- Email: [email protected]
Efficient mono-acylation of fructose by lipase-catalyzed esterification in ionic liquid co-solvents
Accepted Manuscript Title: Efficient Mono-acylation of Fructose by Lipase-catalyzed Esterification in Ionic Liquid co-solvents Author: Lu Li, Fangling Ji, Jingyun Wang, Bo Jiang, Yachen Li, Yongming Bao PII: DOI: Reference:
S0008-6215(15)00231-1 http://dx.doi.org/doi: 10.1016/j.carres.2015.08.009 CAR 7053
To appear in:
Carbohydrate Research
Received date: Revised date: Accepted date:
2-7-2015 17-8-2015 17-8-2015
Please cite this article as: Lu Li, Fangling Ji, Jingyun Wang, Bo Jiang, Yachen Li, Yongming Bao, Efficient Mono-acylation of Fructose by Lipase-catalyzed Esterification in Ionic Liquid cosolvents, Carbohydrate Research (2015), http://dx.doi.org/doi: 10.1016/j.carres.2015.08.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Efficient Mono-acylation of Fructose by Lipase-catalyzed Esterification in Ionic Liquid co-solvents
Lu Li1, Fangling Ji1, Jingyun Wang1, Bo Jiang1, Yachen Li2*, Yongming Bao1*
1
School of Life Science and Biotechnology, Dalian University of Technology, Dalian,
China 116024 2
Department of Occupational and Environmental Health, Dalian Medical University,
Dalian, China 116044
∗
To whom correspondence should be addressed: School of Life Science and
Biotechnology, Dalian University of Technology. NO. 2 Linggong Road, Ganjingzi District, Dalian, 116024, China. Tel: +86-411-84706344. Fax: +86-411-84706344; E-mail: [email protected], [email protected].
1
Page 1 of 32
ABBREVIATIONS ILs, ionic liquids; 2M2B, 2-methyl-2-butanol; CALB, Candida antarctica lipase B; [BMIM][TfO], 1-butyl-3-methylimidazolium trifluoromethanesulfonate; [BMIM][BF4], 1-butyl-3-methylimidazolium
tetrafluoroborate;
[BMIM][PF6],
1-butyl-3-methylimidazolium hexafluorophosphate; ATR-FTIR, attenuated total reflectance-fourier transform infrared; HPLC-ELSD, high performance liquid chromatography-evaporative light scattering detector; NMR, nuclear magnetic resonance; API-MS, atmospheric pressure ionization-mass spectrometry; ANOVA, analysis of variance.
2
Page 2 of 32
Highlight • Markedly improved fructose mono-acylation catalyzed by CALB in ILs/2M2B co-solvent system. • Selectivity of mono-acylation in ILs/2M2B co-solvent governed by the change of CALB kinetic parameters • Kinetic parameters variance determined by conformational changes of CALB binding substrates. Graphical Abstract
ABSTRACT Fructose monoesters are eco-friendly nonionic surfactants in various applications. Selective preparation of mono-acylated fructose is challenging due to the multiple hydroxyl sites available for acylation both chemically and enzymatically. Ionic liquids 3
Page 3 of 32
(ILs) have profound impacts not only on the reaction media but also on the catalytic properties of enzymes in the acylation process. In this study, utilizing an IL co-solvents system, selective synthesis of mono-acylated fructose with lauric acid catalyzed by immobilized Candida antarctica lipase B (CALB) was investigated. The imidazolium-based ILs selected as co-solvents with 2-methyl-2-butanol (2M2B), markedly improved the ratios of monolauroyl fructose in the presence of 60% [BMIM][TfO] (v/v) and 20% [BMIM][BF4] (v/v), in which the mono-acylated fructose was 85% and 78% respectively. Based on a Ping-Pong Bi−Bi model, a kinetic equation was fitted, by which the kinetic parameters revealed that the affinity between fructose and acyl-enzyme intermediate was enhanced. The inhibition effect of fructose on free enzyme was weakened in the presence of IL co-solvents. The conformation of CALB binding substrates also changed in the co-solvent system as demonstrated by Fourier transform infrared spectra. These results demonstrated that the variation of CALB kinetic characteristics was a crucial factor for the selectivity of mono-acylation in ILs/2M2B co-solvents. Keywords: Sugar ester of fatty acid; Monolauroyl fructose; Candida antarctica lipase B; Ionic liquids; Kinetics
1. INTRODUCTION Sugar esters are biodegradable and nontoxic bio-surfactants which are very useful for many commercial applications. One of the attractions is fructose laurate, due to its highest cell growth inhibitory against Streptococcus mutans among different sugar 4
Page 4 of 32
esters1. Chemical synthesis tends to result in high energy cost and environmentally unsafe conditions. As a potential alternative, enzymatic preparation has mild and eco-friendly reaction conditions, as well as the inherent selectivity maintained by enzymes during reaction process2. Lipases (EC 3.1.1.3) are considered as better catalysts for esterification, amidation and transesterification of esters in anhydrous media3. With regard to the direct esterification of fructose with lauric acid, traditional chemical reactions can result in a broad distribution of one to multi-ester species in the products. On the other hand, lipases selectively acylate the primary hydroxyl group rather than secondary groups with long chain fatty acid in anhydrous organic solvents. As the result, the products are composed of a narrow range of mono- and di-esters (Figure 1), and it is more convenient for the separation of products. As reported, Cao’s4 and Šabeder’s co-workers5 obtained relatively higher conversions of fructose esters with lipases-mediated acylation, achieving acylation rates of 86% and 82%, respectively. However, these products were mixtures of esters in different esterification degrees, with a higher proportion of di-ester in the products. This phenomenon is considered inevitable in the direct esterification process. Ionic liquids (ILs) have been recognized as an alternative to traditional organic solvents for the synthesis of sugar esters, not only because they offer many advantages as solvents, such as their low vapor pressure, high thermal stability and good solubilization of a wide range of carbohydrates6, but also as reaction media that can enhance the reactivity, selectivity, and stability of enzymes7-10. Several groups have 5
Page 5 of 32
reported that anhydrous ILs containing [BF4]-, [PF6]- and [TfO]- anions can be used for biotransformation of sugars. In these ILs, desired monoacylated compounds were synthesized and/or high ester yields were achieved10-12. Co-solvent system of ILs/organic solvent is a balanced approach between mass transfer limitation of pure ILs and selectivity of lipase. Our previous work revealed that in pure 2M2B, CALB displayed selectivity towards low degree of esterification products (mono-acylated fructose) and was an unfavorable solvent for di-ester synthesis13. Therefore, with the goal of further enhancing the selectivity and obtaining higher proportion of mono-acylated fructose laurate, our study utilized three imidazolium-based ionic liquids ([BMIM][BF4], [BMIM][PF6] and [BMIM][TfO]) to constitute an ILs/2M2B co-solvent system as the reaction media in the acylation process of fructose. To probe the key factors that affect the activity and selectivity in mono-ester synthesis, we first experimentally surveyed the productivity of fructose laurate and the selectivity of mono-acylation in different volume ratios of the ILs/2M2B co-solvent system. Then the optimal reaction media were further studied for the characteristics that are particularly related to the solubility of substrate, mass transfer limitation, and the kinetics and structure variation of CALB. 2. RESULTS AND DISCUSSION 2.1 Effect of volume ratios in the ILs/2M2B co-solvent system For the enzymatic direct esterification of fructose with lauric acid, high conversion rate was obtained by the approach of providing excessive insoluble fructose in a small 6
Page 6 of 32
amount of organic solvent-adjuvant5,4. Based on this protocol, we set about conducting the reaction process in co-solvents of ILs/2M2B. In our case, we expected an increase of mono-acylated product in such reaction media. Having developed an optimized co-solvent system, the synthesis process was launched at different volume percentages of ILs in 2M2B. For comparison, the synthetic reaction was also assayed in pure 2M2B and pure ILs. HPLC and mass spectrometry revealed that, within the reaction time of 12 h in either pure 2M2B or co-solvent system, only mono-acylated fructose laurate and di-acylated fructose laurate was detectable in the products. The structures of products were further confirmed by
13
C NMR spectroscopy, a method employed in previous
