Enzymatic synthesis of 1-caffeoylglycerol with deep eutectic solvent under continuous microflow conditions

Biochemical Engineering Journal 142 (2019) 41–49

Contents lists available at ScienceDirect

Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej

Regular article

Enzymatic synthesis of 1-caffeoylglycerol with deep eutectic solvent under continuous microflow conditions

T

Xi Liua, Xiang-Yun Menga, Yan Xua,b, Tao Dongc, Dong-Yang Zhanga,b, Hui-Xiang Guana, ⁎ Yu Zhuangd, Jun Wanga,b, a

Jiangsu University of Science and Technology, Zhenjiang, 212018, PR China Sericultural Research Institute, Chinese Academy of Agricultural Sciences, Zhenjiang, 212018, PR China c National Bioenergy Center, National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, Colorado, USA d Nantong Yuanshengtang Biological Science and Technology Co., LTD, Haian, 226600, PR China b

H I GH L IG H T S

was first obtained from MC and glycerin with co-solvent ChCl-Urea in a microreactor. • 1-CG time with a yield of 96.49% reduced by 75% compared to that using a batch reactor. • Reaction Michaelis constant in a microreactor decreased 5/6 compared to that in a batch reactor. • The transfer coefficients in a microreactor were both 10 times more than in a batch reactor. • Mass • Lipase reuse times (20) in a microreactor increased 2.29-fold compared with a batch reactor.

A R T I C LE I N FO

A B S T R A C T

Keywords: Biocatalysis 1-Caffeoylglycerol Deep eutectic solvent Kinetics Microreactor Mass transfer coefficient

1-caffeoylglycerol (1-CG) is a hydrophilic ester with a high solubility, and has a strong activity to prevent skin cancer caused by ultraviolet. However, the current preparation process to synthesize 1-CG via transesterification from methyl caffeate (MC) and glycerin was not economic due to long reaction time or adding extra field. Therefore, an enzymatic synthesis with deep eutectic solvent chloride-urea (ChCl-urea) in a microreactor was firstly applied under continuous microflow condition to improve the productivity. A maximum 1-CG yield of 96.49% was obtained under the optimized conditions: temperature of 65 °C, flow rate of 2 μL/min, MC concentration of 50 g/L. Compared to the batch reactor, the reaction time reduced by 75%, the Km value decreased by 5/6, the reuse times of lipase increased by 2.29-fold, the external mass and transfer coefficients increased 10 times, which process was verified by numerical simulation. In addition, the produced 1-CG of 2 mg/mL has a high UV damage remediation ability of 75.16% using HaCaT cell model, which reveals 1-CG can repair the damage of cells radiated by the UV. Thus, this approach represents a convenient and cost-saving method to produce 1-CG using microfluidic biocatalysis.

1. Introduction Caffeic acid (3, 4-dihydroxycinnamic, CA) is naturally found in fruits, vegetables, olive oil, and plants. A number of effects of CA have been reported such as antioxidative, preventing cardiovascular disease and carcinogenic inhibition especially skin cancer caused by ultraviolet (UV) [1]. However, the very poor solubility in non-polar media limits its application in food, medicine and cosmetic industry. In view of this feruloylated monoacyl- can be participated in human lipid metabolism and synthesis, which may improve the absorption and maximize the

⁎

effect of ferulic acid [2]. Therefore, CA esterification and glycerin modification of alkyl caffeates were chosen to enhance lipophilicity and solubility of CA. 1-caffeoylglycerol (1-CG) is a hydrophilic ester with solubility 3 times higher than CA in water at 20 °C [3]. Some studies reported that 1-CG was prepared by the transesterification of ethyl caffeic (EC) with glycerol using Novozym 435 or ionic liquids as catalysts [4]. In our work, lipase-catalyzed transesterification of 1-CG was performed using methyl caffeate (MC) and glycerin in a batch reactor and the maximum yield reached 90.6% under the optimal condition [1]. The results revealed the feasibility of the transesterification process

Corresponding author at: Jiangsu University of Science and Technology, Zhenjiang, 212018, PR China. E-mail address: [email protected] (J. Wang).

https://doi.org/10.1016/j.bej.2018.11.007 Received 23 July 2018; Received in revised form 10 November 2018; Accepted 12 November 2018 Available online 15 November 2018 1369-703X/ © 2018 Elsevier B.V. All rights reserved.

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All other reagents were analytical grade.

