates of 0.15, 0.30 and 0.60 g·L − 1 ·h − 1 , respectively.

Conclusions
In conclusion, we describe the first use of cyclodextrin as a chemical chaperone in the production of recombinant γ-CGTase in E. coli. The addition of β-cyclodextrin probably enhancing the soluble expression of γ-CGTase by inhibiting the formation of inclusion bodies (Fig. 1b) The results were obtained from three independent experiments a scheme depicting the relevant processes is presented in Fig. 7. The addition of β-cyclodextrin was combined with process optimization strategies to enhance the production of soluble γ-CGTase by E. coli. A modified two-stage glycerol feeding strategy in the presence of β-cyclodextrin enhanced the production of soluble B. clarkii γ-CGTase in a 3-L fermenter. Through this cultivation approach, the total soluble activity and specific productivity reached 50.29 U·mL − 1 (5.61 mg·mL − 1 ) and 22.4 U g − 1 cell ·h − 1 , respectively. Since β-cyclodextrin is inexpensive and not toxic to microbes, it has the potential to be universally applied to the economical production of soluble recombinant enymes on large scales.

Media and feeding solutions
LB medium was used to prepare seed cultures and the TB medium was used in shake-flask cultures, see reference [17]. The medium used for γ-CGTase production in the 3-L fermenter contained (g·L

Bioreactor
An overnight seed culture was diluted (8% v/v) into a semi-synthetic medium (described in "The effect of cyclodextrin concentration on soluble γ-CGTase production by E. coli" section) for fed-batch cultivation. Fed-batch cultivation, which was performed in a 3-L fermenter (BioFlo 115, New Brunswick Scientific Co., Ltd), consisted of three phases that were performed as described by reference [17] with some modifications. The first phase, characterized as a batch phase, had an initial glycerol concentration of 8 g·L − 1 and a culture temperature of 37°C. After inoculation, the dissolved oxygen (DO) value decreased immediately and there was a gradual decrease in pH and glycerol. The end of glycerol consumption was signaled by a sharp increase in both DO and pH value. As the DO and pH value kept growing, the second, exponential feeding (preinduction) phase of fed cultivation was started. When a DCW of 30 g·L − 1 was reached, the inducer was fed at a range from 0.15 to 0.6 g·L − 1 ·h − 1 and the culture temperature was decreased to 25°C (unless otherwise stated) for γ-CGTase production. This began the third, post-induction phase. During the whole process, the pH was kept at 7.0 by automatic addition of ammonia solution (25%, v/v). Antifoam was added manually when necessary. To maintain the DO level at around 30% of air Fig. 7 Scheme with simplified protein synthesis and folding pathways comparing CGTase expression by E. coli in the absence and presence of cyclodextrin. The unfolded CGTase protein is drawn as a green line, while the signal peptide is drawn as a bold brown line. The orange segments in the green lines represent amino acids with hydrophobic side chains, and the blue rings represent CD molecules. Unfolded CGTase is synthetized by the ribosome complex and transported through the Sec-dependent pathway with the help of cellular proteins SecB, DnaK, SRP and SecA, as well as the membrane protein Sec-complex. In the periplasmic space, unfolded CGTase is either folded into an active protein or misfolded and aggregated into inclusion bodies saturation, the agitation speed was varied from 200 to 900 RPM. The air flow rate was 1.8 L·min − 1 . Dissolved oxygen concentration, pH, temperature, and impeller speed were recorded using Advanced Fermentation Software (AFS) from New Brunswick Scientific Co. Inc.

Determination of biomass
Cell growth was monitored during cultivation as described by Duan [14].

Cell fractionation
Cell fractionation was performed as described by Duan [14] with some modifications. The extracellular fraction was obtained by centrifugation of the culture broth at 13,800 g for 10 min at 4°C, and the supernatant was used as the extracellular fraction. The cell pellets arising from the centrifugation were harvested and resuspended in 1 mL of 30 mM Tris-HCl solution (pH 8.0) containing 25% (w/v) sucrose and 1 mM EDTA. The cell suspension was incubated on ice for 2 h and pelleted by centrifugation at 13,800 g for 10 min at 4°C. The supernatant was collected as the periplasmic fraction. The total soluble γ-CGTase activity was the sum of periplasmic enzyme activity and the extracellular enzyme activity.