reports14,15. We found that mono-ester was a mixture composed of 1-α-fructofuranose laurate, 1-β-fructopyranose laurate, 6-lauroyl-α-D-fructofuranose and 6-lauroyl-β-Dfructofuranose; the di-esters were composed of 1, 6-dilauroyl-α-D-fructofuranose laurate and 1, 6-dilauroyl-β-D-fructofuranose laurate. As shown in Figure 2, the addition of ionic liquids resulted in fluctuation in the yields of products and especially the increased selectivity of mono-acylation. The proportion of mono-acylated product had a clear improvement (P < 0.05) in the [BMIM][BF4]/2M2B and [BMIM][TfO]/2M2B co-solvent system. As shown in Figure 2-A, we found that with the increase of [BMIM][BF4] component, yield and selectivity of mono-acylation content were gradually increased, beyond which further increase in ILs content resulted in a substantial decline. At the point of 20% [BMIM][BF4], we detected the highest yield of mono-ester (27.05%) and significantly higher selectivity 7
Page 7 of 32
of mono-acylation (78.30%) than pure 2M2B (72.28%) and pure IL (33.57%). After that point, yield of mono-ester decreased and the overall conversion rate was low. Meanwhile, the increase of di-ester led to poor selectivity for mono-ester synthesis. Similarly, enhanced yield of mono-ester was observed with the increase of [BMIM][TfO] ratios in co-solvent system (Figure 2-B). The selectivity of mono-acylation achieved the highest level in ratio of 60% [BMIM][TfO]/2M2B (83.33%). However, unlike the case in [BMIM][BF4]/2M2B, pure [BMIM][TfO] as solvent did not bring about a significant decline in conversion rate and selectivity. Compared to the case in pure 2M2B, it still maintained at a relatively high level. [BMIM][PF6] was also investigated as co-solvent in the reaction, where obvious decrease of both conversion and mono-acylation was detected (data not shown). Compared to other ILs, almost no activity of CALB presented in pure [BMIM][PF6]; the same phenomenon was also observed by Devi et al. in CALB-catalyzed production of biodiesel in this ionic liquid16. Since [BMIM][PF6] has the highest viscosity and poor solubility of fructose (Table 1), one possible explanation is that enzyme reaction in [BMIM][PF6] can be limited by the dissolution rate and mass transfer rate of substrate. It is worth noting that higher conversion and selectivity of mono-acylation was achieved in high ratio of [BMIM][TfO]/2M2B, but not in [BMIM][BF4]/2M2B. In addition, 20% [BMIM][BF4]/2M2B co-solvent system resulted in significantly higher selectivity of mono-acylation than 2M2B did, even though they have almost similar fructose solubility and viscosity. Since the three ionic liquids used are notably different in their properties, such as polarity, hydrophobicity, anion nucleophilicity, 8
Page 8 of 32
hydrogen-bond basicity and viscosity17, and also that they possess distinct IL ratios and solubility of fructose (Table 1), we reasoned that not only the solubility of the substrate, especially fructose, but also limitation of mass transfer and functional interaction between enzyme and ILs, or the combined effects of above could be accountable for these phenomena. To investigate the key factors in dominating mono-acylated product yields and selectivity of mono-acylation in different co-solvents system, a series of experiments addressing these aspects were designed and carried out below. 2.2 Effect of rotation To understand the effect of mass transfer in a co-solvent system, reactions were proceeded under different rotating speeds in a 20% [BMIM][BF4]/2M2B and 60% [BMIM][TfO]/2M2B co-solvent system (Figure 3). As rotating speeds shifted, reaction rate gradually improved in both solvents. However, when speed exceeded 200 rpm, the growth rate for yields of both products slowed down. From Figure 3-A, yields of mono-ester and di-ester in 20% [BMIM][BF4]/2M2B mixture increased at the same rate, the selectivity of mono-acylation was not significantly affected by mass transfer in this co-solvents system. While in the [BMIM][TfO] group (Figure 3-B), poor selectivity of mono-acylation was found at low speeds (below 200 rpm). This feature was significantly improved with a further increase in rotating speed, but it was not enhanced above 200 rpm. Many previous reports indicated that viscosity of ILs may have an influence on reaction rate and enzyme activity18,19. As mentioned above, selectivity of mono-acylation was also changed in the [BMIM][TfO] group. This phenomenon is 9
Page 9 of 32
likely to be explained by the dissolution rate of fructose. Mass transfer rate is closely related to dissolution and diffusion rate of substrates. Since mono-ester was far more soluble than fructose in our co-solvent system, the amount of soluble fructose is particularly important when enzyme chose its acyl acceptor. At a low speed, the more dissolution rate was limited, the less soluble fructose interacted with enzyme. Compared with 20% [BMIM][BF4]/2M2B mixture, this effect was more prominent in medium with a higher ratio of [BMIM][TfO], which has higher solubility of fructose and viscosity (Table 1). As rotating speed increased, this limited effect gradually weakens. Selectivity of mono-acylation in 60% [BMIM][TfO]/2M2B was independent of the rotating rate for values of the rate above 200 rpm. Taken together, we reasoned that selectivity of mono-acylation in a two optimal co-solvent system was not primarily determined by mass transfer limitation set by viscosity in the adopted speed of our experiment (200 rpm). 2.3 Reaction time course In comparison of specific reaction process with each reaction mixtures, Figure 4 depicts the time course of the conversion and selectivity of mono-acylation catalyzed by CALB in 20% [BMIM][BF4]/2M2B mixture, 60% [BMIM][TfO]/2M2B mixture and pure 2M2B at 50°C. As shown, for all of the co-solvent systems tested, acylated products of the reaction appeared in a sequential way, i.e., the formation of di-ester is associated with the depletion of substrates and the formation of a substantial amount of mono-acylated products. At the initial stage of the reaction, only monoester was produced. The formation of monoester in all of the co-solvent systems proceeds rapidly 10
Page 10 of 32
within 50 min, far more than that in pure 2M2B. The amount of monoester in 60% [BMIM][TfO]/2M2B mixture was increased to 0.045 mmol, and 0.021 mmol in 20% [BMIM][BF4]/2M2B mixture, whereas 0.009 mmol in pure 2M2B. After 50 min, the velocity of mono-ester formation in all of the co-solvent system tested slowed down. Di-esters for all of the tested solvents were detectable at that point and then gradually increased in a nearly identical formation velocity. Based on this, we concluded that, within 12 h, it is the synthesis of mono-ester that leads to such selectivity during the acylation process. In other words, within 12 h the reaction step of fructose to mono-ester synthesis determines the selectivity of mono-acylation but not the synthesis step from monoester to di-ester. Consequently, as no di-ester was detected during the initial reaction period, the kinetics study of monoester formation can be greatly simplified. 2.4 Kinetic constants Direct esterification of lauric acid with fructose catalyzed by a lipase is often modeled using the Ping Pong Bi−Bi kinetic mechanism20-22. For this mechanism, a general scheme for this type of reactions was established (Figure 5). The initial rates V is expressed as in the following equation (3)， v
V m a [x F ] [ L A ] K m F[ L A]
Km
L
[A F]
(3) [ F ] [L A]
Where v is the initial reaction rate; Vmax is the maximum reaction rate; [LA] and [F] are the concentrations of lauric acid and fructose, respectively; KmLA and KmF are kinetic constants for lauric acid and fructose, respectively. 11
Page 11 of 32
If there are inhibitions from any of the substrates, it is necessary to include the corresponding inhibition constants in the equation. From Figure 6, it showed that lauric acid does not negatively affect the reaction rate. Thus, a Ping-Pong Bi−Bi kinetic model with competitive inhibition by fructose can be established as shown in the following equation (4)， V m a x [ F ][ L A ]
v
K m F [ L A ] K m L A [ F ](1
[F ]
(4) ) [ F ][ L A ]
Ki
Where Ki is the inhibition coefficient of fructose. Considering that the fructose dissolution rate could complicate the analysis of enzyme kinetics, fructose concentration was fixed at the saturation point in our tested reaction media. However, limited solubility of fructose in 2M2B provided us narrow intervals to test the effect of fructose concentrations on the reaction rate. Much lower concentration of fructose would not be detectable in the initial period of reactions or could lead to larger errors via HPLC. Therefore, we intended to fit our available data with both of the equations. Kinetic constants-Vmax, KmLA, KmF and Ki were estimated from the initial rate data (Figure 6) coupled with equation (3) and equation (4), and refined by using the Polymath 5.1 software package. By knowing the experimental initial rates with respect to initial concentrations, all of the rate constants were calculated through nonlinear regression to the rate equation. The best fit values of the two equations in pure 2M2B obtained are presented as examples in Table 2. The fitting for both of the equations was excellent, with R2 and adjusted R2 being above 0.98. This indicates that the theoretical 12