of 1-CG and provided an efficient pathway for the enzymatic synthesis of 1-CG. However, batch reactor has some inherent drawbacks such as long reaction times which will limit efficient production of 1-CG. Hence, more efficient method to synthesize 1-CG from alkyl caffeate and glycerin need to be explored. Microreactors are usually defined as miniaturized reaction systems fabricated by using, at least partially, methods of microtechnology and precision engineering [5]. Nowadays, microreactor technology applications in the biocatalysis field are gaining popularity for these benefits: faster mass transfer, high surface area-to-volume ratio, exquisite control and easier scale-up of production capacity which make the microreactor more effective for enzyme-catalyzed reactions than batch reactor [5,6]. Cech J reported when 1-heptyl-methylimidazolium bis (trifluoromethylsulfonyl) imide ([C7mmim] [Tf2N]), Novozym 435 and microreactor was used as the solvent, catalyst and reactor respectively to synthesize isoamyl acetate, the yield of product could reach 92%, the lipase selectivity much higher than that in batch reactor. To further increase the reaction efficiency, co-solvent was applied to improve the solubility of MC due to its advantages, especially easy and cheap preparation, biodegradability, and the precursors used are renewable, nontoxic, and natural compounds [7]. Zhao investigated that, Novozym 435 obtained the highest enzyme activity (> 1 μmol·min−1/g) and selectivity (> 99%) in the choline chloride/urea system when used to catalyze the transesterification of ethyl sorbate and 1-propano [8]. In addition, the introduction of co-solvents into the reaction mixture decreases the severity of the reaction parameters and can make this process practical, and can also decrease the amount of glycerin and allow the reactions to be carried out at milder temperatures [9]. Therefore, the co-solvents can also improve the enzyme activity and selectivity. Hence, we screened a series of co-solvent with microfluidics to synthesize 1-CG. Enzyme kinetics is an important means to evaluate different reactor performance and estimate the extent of substrate transesterification under known condition [10,11]. The development of tailormade, low cost products is important for the innovation in reactor, therefore, kinetics modeling plays a role as an engineering practice in accelerating enzymatic reactions which indicates the behavior of substrate and enzyme [12]. In addition, enzyme kinetics modeling can use many parameters to explore pathways and reaction mechanisms of complex macromolecular substrates prior to developing innovative process control strategies to ensure stability and desired efficacy [13,14]. Hence, enzyme kinetics modeling was used to evaluate catalytic efficiency of enzyme and mass transfer of different reactors. The present work provided a simple and efficient approach to the enzymatic transesterification synthesis of 1-CG from MC and glycerin in co-solvent with independent design grooved type reactor. In addition, kinetic models were applied to optimize the reaction and study the reaction mechanism. For this purpose, a series of solvents were screened to increase the solubility of substrate. Besides, the reaction conditions were studied to determine the optimal conditions. Additionally, secondary structure and reusability of enzyme were detected to verify the performance of microreactor. Finally, 1-CG was used to cultivate HaCaT cell to explore the anti-UV effect of 1-CG.

2.2. Preparation and characterization of chloride-urea Choline chloride (14 g) and urea (12 g) were mixed in a threenecked flask and was stirred with a magnetic stirrer [15]. The reaction temperature was held at 80 °C in an oil bath heater. The reaction was completed when the solution altered to one phase with clear homogeneous appearance phase. The solution was evaporated with a rotary evaporator at 60 °C and dried for 24 h to obtain the eutectic solvents. The characterization of eutectic solvents was checked by 1H NMR analysis to verify the structure and purity of the product. 1H NMR (400 MHz) spectra were recorded on a Bruker AVANCE Spectrometer 400 (Bruker Biospin Co., Billerica, MA, USA). The structure of it was δH (D2O, 400 MHz): 4.01 (m, 2H, > N+-C-CH2-O-), 3.46 (t, 2H, > N+CH2-), 3.15 (s, 2H, > C-CH2-C <), 3.12 (s, 9H, -N+-(CH3)3). 2.3. Enzymatic synthesis of 1-CG with co-solvents The enzymatic synthesis of CA monolyceride was performed in a 5 mL bottle sealed with screw cap. Co-solvent acetone, dimethyl sulfoxide (DMSO), choline chloride-urea (ChCl-urea), choline chlorideglycerin (ChCl-glycerin) and ionic liquid [Bmim][Tf2N] were used as reaction medium, respectively. MC, glycerin and Novozym 435 were added in a bottle and placed in shaker with 180 rpm. After the reaction, the reaction solution was dissolved in ethanol, and the lipase was filtered off. The suspension liquid obtained by rotary evaporation of filtrate was extracted with ethyl acetate three times. The obtained solution was rotary evaporated to remove the ethyl acetate. The product was washed with distilled water three times, and diluted to one tenth and the purity was detected with HPLC. The white powders were collected after washing/centrifugation and dehydrated at −40 °C for 2 h. 2.4. HPLC analysis The HPLC quantitative analyses were carried out using HPLC (Waters 2695) with a PDA detector (Waters 2998). Reactants and products were analysed with a C18 reverse phase column (5 μm, 150 × 4.6 mm, i, d.; W. R. Grace & Co., Deerfield, IL, USA) using a binary gradient at a flow rate of 1 mL/min. Solvent A was methanol and solvent B was methanol/water (1:4 v/v, 1% ammonium acetate). The gradient profile commenced from 90% B to 10% B linearly in 30 min, returning to 90% B at 40 min followed by 5 min isocratic to equilibrate. The eluate was monitored at 325 nm using a photodiode array detector (PDA). The external standard method was adopted to calculate the total conversion of MC and the yield of 1-CG. The computational formulae are Eqs. (1) and (2) [16]:

MC conversion(%) =

Converted molar amount of MC (mol) Initial molar amount of MC (mol) × 100

1-CG yield(%) = 2. Materials and methods

(1)

Molar amount of 1-CG(mol) × 100 Initial molar amount of MC (mol) (2)

2.1. Lipase and chemicals 2.5. Continuous synthesis of 1-CG in an microreactor packed with immobilized enzyme

CA, ferulic acid and cinnamic acid were purchased from Nanjing Zelang Pharmaceutical Co., Ltd. (Nanjing, China). Novozym 435 was purchased from Novo Nordisk A/S (Bagsvaerd, Denmark). Choline chloride and urea were purchased from Sigma Chemical Co. (St. Louis, MO), PDMS microchannel reactor was purchased from Suzhou Wen Hao Chip Technology Co., Ltd (Suzhou, China). Methanol was chromatographically pure for HPLC (purchased from Tedia Co.; Fairfield, OH, USA). Other chemicals were all purchase from Shanghai

The packed-bed reactor was designed with polydimethylsiloxane which has a good chemical inertness with dimension 10 mm × 0.9 mm × 75 mm (W × H × L), respectively. One inlet and one outlet were set in both ends of the microreactor to ensure the substrate was fed from one inlet and flow out of the reactor from an outlet which was on the other side of the reactor. Novozym 435 of 100 mg was put in the 42

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

reaction channel as a biocatalyst. The constant current pump (LSP021B, Longer Pump Co. Ltd., Beijing, China) was used as a powerplant which feed the substrate (the homogeneous solution of MC, glycerin and co-solvent) into the microreactor with flow rates ranging from 2 μL/min to 8 μL/min. Considering the viscosity, the reaction temperatures were set at 50 °C, 60 °C, 65 °C, 70 °C and 75 °C respectively to reduce the viscosity of glycerin, the substrate concentration were 30 g/ L, 40 g/L, 50 g/L, 60 g/L and 70 g/L. Effluent of 50 μL was collected and diluted for HPLC detection.

UV damage repair degree of sample(100%) ODSample group -ODIrradiation group A − A0 = *100% = i *100% ODSample group Ai

(5)

2.10. Statistical analysis All detections of each parameter and antioxidant capacity were performed in triplicate. The analysis of variance (ANOVA) method was used, and the standard deviation was calculated to ensure the reliability of the results. Significance was determined at a 95% level of probability. For antioxidant capacity data, p values < 0.01 were considered significant.

2.6. Reusability and stability of enzyme in the microreactor The stability test of enzyme was performed first under the optimal substrate concentration and temperature. The microreactor was thoroughly washed with fresh solvent after each reaction. All products were collected and detected with HPLC to compare the change of yield of product in order to identify catalytic activity of lipase.

3. Results and discussion 3.1. Enzymatic synthesis of 1-CG via transesterification with different solvents

2.7. Kinetics analysis of the reaction in the microreactor It is the prerequisite that MC can be dissolved in the glycerin which ensured the progress of the reaction. Therefore, solvents were added into the reaction system in proportion. As shown in Table 1, all the selected solvents can activate the enzymatic synthesis of 1-CG. DMSO, [Bmim][Tf2N] and ChCl-urea used as reaction medium attained higher conversion of MC and higher yield of CA monoglyceride which were both more than 50%. Among these co-solvents, the highest conversion (91.80%) and highest yield (85.22%) were observed with ChCl-urea. Though organic solvents can increase the yield of 1-CG, it has higher toxicity and was difficult to be removed [19]; For ionic liquid [Bmim] [Tf2N], it has garnered interests in green engineering chemistry recently [19]. However, the preparation progress of ionic liquid is complex and difficult for separation and purification leading to high price. The deficiency may influence large-scale production. ChCl-urea is a green, biodegradable, and cost-effective solvent which can be produced via a simple approach. It has attracted increasing attention in the biocatalysis industry. Therefore, the ChCl-urea was chosen for synthesis of 1-CG from MC alkyl and glycerin.

The synthesis kinetics of CA monoglyceride was studied to elucidate enzyme-catalyzed mechanism in this reaction. The concentration of MC ranged from 0.02-0.6 mol/L, the flow rate ranged from 2 μL/min to 8 μL/min and other reaction conditions were set as the optimal factors. The initial reaction rates were expressed as the yield of product per minute. The relationship of initial reaction rates and the initial substrate concentration were expressed with function and the kinetics constants were determined by the coefficients of the function [17]. 2.8. Characteristics and secondary structure change of lipase measured by FTIR spectrometry Dimethyl sulfoxide (DMSO) and Novozym 435 were mixed with agitation at 37 °C for 30 min. The mixture was separated with suction filter, and the solid matter washed by DMSO. The measurement of the Novozym 435 secondary structure was achieved using FTIR spectrometry measured by a Varian 670 IR spectrometer. The secondary structure elements based on the information of the amide I region and the band assignment were manipulated using Omnic and Peakfit software [18].

3.2. Continuous flow synthesis of 1-CG in the microreactor Further experiments using ChCl-urea as the media and Novozym 435 as the catalyst to synthesize 1-CG was performed in the microreactor. As shown in Fig. 1A and B, when the temperature was 50 °C, the conversion of MC was only 20.13% at the flow rate of 8 μL/min. With low temperature, the glycerin with high viscosity inhibited the substrate contracting with enzyme, which make much active site of enzyme in an idle state. Besides, the low temperature resulted in low

2.9. The model construction of UV damage to HaCaT cell The UV light source of 313 nm was chosen to simulate UVB to irradiate HaCaT cell with distance of 10 cm at 200 μW/cm2. The ultraviolet radiation dose can be expressed by the formula Eq (3).

Ultraviolet radiation dose(mJ/cm2) = radiation time × UVB intensity Table 1 Comparison of lipase-catalyzed synthesis of 1-CG in different media.

(3) The ultraviolet radiation dose of specific wavelength determined by the model construction of UV damage to HaCaT cell. The cells were divided into three groups (blank control group, negative control group and sample group). Blank control group contained CK1 (A1) which was cell without UV radiation added with 30 μL DMEM medium, negative control group contained CK2 (A0) with 30 μL DMSO solution (10% mass fraction) without UV radiation, sample group (Ai) contained two groups: (1) cells with 30 μL product solution of different concentration with UV radiation, (2) cells with 30 μL Vc or BHA solution with UV radiation. The cell viability and degree of sample UV damage can be calculated with the following Eqs (4) and (5).

Irradiation group absorbance values *100% Blank way absorbance values A = 0 *100% A1

Solvents

MC conversion (%)

1-CG yield (%)

Solvent-free system [1] 10 % (v/v) acetonea 10 % (v/v) DMSOa 5 % (v/v) ChCl-ureab 10 % (v/v) ChCl-ureaa 20 % (v/v) ChCl-ureac 5 % (v/v) ChCl-glycerolb 10 % (v/v) ChCl-glycerola 20 % (v/v) ChCl-glycerolc

92.09 65.21 60.59 43.73 91.82 70.03 19.58 32.55 27.68

90.62 53.06 43.21 30.38 85.22 60.03 10.35 20.46 16.23

± ± ± ± ± ± ± ± ±

0.63 0.45 1.16 0.88 0.95 1.11 1.23 0.69 0.75

± ± ± ± ± ± ± ± ±

0.48 1.07 0.98 0.47 0.91 1.46 0.75 0.94 0.65

a MC concentration 50 g/L, MC/lipase = 1:5 (mass ratio), solvent/glycerol = 1:10 (v/v), 75 °C, 180 rpm. b MC concentration 50 g/L, MC/lipase = 1:5 (mass ratio), solvent/glycerol = 1:20 (v/v), 75 °C, 180 rpm. c MC concentration 50 g/L, MC/lipase = 1:5 (mass ratio), solvent/glycerol = 1:5 (v/v), 75 °C, 180 rpm.