Page 12 of 32
model and the actual value are in agreement. As shown in Table 3, in the column without fructose inhibition group, KmF was less than KmLA, indicating that the affinity of the enzyme towards fructose was much greater than that to lauric acid. It is reasonable to assume that such affinity can induce fructose to form a dead-end complex by binding to the enzyme. On the other hand, in the inhibition group the higher affinity of enzyme towards lauric acid could fully describe this mechanism (Figure 5) with acyl donor as the first binding substrate23. Thus, we believe that this fitting model may exhibit more rational kinetic constants. The final kinetic equation with inhibition constant was used to estimate the initial reaction rates. By using kinetic constant values (Table 3), the preference of the CALB for different substrates in these co-solvent systems can be compared. With the addition of ILs, obtained values of kinetic constants exhibited marked differences. As expected, the highest Vmax (0.230 mmol·min-1·g-1) was found in 60% [BMIM][TfO]/2M2B co-solvent system, followed by 20% [BMIM][BF4]/2M2B (0.194 mmol·min-1·g-1), with the lowest one (0.187 mmol·min-1·g-1) being in 2M2B. As shown in Table 3, the increase of KmLA in three solvents followed the same order as Vmax, suggesting that the affinity between enzyme and lauric acid declined with the addition of ILs. This effect can be seen in Figure 6, where the reaction rate was faster in 2M2B than in co-solvent systems at low concentration of lauric acid. With the increase of lauric acid concentration, the relatively lower fructose concentration started to determine the whole reaction rate when fructose concentrations were fixed in these reaction systems. Most notably, the acyl-CALB (Figure 5) affinity herein for fructose was enhanced in 13
Page 13 of 32
the co-solvent systems (reduced KmF). Moreover, the inhibition effect of fructose in both of the co-solvent systems was weakened compared with pure 2M2B (increased Ki). Collectively, these varied affinities contributed to the higher reaction rate in the co-solvent systems at high lauric acid concentration. 2.5 The secondary structure of immobilized CALB by FTIR spectrometry The interactions of enzyme-solvent play an important role in enzyme structure and function. It is reasonable to speculate that the conformational change in secondary structure caused by protein-solvent interactions is a probable cause for the variation of lipase activity, or even the selectivity. It has been reported that organic solvent and ILs can drive conformational changes and affect enzyme stability and activity24-28. To monitor secondary conformational variations of enzyme binding substrates in the different ILs/2M2B co-solvent system, we calculated the composition of the lipase secondary structure with the aid of ATR-FTIR spectroscopy. First, in the control group (Table 4), no substrate was found to be bound to CALB in the co-solvent system. There are obvious changes in the composition of the secondary structure in different solvents. The addition of ILs changed the whole conformation of lipase. We believe that these changes would contribute to the variation in activity and selectivity in our experiments. When CALB binding to lauric acid and fructose, respectively, as revealed by the kinetics studies, second structure changed in co-solvent systems. Our results provided direct evidence for the changes in substrate affinity and inhibition. With one exception, however, there was no obvious changes in fructose binding groups between pure 2M2B and 20% [BMIM][BF4]/2M2B. When the control group was compared with the 14
Page 14 of 32
binding groups in corresponding co-solvent system, notable conformation changes were found when CALB binds to fructose in pure 2M2B. The proportions of both α-helix and β-sheet reduced from 15.35% to 10.63% and from 42.06% to 34.68%, respectively, with the other structural contents increased accordingly. In contrast, in 20% [BMIM][BF4]/2M2B, the proportions of secondary structures were nearly identical. This more flexible enzyme structure observed in pure 2M2B that seem to increase the inhibition effect of fructose on free enzyme. By contrast, the flexibility of CALB was limited in the co-solvent system. Based on the study of molecular dynamic simulations by Patrick R. Burney and Jim Pfaendtner29, it was also found that ILs significantly dampen protein dynamics of lipase, and consequently, it is difficult for ILs to drive enzyme deviating from its initial structure compared with organic solvent. Therefore, ILs component does shape the enzyme in response to the function of lipase. 2.6 Operational stability The operational stability or reusability of CALB in ILs/2M2B co-solvents system was achieved by 8 batches esterification of fructose with lauric acid at the same condition (Figure 7). The relative activity of CALB was determined by mono-ester yield of each batch to the corresponding value of the first batch, taking the relative activity of the first batch as 100%30. As shown in Figure 7, although the activity of CALB in both co-solvents system were reduced with increased batches, the relative activities were remained ca. 80% in 20% [BMIM][BF4]/2M2B after 6 batches and ca. 60% in 60% [BMIM][TfO]/2M2B after 7 batches, respectively. It indicated that 2M2B exerted certain protective effects 15
Page 15 of 32
on enzyme activity against inactivation by ILs. The decrease of enzyme relative activity (P < 0.05) from third batch might be explained by the inactivation of enzyme, rather than effects by substrates or products. In the next 5 batches, differences of enzyme relative activity were not significant (P > 0.05) among these batches, and it indicated that CALB could maintain good stability in both co-solvents system. Moreover, the enzyme relative activities were slightly increased, and Joel L. Kaar et al.31 speculated that ionic liquids could swell the acrylic resin (immobilized carrier) causing improved mass transfer. 3. CONCLUSION In the present work, the catalytic behavior of CALB was investigated in the acylation of fructose with lauric acid in the ILs/2M2B co-solvent system. For improving the mono-ester synthesis, three ILs which were reported as favorable solvents in mono-acylation and possessing different dissolving capacities of fructose were employed as co-solvents with 2M2B. As expected, higher selectivity of mono-acylation was obtained by using the co-solvents composed of [BMIM][TfO] and [BMIM][BF4], respectively. In an attempt to elucidate the underlying mechanisms, we studied the probable causes from the addition of ILs for the changes in solubility of substrates, mass transfer, kinetics features and structural changes of CALB. It is evident from the time-course study that mono-ester synthesis determines the selectivity of mono-acylation in the entire reaction process. Mass transfer limitation caused by ILs viscosity showed that such enhanced selectivity was independent of rotating speed in our operational conditions. With A Ping-Pong Bi−Bi kinetic model and with dead-end 16
Page 16 of 32
complex inhibition of fructose fitting our experimental data, kinetic feature of CALB was affected by co-solvent system, in which the affinities and inhibition of substrates were changed. These observations were also supported by the conformational studies via FTIR when CALB binds different substrates. In conclusion, the variation of enzyme kinetic characteristics of CALB was found to be a crucial parameter in governing mono-ester productivity and selectivity of mono-acylation in ILs/2M2B co-solvents. Furthermore, our data showed that lipase catalysis can be an ideal and favorable approach for mono-acylation of polyhydroxy substrates, such as fructose. 4. EXPERIMENTAL 4.1 Materials Novozym 435 was a gift from Novozymes (China) Investment Co. Ltd (Beijing, China). Lewatit VP OC 1600 resin was also a gift from LANXESS Chemical (China) Co.Ltd. Commercial ionic liquids, 1-butyl-3-methylimidazolium trifluoromethanesulfonate
([BMIM][TfO]),
1-butyl-3-methylimidazolium
tetrafluoroborate
([BMIM][BF4]) and 1-butyl-3-methylimidazolium hexafluorophosphate ([BMIM][PF6]) with purity of 99% were purchased from Shanghai Cheng Jie Chemical Co. Ltd. (Shanghai, China). D-fructose (≥99%) (Amresco, USA), lauric acid (≥98.0%) (Sinopharm, Shanghai, China), 2-methyl 2-butanol (2M2B, 98%) (Guang Fu, Tianjin, China). All other chemicals were also from commercial sources and of the highest purity available.