Cell activity(%) =

(4) 43

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

Fig. 1. The effect of temperature (A and B), MC concentration (C and D) and residence time (E and F) on the yield of 1-CG and conversion of MC in a microreactor. Reaction conditions: (A) and (B) Novozym 435 90 mg, MC concentration 30 g/L, glycerol 1.8 mL, ChCl-urea 0.2 mL, temperature 50–75 °C; (B) and (C) Novozym 435 90 mg, MC concentration 20–70 g/L, glycerol 1.8 mL, ChCl-urea 0.2 mL, temperature 65 °C; (E) Novozym 435 90 mg, MC concentration 20–70 g/L, glycerol 1.8 mL, ChCl-urea 0.2 mL, temperature 65 °C; (F) Novozym 435 90 mg, MC concentration 50 g/L, glycerol 1.8 mL, ChCl-urea 0.2 mL, temperature.50–75 °C.

substrate concentration (30∼50 g/L) at 2 μL/min while increasing at a faster flow rate. Fig. 1C shows that with the flow rate of 2 μL/min, the rising trend of MC conversion and 1-CG yield was consistent. At a faster flow rate, 1-CG yield increased with the substrate concentration from 30∼50 g/L and decreased with higher concentration. Thus, the highest 1-CG yield (96.49%) was obtained with the substrate concentration of 50 g/L at a flow rate of 2 μL/min. When the substrate concentration and flow rate was within a range of 40∼50 g/L, a fast flow of many MC moved past the enzyme to perform the transesterification reaction while continuous flow contribute continuous generation of 1-CG

molecular motion rate which determined the mixing and collision. When the temperature increased to 65 °C, 1-CG yield maximized to 96.44%. At elevated temperature the viscosity of substrate dramatically dropped, making the reaction process smoother and more collision of substrate molecular [20,21]. Further increasing temperature will drop the yield slightly, indicating side reaction and recession of enzyme activity at high temperature [22]. This finding was consistent with our previous report [23]. The effect of substrate concentration on MC conversion and 1-CG yield is shown in Figs. 1C and D. MC conversion increased at a higher 44

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

combines with lipase and releases methanol and CA, and then CA will combine with lipase and glycerine to generate 1-CG. The reaction process belongs to Ping Pong reaction model, thus the Ping Pong Bi Bi model was used for kinetics analysis of transesterification reaction catalyzed by lipase [27]. In the present study, the kinetics model was based on the decrease of substrate concentration, and the reaction rate which is the following Eq. (6).

causing an increase in yield [24]. Suitable concentration can shift the equilibrium toward 1-CG production while too much MC and glycerin will affect the substrate combining with the enzyme which led to the decreasing of 1-CG yield. Figs. 1E and F show the effect of residence time on the enzymatic synthesis of 1-CG from MC and glycerin in the microreactor. The reactor employs a residence time at a range of 37.5–150 min with a flowing rate of 2–8 μL/min. Higher 1-CG yield was obtained with a longer interaction time between substrate and enzyme. Fig. 1E shows the effect of residence time on the incorporation under the various temperature conditions. With a short residence time ranging from 37.5 min to 75 min, the 1-CG yield steadily increased and obtained the highest yield (96.49%). However, when the residence time is over 75 min the incorporation did not increase markedly. Fig. 1F shows the effect of residence time on the 1-CG yield under different substrate concentration conditions. The maximum yield of 1-CG was obtained at a residence time of 150 min (flow rate of 2 μL/min) with the substrate concentration of 50 g/L. The two substrate ratios increased linearly at a low flow rate (residence time ranged from 37.5 min to 70 min). Thus, as the residence time decreases, the laminar flow may transform into turbulent flow. The substrate passes through the surrounding of the enzyme instead of entering the active site of the enzyme under the turbulent flow [25,26]. In summary, the 1-CG can reach 96.46% under the optimal conditions: Temperature of 65 °C, substrate concentration of 50 g/L and residence time of 150 min. However, for a batch reactor, the maximum yield of 1-CG 90.63% with temperature of 75 °C, substrate concentration of 70 g/L and reaction time of 10 h which resulted in a lot of economic and energy loss compared with microreactor. The reusability of enzymes is of great practical significance in industrial production. Therefore, the reusability of the Novozym 435 for the synthesis of 1-CG in the microreactor was evaluated at the optimal condition. Fig. 2 shows the activity of the enzyme over 20 runs in the microreactor. Under the optimal conditions, the yield of 1-CG still had a remainder of more than 50% after 20 runs. And then, a noticable decrease of yield occurred with further cycles. Hence, the results showed that the microreactor can ensure catalytic activity and site-specific selectivity of enzyme.

rA = kCAα CBβ

(6)

where: rA is the reaction rate of MC; k is reaction rate constant; CA is the concentration of MC (mol/L); α is the reaction order of MC; CB is the concentration of glycerin (mol/L); β is the reaction order of glycerin. In the kinetics model, the concentration of glycerin was much excessive in comparison with the MC. It was confirmed that the transesterification reaction rate was more sensitive to the changes in MC concentration [28]. Therefore, kCBβ was constant due to the excessive glycerine concentration. Then, Eq. (6) was simplified as Eq. (7):

rA = k1 CAα

(7)

x is assumed to be the conversion rate of MC and CA0 is the initial concentration of MC, the concentration MC at any time is shown in Eq. (8):