17
Page 17 of 32
4.2 General procedure for lipase-catalyzed esterification in ILs/2M2B co-solvent system All reactions were carried out in 1.5 ml clear reaction vial sealed with a screw cap. Equimolar fructose and lauric acid (0.222 mmol) and 100 mg of molecular sieve (3 Å) were added into 0.6 ml mixed solvents of [BMIM][TfO]/2M2B, [BMIM][BF4]/2M2B and [BMIM][PF6]/2M2B, respectively. These mixtures were then placed in an air-bath shaker for 12 h at 200 rpm and 50 °C to dissolve certain amount of substrates. The reactions were initiated by adding 25 mg CALB into each mixture. Periodically, these samples were extracted by the addition of 0.6 ml methanol and 15 s of vortexing shake. 50 μl of liquid phase were withdrawn and analyzed via high performance liquid chromatography (HPLC). 2M2B and ILs were dehydrated with molecular sieves (4 Å) over 48 h before reactions. All measurements were conducted in triplicate. 4.3 Yields of products and selectivity of mono-acylation The yield of products was calculated by the following equation (1): Y ie ld (% )=
M
P
M
F
100
(1)
Where MP is the generated amount of products (mole), MF is the initial amount of fructose (mole). The selectivity of mono-acylation was defined by the following equation (2): M
S e le c tiv ity ( % ) = M
ME
100
ME
M
(2)
DE
Where MME and MDE are the generated amount of mono-ester and di-ester (mole), respectively. The generated mono-ester and di-ester were measured by with a HPLC 18
Page 18 of 32
system (Waters 600 controller) equipped with an evaporative light scattering detector (ELSD) and a 150 mm × 4.6 mm (5 μm) Hypersil ODS2-C18 column. 50 μl samples from the reaction mixture were withdrawn at final reaction times, then filtered and injected into a HPLC. The temperature of drift tube in ELSD was set at 45°C at the pressure of 20 psi (nitrogen gas). The mobile phase system was acetonitrile/water (v/v). The gradient eluted program was used as following profile: 70% acetonitrile at a flow rate of 0.9 ml/min from 0-4 min, followed by 100% acetonitrile from 4-20 min at a flow rate of 1 ml/min. The retention time for lauroyl-D-fructose was 3.4 min, and for 1,6-diacyl-D-fructofuranose was 14 min. Mono-ester and di-ester were isolated and collected by flash column chromatography (elution with petroleum ether and ethyl acetate sequentially). The purity of collected fractions was determined by HPLC. 4.4 Structural characterization of fructose lauric acid esters 13
C NMR spectra were obtained from Varian INOVA 400 (101 MHz for
13
C)
spectrometer in CD3OD. Atmospheric pressure ionization (API) mass spectra were recorded on HP1100LC/MSD spectrometer. Shifts of fructose carbons of four isomers of fructose mono-esters were recorded as follows: (a) 6-lauroyl-β-D-fructofuranose, 13
C NMR (101 MHz, CD3OD): δ = 103.39 (C-2), 80.09 (C-5), 77.43 (C-3), 76.96 (C-4),
66.74 (C-6), 64.34 (C-1). (b) 1-lauroyl-β-D-fructopyranose,
13
C NMR (101MHz,
CD3OD): δ = 98.16 (C-2), 71.39 (C-4), 70.89 (C-5), 69.71 (C-3), 66.58 (C-1), 64.66 (C-6). (c) 6-lauroyl-α-D-fructofuranose,
13
C NMR (101MHz, CD3OD): δ = 106.00
(C-2), 84.06 (C-3), 80.42 (C-5), 79.00 (C-4), 65.42 (C-6), 64.92 (C-1). (d) 1 -lauroyl-β-D-fructofuranose,
13
C NMR (101MHz, CD3OD): δ = 101.44 (C-2), 83.31 19
Page 19 of 32
(C-5), 78.43 (C-3), 76.33 (C-4), 65.56 (C-1), 63.98 (C-6). Shifts of groups of four lauric acid isomers: δ = 175.35, 175.29, 175.26, 174.89 (C=O, 4 isomers), 25.70-34.89 (CH2), 14.45 (CH3). API-MS [M + Na] + m/z: 385. Shifts
of
the
two
isomers
1,6-dilauroyl-β-D-fructofuranose, 80.15
(C-5),
77.96
(C-3),
1,6-dilauroyl-α-D-fructofuranose,
of
13
fructose
di-esters
are
as
follows:
(a)
C NMR (101MHz, CD3OD) δ = 101.67 (C-2),
77.06 13
(C-4),
66.72
(C-1),
65.30
(C-6).
(b)
C NMR(101MHz, CD3OD) δ = 104.90 (C-2),
83.73 (C-3), 80.54 (C-5), 78.55 (C-4), 65.92 (C-1), 65.30 (C-6). Shifts of acyl skeleton: δ = 175.26, 174.77 (C=O, β-anomer), 175.14, 175.06 (C=O, α-anomer), 23.73-34.92 (CH2), 14.48 (CH3). API-MS [M + Na] + m/z: 567. 4.5 Measurement of fructose solubility Fructose (0.222 mmol) was added to glass vials containing 0.6 ml 2M2B, ILs or ILs/2M2B, respectively. The suspension was stirred at 200 rpm 50°C After 12 h, the supernatant was removed after centrifugation of the suspension. Then 1 ml deionized water was added to the vial. The residual fructose was dissolved to determine sugar content. The fructose concentration was confirmed by DNS method with a fructose standard. 4.6 Determination of kinetic constants in co-solvent system Initial reaction rates, expressed as micromole produced mono-ester per minute and per gram of enzyme, were determined by the amount of lauroyl-D-fructose generated within 50 min via HPLC. To avoid fructose dissolution rate complicating the analysis of enzyme kinetics, different solvents with saturated fructose prior to the reaction were 20
Page 20 of 32
incubated and the concentration of fructose were kept constant. The effect of lauric acid concentration on the reaction rate was investigated by esterifying fixed initial quantities of fructose with different initial concentrations of lauric acid (20 mM to 600 mM). In all experiments it was used an input amount of 25 mg immobilized CALB. In order to get the parameters of the kinetic model, initial rates were fitted to the proposed reaction rate equation by non-linear regression analysis with the computer program polymath 5.1. 4.7 Conformation analysis of CALB by Fourier Transform Infrared (FTIR) spectrometry Aided by attenuated total reflectance (ATR) - FTIR spectroscopy, the secondary structure of immobilized CALB can be directly measured, which caused by the characteristic peptide group vibrations, amide I band (1700-1600 cm-1) that arises from the C=O stretching vibrations32. Prior to the essay, 50 mg immobilized CALB were incubated for 30 min with 46.30 mM dissolved fructose or lauric acid in 2M2B, 60% [BMIM][TfO]/2M2B mixture and 20% [BMIM][BF4]/2M2B mixture, respectively. Then experiments were performed without substrates under same conditions as control groups. Besides, for subtracting the absorption of immobilized enzyme carrier (Lewatit VP OC 1600 resin), the carrier was strictly treated and collected under same conditions. FTIR absorption spectra from 4000 to 640 cm-1 were collected using a Nicolet 6700 FTIR spectrometer equipped with a nitrogen-cooled, mercury-cadmium-tellurium detector at a resolution of 4 cm-1. Spectroscopic data were recorded and analyzed with 21
Page 21 of 32
Thermo Scientific OMNICTM software (OMNIC 8.2.387, Thermo Fisher Scientific Inc.). The spectra were smoothed with a binomial 5 points smoothing function, then second derivative spectra were obtained following the Savitsky–Golay method (3rd grade polynomial)33. The center position for each amide I sub-peak from 1700 to 1600 cm−1 was determined on the basis of previous assignments34-36 and confirmed by using second-derivative peak analysis. Fourier-self-deconvolution of the amide I band was performed by using an enhancement factor K=2.4 and half-bandwidth=20 cm−1. Curve-fitting process was accomplished with Gaussian function by employing OMNIC software. 4.8 Statistical analysis Statistical analyses were performed using computer program OriginPro (version 8.5). A one-way analysis of variance (ANOVA) with Tukey test was applied to determine significant differences among means. The significance value was set at p < 0.05.
ACKNOWLEDGEMENT This work was supported by the Major Basic Research Program of China (Grant No. 2009CB724700). We thank Novozymes (Beijing, China) for kindly supplying immobilized lipase and LANXESS Chemical (Shanghai, China) for their generous 22
Page 22 of 32
donation of macroporous resin. We thank Limei Luo (Analysis Center, School of Chemical Engineering) and Kun Jin (State Key Laboratory of Fine Chemicals) for their assistances in FTIR spectroscopy and NMR measurement.
REFERENCES 1.
Watanabe, T.; Katayama, S.; Matsubara, M.; Honda, Y.; Kuwahara, M. Curr Microbiol 2000, 41, 210-213.
2.
Pyo, S. H., and D. G. Hayes. . In Biobased Surfactants and Detergents: 23
Page 23 of 32
Synthesis, Properties, and Applications; D.G. Hayes, D. K., D.K.Y. Solaiman, and R.D. Ashby Ed.; AOCS Champaign, IL, 2009; pp. 323-350. 3.
Gandhi, N. N.; Patil, N. S.; Sawant, S. B.; Joshi, J. B.; Wangikar, P. P.; Mukesh, D. Catalysis Reviews 2000, 42, 439-480.
4.
Cao, L.; Fischer, A.; Bornscheuer, U. T.; Schmid, R. D. Biocatal Biotransform 1996, 14, 269-283.
5.
Šabeder, S.; Habulin, M.; Knez, Ž. J Food Eng 2006, 77, 880-886.
6.
Zakrzewska, M. E.; Bogel-Łukasik, E.; Bogel-Łukasik, R. Energy Fuels 2010, 24, 737-745.
7.
Ganske, F.; Bornscheuer, U. T. Org Lett 2005, 7, 3097-3098.