C A = CA0 (1 − x )

(8)

Therefore,

rA = −

dCA d (CA0 (1 − x )) dx = = CA0 = k1 CAα dt dt dt

(9)

Thus, Eq. (7) can be obtained:

k1 CA0

k2 =

(10)

Using integral method, Eq. (11) was derived from Eq. (7):

ln

dx = α ln[CA0 (1 − x )] + ln k2 dt

(11)

In Eq. (11) kinetic parameters, α can be calculated by the linear regression, which was shown in Fig. 3. As the flow rate ranged from 2 μL/min to 4 μL/min, the km values were 12.73 mM to 16.41 mM (Table 2). The results indicated that the km values increase with increasing flow rate. In addition, higher value of km may result in poor catalytic effect which can be confirmed from the performance of batch reactor and microreactor. This phenomenon indicated that mass transfer appeared to be playing an important role in the enzymatic transesterification reaction [29]. At a lower flow rate, substrate will be fully mixed and collided with the enzyme which make the substrate contact with the active site of the lipase to be sufficient. This action ensured the reaction efficiency and yield of 1-CG in the microreactor. Therefore, 2 μL/min was the suitable flow rate for a packed bed microreactor. As shown in Table 2, though the km values increased with increasing flow rate, all the three km values are still far less than the km value in the batch reactor. It is assumed that better mass transfer and heat transfer contribute to the improvement of kinetic parameter. Thus, we tried to explain this process by calculating the mass transfer in the microreactor. Internal mass transfer in the reaction are always calculated with the following Eqs (12) and (13) [29]:

3.3. Kinetic analysis of the microreactor For the enzymic reaction of two substrates, the dynamic mechanism involved divides into two kinds: sequential reaction model and Ping Pong reaction model. For the reaction of MC and glycerine, MC

η=

Fig. 2. Reusability of Novozym 435 in a continuous flow transesterification synthesis of 1-CG from MC and glycerol. Reaction conditions: flow rate 2 μL/ min, 90 mg enzyme, MC concentration 50 g/L, glycerol 1.8 mL, ChCl-urea 0.2 mL, temperature 65 °C and reaction time 2.5 h.

tD = tR

KL =

2De d

De / KL2 K eq × S0 / r0

(12) (13)

Where η is the internal mass transfer coefficient which is the ratio of tD and tR, De is the effective diffusivity. Keq is the equilibrium constant for 45

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

Fig. 3. Kinetic plots of the lipase-catalyzed synthesis of 1-CG from MC and glycerol using different reactors. (A) Lilly-Hornby plots of the continuous flow enzyme microreactor at different flow rates of 2 μL/min, 6 μL/min and 8 μL/min under different MC concentration (0.02-0.6 mol/L); (B) Lineweaver-Burk plot of reciprocal initial rates vs. reaction conditions: enzyme load-12%, MC concentration 0.02-0.6 mol/L, glycerol 60 mmol and 180 rpm in an incubator. 4

MC conversion, which is calculated to be 0.1 at 65 °C. S0 is the initial concentration of MC (0.277 mol/L), r0 is the initial rate (5.73 × 10−−5 mol/(m3∙s), KL is the molecular diffusion coefficient for liquid phase diffusion, and d is the diameter of Novozym 435 (0.3 mm), De can be defined as the following Eq. (14):

ε De = ⎛ ⎞ DAM ⎝τ ⎠

). Hence, compared to the external mass transfer, internal mass transfer coefficient can significantly affect the reaction. In addition, the mass transfer in microreactor is much more than that in batch reactor which verified that the microreactor had better mass transfer effect. The usual approach to the modelling of fluid flow in microchannels can be achieved by using a formulation in which a common flow field is shared by all of the phases. For an incompressible, Newtonian fluid, the hydrodynamics of such flows can be described by the Navier-Stokes (N–S) equation expressed as follows:

(14)

where DAM is the molecular diffusion coefficient [30]. Rausis, ε is the particle porosity of Novozym 435, τ is the tortuosity factor. The value of ε/τ is taken 0.5 for resin catalysis [31]. The DAM can be evaluated as the following eq:

DAM =

7.4 ×

10−8

φM × T

(VA )0.6μ

(15)

(robs ρp ) Rc n (16)

K c CA

(17)

ρ (u•∇) u = ∇•[−pI + η (∇u + (∇u)T )] + F

(18)

δts ρCp ∂T / ∂t + ∇•(−k∇T ) = Q − ρCp u•∇T

(19)

where ρ is water density, p is pressure, u is the velocity at the inlet of the microchannel, η is the dynamic viscosity, and Cp is the specific heat of the glycerin, F is the surface tension force approximated as a body force in the vicinity of the interface, and δts is the thermal boundary layer thickness. The bulk properties, such as density and viscosity, are determined from the volume-fraction weighted average of the properties of the fluids. Numerical simulation liquid flow was applied in the microreactor with the COMSOL Reaction Engineering Lab 3.5a. The density ρ (1.263 kg/m3), viscosity η (1.412 Pa·s), volume force F (0 N/m3), relative permittivity εr (80) and mean free path of gas molecules λmfp (10−6) were imported the software and it can solve the model with the built-in program. In addition, with the result of speed field, the temperature field can be solved by adding heat transfer coefficient k (0.268 W/m·K), density ρ (1.263 kg/m3), atmospheric pressure Cp (2.40 J/kg·K), specific heat rate γ (1). Fig. 4A shows the vorticity of the fluid in the groove. The vorticity (range from 0 to 0.0318 1/s) became great in the bottom of the groove where packed with the enzyme. This phenomenon indicated that the substrate can be fully mixed and

where φ and M are the association factor and the molar mass of the solvent. VA is the molar volume at boiling point and μ is the viscosity of the solution. VA can be calculated with the group contribution method. Through detection, the viscosity of glycerin was 81.3 mPa·s and VA was 72.68 cm3/mol. De (1.13 × 10−7 cm2/s) and η (2.055) > 1 which shows the internal diffusion in the microreactor play an important role in the transesterification reaction and can be calculated with the formula. External mass transfer can be calculated with the following eq [32]:

CM =

∇•u = 0

where the CM is the external mass transfer coefficient, γobsρp is the initial rate (5.73 × 10−−5 mol/(m3∙s), the n is reaction order (2), Rc was the particle size of the lipase, the Kc is the mass transfer coefficient which can be obtained from Dwidevi-Upadhyay mass transfer correlation as 2.2 × 10-5 m/s and the CA is bulk concentration [33]. The CM = 6.9 × 10-3 < 1 but still much more than batch reactor (7.3 × 10-

Table 2 Kinetic constants for the lipase-catalyzed synthesis of 1-CG using a continuous flow packed bed microreactor. Reactor a

Packed bed microreactor Packed bed microreactorb Packed bed microreactorc Batch reactord a b c d

Flow rate (μL/min)

Km (mM)

R2

η

CM

2 4 8 –

12.73 16.41 26.46 79.1

0.956 0.971 0.950 0.991

2.055 – – 0.115

6.9 × 10−3

MC concentration 50 g/L, lipase amount : 90 mg, flow rate : 2 μL/min at 65 °C. MC concentration 50 g/L, lipase amount : 90 mg, flow rate : 4 μL/min at 65 °C. MC concentration 50 g/L, lipase amount : 90 mg, flow rate : 8 μL/min at 65 °C. MC concentration 60 g/L, lipase amount : 60 g/L, at 75 °C. 46

7.3 × 10−4

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

Fig. 4. Analysis of mass transfer and heat transfer with numerical method: (A) Vorticity of the fluid in the microreactor (B) Speed field of the fluid in the microreactor (C) Temperature field of the fluid in the microreactor.

47

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

contact with the enzyme which contribute to the high yield of the 1-CG [33]. At the upper layer of the groove, the vorticity of the fluid was lower which can steadily pump substrate and output the product. Thus, microreactor can ensure the substrate fully mixed and reaction with a stable flow process. Speed field of the fluid in the microreactor is shown in Fig. 4B. When the fluid flow (ranging from 0 to 1.218 × 10-5) through the channel, the speed was faster than that of inflow or outflow. In addition, the speed changed a lot at the inlet and outlet of groove. It is reasonable to expect that the flow form of fluid has changed (from laminar flow to turbulence) when flow into the groove which packed with enzyme [34,35]. Fig. 4C shows the temperature field (range from 0 to 1.13 × 105 K/m) of the microreactor. The temperature decreased from bottom to upper layer of the microreactor which can ensure the contact section to be locally heated as well as prevent oxidation of the product [36]. Additionally, the local heat of effective area can maximize the use of energy which reduces the energy cost in order to achieve large-scale production. In summary, kinetics approach can analyze reaction parameters and explain the high efficiency of microreactor.

Table 3 Quantitative estimation (%) of the secondary structure elements of Novozym 435 in different reactor. Lipases

α-helix

β-sheet

Other coil

Original lipase Lipase in the microreactor Lipase in the batch reactor

19.6 14.8 0

48.7 50.1 67.2

31.7 35.1 32.8

3.4. Secondary structure change of Novozym 435 in microreactor and batch reactor The secondary structure of Novozym 435 can be detected with FTIR since the peptide bond vibrations of the proteins packaged with resin can absorb infrared wavelengths. The C]O stretching vibration can form an absorption peak at approximately 1600–1700 cm−1 which is called the amide I region. Its good sensitivity in conformational changes always enables it to be used in determination of secondary structure of protein [37]. The IR spectrum of lipase Novozym 435 in different reactors is shown in Fig. 5 and its secondary structure elements was calculated with the data shown in Table 3. The lipase was produced from modified Aspergillus oryzae, so the structure change of lipase before and after the reaction were detected. The α-helix and β-sheet contents of the original Novozym 435 were 19.6% and 48.7%, respectively. As Table 3 shows, the Novozym 435 in the batch reactor appeared to have a little change in the secondary structure change (α-helix decreased by 4.8% and β-sheet increased by 1.4%) while it is remarkably changed in the microreactor (α-helix decreased to 0 and βsheet increased by 18.5%). In total, the active site of lipase can be significantly affected with the decreasing of α-helix and increasing of βsheet contents [30,38]. Partially, the lower the α-helix, the easier the substrates access the active site [39]. Therefore, the change result shows that microreactor contributes to maintaining the enzymatic activity of lipase. The phenomenon may be related to that before full enzyme deactivation occurs, more than one type of interaction must be

Fig. 6. The cell activity treated by different samples after UVB irradiation (720 mJ/cm2).