8.
Kim, M.-J.; Choi, M. Y.; Lee, J. K.; Ahn, Y. J Mol Catal B: Enzym 2003, 26, 115-118.
9.
Klähn, M.; Lim, G. S.; Seduraman, A.; Wu, P. PCCP 2011, 13, 1649-1662.
10.
Park, S.; Kazlauskas, R. J. J Org Chem 2001, 66, 8395-8401.
11.
Katsoura, M.; Polydera, A.; Katapodis, P.; Kolisis, F.; Stamatis, H. Process Biochem 2007, 42, 1326-1334.
12.
Lee, S. H.; Dang, D. T.; Ha, S. H.; Chang, W. J.; Koo, Y. M. Biotechnol Bioeng 2008, 99, 1-8.
13.
Li, L.; Ji, F.; Wang, J.; Li, Y.; Bao, Y. Enzyme Microb Technol 2015, 69, 46-53.
14.
Scheckermann, C.; Schlotterbeck, A.; Schmidt, M.; Wray, V.; Lang, S. Enzyme Microb Technol 1995, 17, 157-162. 24
Page 24 of 32
15.
Arcos, J. A.; Bernabé, M.; Otero, C. Enzyme Microb Technol 1998, 22, 27-35.
16.
Devi, B.; Guo, Z.; Xu, X. AICHE J 2011, 57, 1628-1637.
17.
Zhao, H.; Baker, G. A.; Song, Z.; Olubajo, O.; Zanders, L.; Campbell, S. M. J Mol Catal B: Enzym 2009, 57, 149-157.
18.
Lozano, P.; de Diego, T.; Guegan, J. P.; Vaultier, M.; Iborra, J. L. Biotechnol Bioeng 2001, 75, 563-569.
19.
Van Rantwijk, F.; Sheldon, R. A. Chemical reviews 2007, 107, 2757-2785.
20.
Husson, E.; Garcia-Matilla, V.; Humeau, C.; Chevalot, I.; Fournier, F.; Marc, I. Enzyme Microb Technol 2010, 46, 338-346.
21.
Janssen, A. E.; Vaidya, A. M.; Halling, P. J. Enzyme Microb Technol 1996, 18, 340-346.
22.
Goto, M.; Kamiya, N.; Miyata, M.; Nakashio, F. Biotechnol Prog 1994, 10, 263-268.
23.
Romero, M.; Calvo, L.; Alba, C.; Daneshfar, A. J Biotechnol 2007, 127, 269-277.
24.
Liu, Y.; Zhang, X.; Tan, H.; Yan, Y.; Hameed, B. H. Process Biochem 2010, 45, 1176-1180.
25.
Li, X.; Xu, L.; Wang, G.; Zhang, H.; Yan, Y. Process Biochem 2013, 48, 1905-1913.
26.
Lau, R. M.; Sorgedrager, M. J.; Carrea, G.; van Rantwijk, F.; Secundo, F.; Sheldon, R. A. Green Chem 2004, 6, 483-487.
27.
De Diego, T.; Lozano, P.; Gmouh, S.; Vaultier, M.; Iborra, J. L. 25
Page 25 of 32
Biomacromolecules 2005, 6, 1457-1464. 28.
Liu, Y.; Chen, D.; Yan, Y.; Peng, C.; Xu, L. Bioresource Technol 2011, 102, 10414-10418.
29.
Burney, P. R.; Pfaendtner, J. J Phys Chem B 2013, 117, 2662-2670.
30.
Jin, Z.; Liang, S.; Zhang, X.; Han, S.; Ren, C.; Lin, Y.; Zheng, S. Biotechnol Bioproc E 2013, 18, 365-374.
31.
Kaar, J. L.; Jesionowski, A. M.; Berberich, J. A.; Moulton, R.; Russell, A. J Am Chem Soc 2003, 125, 4125-4131.
32.
Byler, D. M.; Susi, H. Biopolymers 1986, 25, 469-487.
33.
Natalello, A.; Ami, D.; Brocca, S.; Lotti, M.; Doglia, S. Biochem. J 2005, 385, 511-517.
34.
Jackson, M.; Mantsch, H. H. Crit Rev Biochem Mol Biol 1995, 30, 95-120.
35.
Kong, J.; Yu, S. Acta Biochim Biophy Sin 2007, 39, 549-559.
36.
Haris, P. I.; Severcan, F. J Mol Catal B: Enzym 1999, 7, 207-221.
37.
Izutsu, K. Electrochemistry in nonaqueous solutions; John Wiley & Sons, 2009.
38.
Viswanath, D. S.; Ghosh, T. K.; Prasad, D. H.; Dutt, N. V.; Rani, K. Y. Viscosity of liquids; Springer, 2007.
FIRURE LEGENDS Figure 1. Enzymatic synthesis scheme of fructose monolaurate and dilaurate. Figure 2. Effects of volume ratios on esterification in ILs/2M2B co-solvent system. Different ratios of [BMIM][BF4](A), [BMIM][TfO](B) as co-solvent in 2M2B (v/v). 26
Page 26 of 32
Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50°C, 200 rpm (orbital stirrer), 12 h. Asterisks indicate significant differences (P < 0.05). Figure 3. Effects of rotating speed on esterification in ILs/2M2B co-solvent system. Reaction media, 20% [BMIM][BF4]/2M2B mixture (A), 60% [BMIM][TfO]/2M2B mixture (B). Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50°C, 200 rpm (orbital stirrer), 12 h. Figure 4. Time course of CALB-catalyzed production of fructose laurate in ILs/2M2B co-solvent system. Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm (orbital stirrer). Figure 5. Schematic representation of Ping Pong Bi−Bi mechanism. Figure 6. The effect of lauric acid concentration on the reaction rate. The concentration of lauric acid was varied from 20 mM to 600 mM with fixed fructose amount (0.222 mmol). Reaction conditions: 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm (orbital stirrer), 50 min. Figure 7. Operational stability of CALB in ILs/2M2B co-solvent system. Reaction media: 20% [BMIM][BF4]/2M2B mixture (A), 60% [BMIM][TfO]/2M2B mixture (B). Values in each column with different letters present significant differences (P < 0.05). Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm 27
Page 27 of 32
(orbital stirrer). After each batch, enzymes were filtered, then washed with equal volume of 2M2B for three times and finally were used for next reaction cycle with new reactants, molecular sieves and reaction media. The reaction solvent mixture and three washed 2M2B were pooled for mono-ester quantitative detection.
TABLES Table 1. Solubility of fructose in different solvents and viscosity of different solvents
28
Page 28 of 32
Solubility of fructosea (g/l, Mean% ± SD)
Visocosity Range (cP)
2M2B
16.28 ± 1.41
0.57 – 13.80c
[BMIM][PF6]
2.87 ± 1.45
173 – 450b
[BMIM][BF4]
16.74 ± 0.46
92 – 219b
20% [BMIM][BF4]/2M2B
16.50 ± 0.73
-
[BMIM][TfO]
45.31 ± 1.20
90b
60% [BMIM][TfO]/2M2B
34.52 ± 1.06
-
Solvent
a
Detected by DNS method, data represent the mean values of three experiments ± SD.
SD= Standard deviation; bRef.37; cRef.38
Table 2. Kinetic constants of the enzymatic esterification of fructose with lauric acid in 2M2B. 29
Page 29 of 32
Ping Pong Bi−Bi mechanism without substrate inhibition
Ping Pong Bi−Bi mechanism with fructose inhibition
Value
95% confidence
Value
95% confidence
Vmax
0.172
1.612×10-4
0.187
1.819×10-6
KmLA
695.277
1.526
2.184
4.957×10-5
KmF
390.441
0.735
434.392
0.008
-
-
0.261
5.940×10-6
Constant
Ki
Table 3. Kinetic constants of the enzymatic esterification of fructose with lauric acid in ILs/2M2B co-solvent system. 30
Page 30 of 32
Constant
2M2B
20% [BMIM][BF4]
60%[BMIM][TfO]
Vmax
0.187
0.194
0.230
KmLA
2.184
3.611
9.878
KmF
434.392
317.207
303.998
0.261
0.406
1.339
Ki
Table 4. Contents of secondary structure elements of CALB in ILs/2M2B co-solvent systems by FTIR analysis at Amide I region. 31
Page 31 of 32
2M2B Secondary Structure
Controla
Lauric acidb
20% [BMIM][BF4]/2M2B Fructosec
Controla
Lauric acidb
60% [BMIM][TfO]/2M2B
Fructosec
Controla
Lauric acidb
Fructosec
α-Helix (%)
15.4
9.5
10.6
10.6
7.5
10.7
9.4
8.4
8.1
β-Sheet (%)
42.1
39.1
34.7
37.3
41.9
33.6
40.9
42.8
39.6
β-Turn (%)
22.5
22.9
24.2
24.3
22.9
24.6
27.1
25.0
29.8
Random coil (%)
12.0
22.7
27.1
24.3
18.9
26.2
18.1
17.0
18.2
a
Control: binding no substrate; bCALB binding lauric acid; cCALB binding fructose.