broken followed by different changes of the native structure [18]. Hence, microreactor attributes to change of lipase Novozym 435 which results in a higher yield of 1-CG. 3.5. Repair ability of 1-CG on UV-damage of HaCaT cells We initially investigated the effect of 1-CG, MC, CA, BHA (0.1–2.5 mg/mL) on the UVB-mediated decrease in cell viability [40]. HaCaT cells were treated with the above four solutions for 24 h after being exposed to UVB. As Fig. 6 shows, UVB irradiation of HaCaT cells led to the decrease of cell viability. Compared to other three solutions, The UVB-induced cell death was significantly attenuated by treatment of cells with 1-CG in every concentration gradient. With the low concentration, MC showed the poorest effect on the rehabilitation of the cells while it gradually increased stronger than CA and BHA with the increasing concentration. When the concentration was 2 mg/mL, the cell activity reached the highest (75.16%, p < 0.05) much more than cell untreated which shows the strongest effect on the repairment of HaCaT cells. The cell activity maintains the degree with further increasing concentration. The result reveals 1-CG can repair the damage of cells radiated by the UV. 4. Conclusions The 1-CG yield of 96.46% was obtained from MC and glycerin with ChCl-urea in a microreactor under the optimal conditions: MC concentration of 50 g/L, temperature of 65 °C, residence time of 2.5 h and flow rate of 2 μL/min. Compared to the batch reactor (90.63%, 10 h), the reaction time shortened by a quarter, the yield increased 5.86%, the Km value decreased by five sixths, internal (2.055 > 1) and external (6.9 × 10−3) mass transfer coefficient both increased 10 times. The lipase can be reused 20 times. In addition, the repair ability of 1-CG on the UV damaged HaCaT cell is first to be reported. Conflict of interest

Fig. 5. IR spectrum for lipase Novozym 435 in different reactors.

The authors declare no competing financial interest. 48

Biochemical Engineering Journal 142 (2019) 41–49

X. Liu et al.

Acknowledgements

Sci. 81 (1992) 569–571. [17] J. Wang, S.S. Gu, H.S. Cui, X.Y. Wu, F.A. Wu, A novel continuous flow biosynthesis of caffeic acid phenethyl ester from alkyl caffeate and phenethanol in a packed bed microreactor, Bioresour. Technol. 158 (2014) 39–47. [18] Y. Xu, S. Sheng, X. Liu, C. Wang, W. Xiao, J. Wang, F.A. Wu, Cooperative reinforcement of ionic liquid and reactive solvent on enzymatic synthesis of caffeic acid phenethyl ester as an in vitro inhibitor of plant pathogenic bacteria, Molecules. 22 (2017) 0072. [19] A.M. Klibanov, Improving enzymes by using them in organic solvents, Nature 409 (2011) 241–246. [20] G.D. Yadav, S. Devendran, Lipase catalyzed synthesis of cinnamyl acetate via transesterification in non-aqueous medium, Process Biochem. 47 (2012) 496–502. [21] C. Balachandran, V. Duraipandiyan, N.A. Al-Dhabi, K. Balakrishna, N.P. Kalia, V.S. Rajput, I.A. Khan, S. Ignacimuthu, Antimicrobial and antimycobacterial activities of methyl caffeate isolated from Solanum torvum Swartz, Fruit Indian J. Microbiol. 52 (2012) 676–681. [22] J.K. Pittman, A.P. Dean, O. Osundeko, The potential of sustainable algal biofuel production using wastewater resources, Bioresour. Technol. 102 (2011) 17–25. [23] J. Wang, X.D. Wang, X.Y. Zhao, X. Liu, T. Dong, F.A. Wu, From microalgae oil to produce novel structured triacylglycerols enriched with unsaturated fatty acids, Bioresour. Technol. 184 (2015) 405–414. [24] X.G. Zou, J.N. Hu, M.L. Zhao, X.M. Zhu, H.Y. Li, X.R. Liu, R. Liu, Z.Y. Deng, Lipozyme RM IM-catalyzed acidolysis of Cinnamomum camphora seed oil with oleic acid to produce human milk fat substitutes enriched in medium-chain fatty acids, J. Agri. Food Chem. 62 (2014) 10594–10603. [25] S. Kundu, A.S. Bhangale, W.E. Wallace, K.M. Flynn, C.M. Guttman, R.A. Gross, K.L. Beers, Continuous flow enzyme-catalyzed polymerization in a microreactor, J. Am. Chem. Soc. 133 (2011) 6006–6011. [26] X.N. Li, T.J. Xiao, W.P. Pei, Y.L. Shi, Z.H. Shu, P.L. Yi, L. Yong, A. Kumud, Characterization of unsaturated fatty acid sustained-release microspheres for longterm algal inhibition, Chemosphere. 120 (2015) 383–390. [27] A.D. Kroumov, A.N. Módenes, B.M. Wenzel, Development of the enzymatic kinetics model of vegetable oils transesterification for biodiesel production, Acta Sci. Pol. Technol. Aliment. 29 (2007) 9–16. [28] M.M. Soumanou, U.T. Bornscheuer, Improvement in lipase-catalyzed synthesis of fatty acid methyl esters from sunflower oil, Enzyme Microb. Technol. 33 (2003) 97–103. [29] S. Matosevic, G.J. Lye, F. Baganz, Immobilised enzyme microreactor for screening of multi-step bioconversions: characterisation of a de novo transketolase-ω-transaminase pathway to synthesise chiral amino alcohols, J. Biotechnol. 155 (2011) 320–329. [30] J. Mutschler, T. Rausis, J.M. Bourgeois, D. Zufferey, I.V. Mohrenz, F. Fischer, Ionic liquid-coated immobilized lipase for the synthesis of methylglucose fatty acid esters, Green Chem. 11 (2009) 1793–1800. [31] Y. Zhang, L. Ma, J.C. Yang, Kinetics of esterification of lactic acid with ethanol catalyzed by cation-exchange resins, React. Funct. Polym. 61 (2004) 101–114. [32] I.B. Ju, H.W. Lim, W. Jeon, J.S. Dong, M.J. Park, Y.W. Suh, Kinetic study of catalytic esterification of butyric acid and n-butanol over Dowex 50Wx8-400, Chem. Eng. J. 168 (2011) 293–302. [33] E. Sert, A.D. Buluklu, S. Karakuş, F.S. Atalay, Kinetic study of catalytic esterification of acrylic acid with butanol catalyzed by different ion exchange resins, Chem. Eng. Process. 73 (2013) 23–28. [34] Y. Liu, G. Xu, J. Sun, H. Li, Investigation of the roughness effect on flow behavior and heat transfer characteristics in microchannels, Int. J. Heat Mass Transf. Theory Appl. 83 (2015) 11–20. [35] J. Liu, G. Xie, T.W. Simon, Turbulent flow and heat transfer enhancement in rectangular channels with novel cylindrical grooves, Int. J. Heat Mass Transf. Theory Appl. 81 (2015) 563–577. [36] C. Zhang, Y. Chen, M. Shi, Effects of roughness elements on laminar flow and heat transfer in microchannels, Chem. Eng. Process. 49 (2010) 1188–1192. [37] D.W. Sun, Lipase activity and conformation in neat organic solvents, J. Mol. Catal. B Enzym. 19 (2002) 93–102. [38] M.J. Earle, K.R. Seddon, Ionic liquids. Green solvents for the future, Pure Appl. Chem. 72 (2009) 10–25. [39] J. Qin, X. Zou, S. Lv, Q. Jin, X. Wang, Influence of ionic liquids on lipase activity and stability in alcoholysis reactions, RSC Advan. 6 (2016) 87703–87709. [40] S.W. Shin, E. Jung, S. Kim, K.E. Lee, J.K. Youm, D. Park, Antagonist effects of veratric acid against UVB-induced cell damages, Molecules. 18 (2013) 5405–5409.