32
Page 32 of 32
S0008-6215(15)00231-1 http://dx.doi.org/doi: 10.1016/j.carres.2015.08.009 CAR 7053
To appear in:
Carbohydrate Research
Received date: Revised date: Accepted date:
2-7-2015 17-8-2015 17-8-2015
Please cite this article as: Lu Li, Fangling Ji, Jingyun Wang, Bo Jiang, Yachen Li, Yongming Bao, Efficient Mono-acylation of Fructose by Lipase-catalyzed Esterification in Ionic Liquid cosolvents, Carbohydrate Research (2015), http://dx.doi.org/doi: 10.1016/j.carres.2015.08.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Efficient Mono-acylation of Fructose by Lipase-catalyzed Esterification in Ionic Liquid co-solvents
Lu Li1, Fangling Ji1, Jingyun Wang1, Bo Jiang1, Yachen Li2*, Yongming Bao1*
1
School of Life Science and Biotechnology, Dalian University of Technology, Dalian,
China 116024 2
Department of Occupational and Environmental Health, Dalian Medical University,
Dalian, China 116044
∗
To whom correspondence should be addressed: School of Life Science and
Biotechnology, Dalian University of Technology. NO. 2 Linggong Road, Ganjingzi District, Dalian, 116024, China. Tel: +86-411-84706344. Fax: +86-411-84706344; E-mail: [email protected], [email protected].
1
Page 1 of 32
ABBREVIATIONS ILs, ionic liquids; 2M2B, 2-methyl-2-butanol; CALB, Candida antarctica lipase B; [BMIM][TfO], 1-butyl-3-methylimidazolium trifluoromethanesulfonate; [BMIM][BF4], 1-butyl-3-methylimidazolium
tetrafluoroborate;
[BMIM][PF6],
1-butyl-3-methylimidazolium hexafluorophosphate; ATR-FTIR, attenuated total reflectance-fourier transform infrared; HPLC-ELSD, high performance liquid chromatography-evaporative light scattering detector; NMR, nuclear magnetic resonance; API-MS, atmospheric pressure ionization-mass spectrometry; ANOVA, analysis of variance.
2
Page 2 of 32
Highlight • Markedly improved fructose mono-acylation catalyzed by CALB in ILs/2M2B co-solvent system. • Selectivity of mono-acylation in ILs/2M2B co-solvent governed by the change of CALB kinetic parameters • Kinetic parameters variance determined by conformational changes of CALB binding substrates. Graphical Abstract
ABSTRACT Fructose monoesters are eco-friendly nonionic surfactants in various applications. Selective preparation of mono-acylated fructose is challenging due to the multiple hydroxyl sites available for acylation both chemically and enzymatically. Ionic liquids 3
Page 3 of 32
(ILs) have profound impacts not only on the reaction media but also on the catalytic properties of enzymes in the acylation process. In this study, utilizing an IL co-solvents system, selective synthesis of mono-acylated fructose with lauric acid catalyzed by immobilized Candida antarctica lipase B (CALB) was investigated. The imidazolium-based ILs selected as co-solvents with 2-methyl-2-butanol (2M2B), markedly improved the ratios of monolauroyl fructose in the presence of 60% [BMIM][TfO] (v/v) and 20% [BMIM][BF4] (v/v), in which the mono-acylated fructose was 85% and 78% respectively. Based on a Ping-Pong Bi−Bi model, a kinetic equation was fitted, by which the kinetic parameters revealed that the affinity between fructose and acyl-enzyme intermediate was enhanced. The inhibition effect of fructose on free enzyme was weakened in the presence of IL co-solvents. The conformation of CALB binding substrates also changed in the co-solvent system as demonstrated by Fourier transform infrared spectra. These results demonstrated that the variation of CALB kinetic characteristics was a crucial factor for the selectivity of mono-acylation in ILs/2M2B co-solvents. Keywords: Sugar ester of fatty acid; Monolauroyl fructose; Candida antarctica lipase B; Ionic liquids; Kinetics
1. INTRODUCTION Sugar esters are biodegradable and nontoxic bio-surfactants which are very useful for many commercial applications. One of the attractions is fructose laurate, due to its highest cell growth inhibitory against Streptococcus mutans among different sugar 4
Page 4 of 32
esters1. Chemical synthesis tends to result in high energy cost and environmentally unsafe conditions. As a potential alternative, enzymatic preparation has mild and eco-friendly reaction conditions, as well as the inherent selectivity maintained by enzymes during reaction process2. Lipases (EC 3.1.1.3) are considered as better catalysts for esterification, amidation and transesterification of esters in anhydrous media3. With regard to the direct esterification of fructose with lauric acid, traditional chemical reactions can result in a broad distribution of one to multi-ester species in the products. On the other hand, lipases selectively acylate the primary hydroxyl group rather than secondary groups with long chain fatty acid in anhydrous organic solvents. As the result, the products are composed of a narrow range of mono- and di-esters (Figure 1), and it is more convenient for the separation of products. As reported, Cao’s4 and Šabeder’s co-workers5 obtained relatively higher conversions of fructose esters with lipases-mediated acylation, achieving acylation rates of 86% and 82%, respectively. However, these products were mixtures of esters in different esterification degrees, with a higher proportion of di-ester in the products. This phenomenon is considered inevitable in the direct esterification process. Ionic liquids (ILs) have been recognized as an alternative to traditional organic solvents for the synthesis of sugar esters, not only because they offer many advantages as solvents, such as their low vapor pressure, high thermal stability and good solubilization of a wide range of carbohydrates6, but also as reaction media that can enhance the reactivity, selectivity, and stability of enzymes7-10. Several groups have 5
Page 5 of 32
reported that anhydrous ILs containing [BF4]-, [PF6]- and [TfO]- anions can be used for biotransformation of sugars. In these ILs, desired monoacylated compounds were synthesized and/or high ester yields were achieved10-12. Co-solvent system of ILs/organic solvent is a balanced approach between mass transfer limitation of pure ILs and selectivity of lipase. Our previous work revealed that in pure 2M2B, CALB displayed selectivity towards low degree of esterification products (mono-acylated fructose) and was an unfavorable solvent for di-ester synthesis13. Therefore, with the goal of further enhancing the selectivity and obtaining higher proportion of mono-acylated fructose laurate, our study utilized three imidazolium-based ionic liquids ([BMIM][BF4], [BMIM][PF6] and [BMIM][TfO]) to constitute an ILs/2M2B co-solvent system as the reaction media in the acylation process of fructose. To probe the key factors that affect the activity and selectivity in mono-ester synthesis, we first experimentally surveyed the productivity of fructose laurate and the selectivity of mono-acylation in different volume ratios of the ILs/2M2B co-solvent system. Then the optimal reaction media were further studied for the characteristics that are particularly related to the solubility of substrate, mass transfer limitation, and the kinetics and structure variation of CALB. 2. RESULTS AND DISCUSSION 2.1 Effect of volume ratios in the ILs/2M2B co-solvent system For the enzymatic direct esterification of fructose with lauric acid, high conversion rate was obtained by the approach of providing excessive insoluble fructose in a small 6
Page 6 of 32
amount of organic solvent-adjuvant5,4. Based on this protocol, we set about conducting the reaction process in co-solvents of ILs/2M2B. In our case, we expected an increase of mono-acylated product in such reaction media. Having developed an optimized co-solvent system, the synthesis process was launched at different volume percentages of ILs in 2M2B. For comparison, the synthetic reaction was also assayed in pure 2M2B and pure ILs. HPLC and mass spectrometry revealed that, within the reaction time of 12 h in either pure 2M2B or co-solvent system, only mono-acylated fructose laurate and di-acylated fructose laurate was detectable in the products. The structures of products were further confirmed by
13
C NMR spectroscopy, a method employed in previous
reports14,15. We found that mono-ester was a mixture composed of 1-α-fructofuranose laurate, 1-β-fructopyranose laurate, 6-lauroyl-α-D-fructofuranose and 6-lauroyl-β-Dfructofuranose; the di-esters were composed of 1, 6-dilauroyl-α-D-fructofuranose laurate and 1, 6-dilauroyl-β-D-fructofuranose laurate. As shown in Figure 2, the addition of ionic liquids resulted in fluctuation in the yields of products and especially the increased selectivity of mono-acylation. The proportion of mono-acylated product had a clear improvement (P < 0.05) in the [BMIM][BF4]/2M2B and [BMIM][TfO]/2M2B co-solvent system. As shown in Figure 2-A, we found that with the increase of [BMIM][BF4] component, yield and selectivity of mono-acylation content were gradually increased, beyond which further increase in ILs content resulted in a substantial decline. At the point of 20% [BMIM][BF4], we detected the highest yield of mono-ester (27.05%) and significantly higher selectivity 7
Page 7 of 32
of mono-acylation (78.30%) than pure 2M2B (72.28%) and pure IL (33.57%). After that point, yield of mono-ester decreased and the overall conversion rate was low. Meanwhile, the increase of di-ester led to poor selectivity for mono-ester synthesis. Similarly, enhanced yield of mono-ester was observed with the increase of [BMIM][TfO] ratios in co-solvent system (Figure 2-B). The selectivity of mono-acylation achieved the highest level in ratio of 60% [BMIM][TfO]/2M2B (83.33%). However, unlike the case in [BMIM][BF4]/2M2B, pure [BMIM][TfO] as solvent did not bring about a significant decline in conversion rate and selectivity. Compared to the case in pure 2M2B, it still maintained at a relatively high level. [BMIM][PF6] was also investigated as co-solvent in the reaction, where obvious decrease of both conversion and mono-acylation was detected (data not shown). Compared to other ILs, almost no activity of CALB presented in pure [BMIM][PF6]; the same phenomenon was also observed by Devi et al. in CALB-catalyzed production of biodiesel in this ionic liquid16. Since [BMIM][PF6] has the highest viscosity and poor solubility of fructose (Table 1), one possible explanation is that enzyme reaction in [BMIM][PF6] can be limited by the dissolution rate and mass transfer rate of substrate. It is worth noting that higher conversion and selectivity of mono-acylation was achieved in high ratio of [BMIM][TfO]/2M2B, but not in [BMIM][BF4]/2M2B. In addition, 20% [BMIM][BF4]/2M2B co-solvent system resulted in significantly higher selectivity of mono-acylation than 2M2B did, even though they have almost similar fructose solubility and viscosity. Since the three ionic liquids used are notably different in their properties, such as polarity, hydrophobicity, anion nucleophilicity, 8
Page 8 of 32
hydrogen-bond basicity and viscosity17, and also that they possess distinct IL ratios and solubility of fructose (Table 1), we reasoned that not only the solubility of the substrate, especially fructose, but also limitation of mass transfer and functional interaction between enzyme and ILs, or the combined effects of above could be accountable for these phenomena. To investigate the key factors in dominating mono-acylated product yields and selectivity of mono-acylation in different co-solvents system, a series of experiments addressing these aspects were designed and carried out below. 2.2 Effect of rotation To understand the effect of mass transfer in a co-solvent system, reactions were proceeded under different rotating speeds in a 20% [BMIM][BF4]/2M2B and 60% [BMIM][TfO]/2M2B co-solvent system (Figure 3). As rotating speeds shifted, reaction rate gradually improved in both solvents. However, when speed exceeded 200 rpm, the growth rate for yields of both products slowed down. From Figure 3-A, yields of mono-ester and di-ester in 20% [BMIM][BF4]/2M2B mixture increased at the same rate, the selectivity of mono-acylation was not significantly affected by mass transfer in this co-solvents system. While in the [BMIM][TfO] group (Figure 3-B), poor selectivity of mono-acylation was found at low speeds (below 200 rpm). This feature was significantly improved with a further increase in rotating speed, but it was not enhanced above 200 rpm. Many previous reports indicated that viscosity of ILs may have an influence on reaction rate and enzyme activity18,19. As mentioned above, selectivity of mono-acylation was also changed in the [BMIM][TfO] group. This phenomenon is 9
Page 9 of 32
likely to be explained by the dissolution rate of fructose. Mass transfer rate is closely related to dissolution and diffusion rate of substrates. Since mono-ester was far more soluble than fructose in our co-solvent system, the amount of soluble fructose is particularly important when enzyme chose its acyl acceptor. At a low speed, the more dissolution rate was limited, the less soluble fructose interacted with enzyme. Compared with 20% [BMIM][BF4]/2M2B mixture, this effect was more prominent in medium with a higher ratio of [BMIM][TfO], which has higher solubility of fructose and viscosity (Table 1). As rotating speed increased, this limited effect gradually weakens. Selectivity of mono-acylation in 60% [BMIM][TfO]/2M2B was independent of the rotating rate for values of the rate above 200 rpm. Taken together, we reasoned that selectivity of mono-acylation in a two optimal co-solvent system was not primarily determined by mass transfer limitation set by viscosity in the adopted speed of our experiment (200 rpm). 2.3 Reaction time course In comparison of specific reaction process with each reaction mixtures, Figure 4 depicts the time course of the conversion and selectivity of mono-acylation catalyzed by CALB in 20% [BMIM][BF4]/2M2B mixture, 60% [BMIM][TfO]/2M2B mixture and pure 2M2B at 50°C. As shown, for all of the co-solvent systems tested, acylated products of the reaction appeared in a sequential way, i.e., the formation of di-ester is associated with the depletion of substrates and the formation of a substantial amount of mono-acylated products. At the initial stage of the reaction, only monoester was produced. The formation of monoester in all of the co-solvent systems proceeds rapidly 10
Page 10 of 32
within 50 min, far more than that in pure 2M2B. The amount of monoester in 60% [BMIM][TfO]/2M2B mixture was increased to 0.045 mmol, and 0.021 mmol in 20% [BMIM][BF4]/2M2B mixture, whereas 0.009 mmol in pure 2M2B. After 50 min, the velocity of mono-ester formation in all of the co-solvent system tested slowed down. Di-esters for all of the tested solvents were detectable at that point and then gradually increased in a nearly identical formation velocity. Based on this, we concluded that, within 12 h, it is the synthesis of mono-ester that leads to such selectivity during the acylation process. In other words, within 12 h the reaction step of fructose to mono-ester synthesis determines the selectivity of mono-acylation but not the synthesis step from monoester to di-ester. Consequently, as no di-ester was detected during the initial reaction period, the kinetics study of monoester formation can be greatly simplified. 2.4 Kinetic constants Direct esterification of lauric acid with fructose catalyzed by a lipase is often modeled using the Ping Pong Bi−Bi kinetic mechanism20-22. For this mechanism, a general scheme for this type of reactions was established (Figure 5). The initial rates V is expressed as in the following equation (3)， v
V m a [x F ] [ L A ] K m F[ L A]
Km
L
[A F]
(3) [ F ] [L A]
Where v is the initial reaction rate; Vmax is the maximum reaction rate; [LA] and [F] are the concentrations of lauric acid and fructose, respectively; KmLA and KmF are kinetic constants for lauric acid and fructose, respectively. 11
Page 11 of 32
If there are inhibitions from any of the substrates, it is necessary to include the corresponding inhibition constants in the equation. From Figure 6, it showed that lauric acid does not negatively affect the reaction rate. Thus, a Ping-Pong Bi−Bi kinetic model with competitive inhibition by fructose can be established as shown in the following equation (4)， V m a x [ F ][ L A ]
v
K m F [ L A ] K m L A [ F ](1
[F ]
(4) ) [ F ][ L A ]
Ki
Where Ki is the inhibition coefficient of fructose. Considering that the fructose dissolution rate could complicate the analysis of enzyme kinetics, fructose concentration was fixed at the saturation point in our tested reaction media. However, limited solubility of fructose in 2M2B provided us narrow intervals to test the effect of fructose concentrations on the reaction rate. Much lower concentration of fructose would not be detectable in the initial period of reactions or could lead to larger errors via HPLC. Therefore, we intended to fit our available data with both of the equations. Kinetic constants-Vmax, KmLA, KmF and Ki were estimated from the initial rate data (Figure 6) coupled with equation (3) and equation (4), and refined by using the Polymath 5.1 software package. By knowing the experimental initial rates with respect to initial concentrations, all of the rate constants were calculated through nonlinear regression to the rate equation. The best fit values of the two equations in pure 2M2B obtained are presented as examples in Table 2. The fitting for both of the equations was excellent, with R2 and adjusted R2 being above 0.98. This indicates that the theoretical 12
Page 12 of 32