This study was financially supported by the Key Research and Development Program (Modern Agriculture) of Jiangsu Province (BE2017322), the Six Talent Peaks Project of Jiangsu Province (2015NY-018), the 333 High-level Talent Training Project of Jiangsu Province (Year 2018), the Shen Lan Young scholars program of Jiangsu University of Science and Technology (Year 2015), the Postgraduate Innovation Project of School Biotechnology of Jiangsu University of Science and Technology (YCX2016X-07) and the Modern Agro-industry Technology Research System of China (CARS-18-ZJ0305). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.bej.2018.11.007. References [1] X.Y. Meng, Y. Xu, J.X. Wu, C.T. Zhu, D.Y. Zhang, G.H. Wu, F.A. Wu, J. Wang, Enzymatic synthesis and antioxidant activity of 1-caffeoylglycerol prepared from alkyl caffeates and glycerol, J. Am. Oil Chem. Soc. 95 (2018) 149–159. [2] J.A. Laszlo, D.L. Compton, F.J. Eller, S.L. Taylor, T.A. Isbell, Packed-bed bioreactor synthesis of feruloylated monoacyl-and diacylglycerols: clean production of a “green” sunscreen, Green Chem. 5 (2013) 382–386. [3] S. Sun, B.A. Hu, Novel method for the synthesis of glyceryl monocaffeate by the enzymatic transesterification and kinetic analysis, Food Chem. 214 (2017) 192–198. [4] S. Sun, X. Hou, B. Hu, Synthesis of glyceryl monocaffeate using ionic liquids as catalysts, J. Mol. Liq. 248 (2017) 643–650. [5] P.L. Urban, D.M. Goodall, N.C. Bruce, Enzymatic microreactors in chemical analysis and kinetic studies, Biotech. Advan. 24 (2006) 42–57. [6] K.F. Schilke, K.L. Wilson, T. Cantrell, G. Corti, D.N. Mcilroy, C.A. Kelly, Novel enzymatic microreactor with Aspergillus oryzae β-galactosidase immobilized on silicon dioxide nanosprings, Biotech. Prog. 26 (2010) 1597–1605. [7] A. García, E. Rodríguez-Juan, G. Rodríguez-Gutiérrez, J.J. Rios, J. FernándezBolaños, Extraction of phenolic compounds from virgin olive oil by deep eutectic solvents (DESs), Food Chem. 197 (2016) 554–561. [8] Z. Hua, G.A. Baker, S. Holmes, New eutectic ionic liquids for lipase activation and enzymatic preparation of biodiesel, Org. Biomol. Chem. 9 (2011) 1908–1916. [9] T. Muppaneni, H.K. Reddy, S. Ponnusamy, P.D. Patil, Y. Sun, P. Dailey, S. Deng, Optimization of biodiesel production from palm oil under supercritical ethanol conditions using hexane as co-solvent: a response surface methodology approach, Fuel 107 (2013) 633–640. [10] V. Gekas, M. Lopez-Leiva, Hydrolysis of lactose-A literature review, Process Biochem. 20 (1985) 2–12. [11] A.P. Jia, G.S. Hu, L. Meng, Y.L. Xie, J.Q. Lu, M.F. Luo, CO oxidation over CuO/ Ce1−xCuxO2−δ and Ce1−xCuxO2−δ catalysts: synergetic effects and kinetic study, J. Catal. 289 (2012) 199–209. [12] R.D. Vasic, U. Kragl, A. Liese, Benefits of enzyme kinetics modeling, Chem. Biochem. Eng. Q. 17 (2003) 7–18. [13] C. Ülger, S. Takaç, Kinetics of lipase-catalysed methyl gallate production in the presence of deep eutectic solvent, Biocatal. Biotrans. 35 (2017) 1–10. [14] C.M. Galanakis, A. Patsioura, V. Gekas, Enzyme kinetics modeling as a tool to optimize food industry: a pragmatic approach based on amylolytic enzymes, Crit. Rev. Food Sci. 55 (2015) 1758–1770. [15] Y. Zheng, L. Ye, L. Yan, Y. Gao, The electrochemical behavior and determination of quercetin in choline chloride/urea deep eutectic solvent electrolyte based on abrasively immobilized multi-wall carbon nanotubes modified electrode, Int. J. Electrochem. Sci. 9 (2014) 238–248. [16] D. Farthing, T. Karnes, T.W. Gehr, C. March, I. Fakhry, D.A. Sica, External-standard high-performance liquid chromatographic method for quantitative determination of furosemide in plasma by using solid-phase extraction and on-line elution, J. Pharm.

49