model and the actual value are in agreement. As shown in Table 3, in the column without fructose inhibition group, KmF was less than KmLA, indicating that the affinity of the enzyme towards fructose was much greater than that to lauric acid. It is reasonable to assume that such affinity can induce fructose to form a dead-end complex by binding to the enzyme. On the other hand, in the inhibition group the higher affinity of enzyme towards lauric acid could fully describe this mechanism (Figure 5) with acyl donor as the first binding substrate23. Thus, we believe that this fitting model may exhibit more rational kinetic constants. The final kinetic equation with inhibition constant was used to estimate the initial reaction rates. By using kinetic constant values (Table 3), the preference of the CALB for different substrates in these co-solvent systems can be compared. With the addition of ILs, obtained values of kinetic constants exhibited marked differences. As expected, the highest Vmax (0.230 mmol·min-1·g-1) was found in 60% [BMIM][TfO]/2M2B co-solvent system, followed by 20% [BMIM][BF4]/2M2B (0.194 mmol·min-1·g-1), with the lowest one (0.187 mmol·min-1·g-1) being in 2M2B. As shown in Table 3, the increase of KmLA in three solvents followed the same order as Vmax, suggesting that the affinity between enzyme and lauric acid declined with the addition of ILs. This effect can be seen in Figure 6, where the reaction rate was faster in 2M2B than in co-solvent systems at low concentration of lauric acid. With the increase of lauric acid concentration, the relatively lower fructose concentration started to determine the whole reaction rate when fructose concentrations were fixed in these reaction systems. Most notably, the acyl-CALB (Figure 5) affinity herein for fructose was enhanced in 13
Page 13 of 32
the co-solvent systems (reduced KmF). Moreover, the inhibition effect of fructose in both of the co-solvent systems was weakened compared with pure 2M2B (increased Ki). Collectively, these varied affinities contributed to the higher reaction rate in the co-solvent systems at high lauric acid concentration. 2.5 The secondary structure of immobilized CALB by FTIR spectrometry The interactions of enzyme-solvent play an important role in enzyme structure and function. It is reasonable to speculate that the conformational change in secondary structure caused by protein-solvent interactions is a probable cause for the variation of lipase activity, or even the selectivity. It has been reported that organic solvent and ILs can drive conformational changes and affect enzyme stability and activity24-28. To monitor secondary conformational variations of enzyme binding substrates in the different ILs/2M2B co-solvent system, we calculated the composition of the lipase secondary structure with the aid of ATR-FTIR spectroscopy. First, in the control group (Table 4), no substrate was found to be bound to CALB in the co-solvent system. There are obvious changes in the composition of the secondary structure in different solvents. The addition of ILs changed the whole conformation of lipase. We believe that these changes would contribute to the variation in activity and selectivity in our experiments. When CALB binding to lauric acid and fructose, respectively, as revealed by the kinetics studies, second structure changed in co-solvent systems. Our results provided direct evidence for the changes in substrate affinity and inhibition. With one exception, however, there was no obvious changes in fructose binding groups between pure 2M2B and 20% [BMIM][BF4]/2M2B. When the control group was compared with the 14
Page 14 of 32
binding groups in corresponding co-solvent system, notable conformation changes were found when CALB binds to fructose in pure 2M2B. The proportions of both α-helix and β-sheet reduced from 15.35% to 10.63% and from 42.06% to 34.68%, respectively, with the other structural contents increased accordingly. In contrast, in 20% [BMIM][BF4]/2M2B, the proportions of secondary structures were nearly identical. This more flexible enzyme structure observed in pure 2M2B that seem to increase the inhibition effect of fructose on free enzyme. By contrast, the flexibility of CALB was limited in the co-solvent system. Based on the study of molecular dynamic simulations by Patrick R. Burney and Jim Pfaendtner29, it was also found that ILs significantly dampen protein dynamics of lipase, and consequently, it is difficult for ILs to drive enzyme deviating from its initial structure compared with organic solvent. Therefore, ILs component does shape the enzyme in response to the function of lipase. 2.6 Operational stability The operational stability or reusability of CALB in ILs/2M2B co-solvents system was achieved by 8 batches esterification of fructose with lauric acid at the same condition (Figure 7). The relative activity of CALB was determined by mono-ester yield of each batch to the corresponding value of the first batch, taking the relative activity of the first batch as 100%30. As shown in Figure 7, although the activity of CALB in both co-solvents system were reduced with increased batches, the relative activities were remained ca. 80% in 20% [BMIM][BF4]/2M2B after 6 batches and ca. 60% in 60% [BMIM][TfO]/2M2B after 7 batches, respectively. It indicated that 2M2B exerted certain protective effects 15
Page 15 of 32
on enzyme activity against inactivation by ILs. The decrease of enzyme relative activity (P < 0.05) from third batch might be explained by the inactivation of enzyme, rather than effects by substrates or products. In the next 5 batches, differences of enzyme relative activity were not significant (P > 0.05) among these batches, and it indicated that CALB could maintain good stability in both co-solvents system. Moreover, the enzyme relative activities were slightly increased, and Joel L. Kaar et al.31 speculated that ionic liquids could swell the acrylic resin (immobilized carrier) causing improved mass transfer. 3. CONCLUSION In the present work, the catalytic behavior of CALB was investigated in the acylation of fructose with lauric acid in the ILs/2M2B co-solvent system. For improving the mono-ester synthesis, three ILs which were reported as favorable solvents in mono-acylation and possessing different dissolving capacities of fructose were employed as co-solvents with 2M2B. As expected, higher selectivity of mono-acylation was obtained by using the co-solvents composed of [BMIM][TfO] and [BMIM][BF4], respectively. In an attempt to elucidate the underlying mechanisms, we studied the probable causes from the addition of ILs for the changes in solubility of substrates, mass transfer, kinetics features and structural changes of CALB. It is evident from the time-course study that mono-ester synthesis determines the selectivity of mono-acylation in the entire reaction process. Mass transfer limitation caused by ILs viscosity showed that such enhanced selectivity was independent of rotating speed in our operational conditions. With A Ping-Pong Bi−Bi kinetic model and with dead-end 16
Page 16 of 32
complex inhibition of fructose fitting our experimental data, kinetic feature of CALB was affected by co-solvent system, in which the affinities and inhibition of substrates were changed. These observations were also supported by the conformational studies via FTIR when CALB binds different substrates. In conclusion, the variation of enzyme kinetic characteristics of CALB was found to be a crucial parameter in governing mono-ester productivity and selectivity of mono-acylation in ILs/2M2B co-solvents. Furthermore, our data showed that lipase catalysis can be an ideal and favorable approach for mono-acylation of polyhydroxy substrates, such as fructose. 4. EXPERIMENTAL 4.1 Materials Novozym 435 was a gift from Novozymes (China) Investment Co. Ltd (Beijing, China). Lewatit VP OC 1600 resin was also a gift from LANXESS Chemical (China) Co.Ltd. Commercial ionic liquids, 1-butyl-3-methylimidazolium trifluoromethanesulfonate
([BMIM][TfO]),
1-butyl-3-methylimidazolium
tetrafluoroborate
([BMIM][BF4]) and 1-butyl-3-methylimidazolium hexafluorophosphate ([BMIM][PF6]) with purity of 99% were purchased from Shanghai Cheng Jie Chemical Co. Ltd. (Shanghai, China). D-fructose (≥99%) (Amresco, USA), lauric acid (≥98.0%) (Sinopharm, Shanghai, China), 2-methyl 2-butanol (2M2B, 98%) (Guang Fu, Tianjin, China). All other chemicals were also from commercial sources and of the highest purity available.
17
Page 17 of 32
4.2 General procedure for lipase-catalyzed esterification in ILs/2M2B co-solvent system All reactions were carried out in 1.5 ml clear reaction vial sealed with a screw cap. Equimolar fructose and lauric acid (0.222 mmol) and 100 mg of molecular sieve (3 Å) were added into 0.6 ml mixed solvents of [BMIM][TfO]/2M2B, [BMIM][BF4]/2M2B and [BMIM][PF6]/2M2B, respectively. These mixtures were then placed in an air-bath shaker for 12 h at 200 rpm and 50 °C to dissolve certain amount of substrates. The reactions were initiated by adding 25 mg CALB into each mixture. Periodically, these samples were extracted by the addition of 0.6 ml methanol and 15 s of vortexing shake. 50 μl of liquid phase were withdrawn and analyzed via high performance liquid chromatography (HPLC). 2M2B and ILs were dehydrated with molecular sieves (4 Å) over 48 h before reactions. All measurements were conducted in triplicate. 4.3 Yields of products and selectivity of mono-acylation The yield of products was calculated by the following equation (1): Y ie ld (% )=
M
P
M
F
100
(1)
Where MP is the generated amount of products (mole), MF is the initial amount of fructose (mole). The selectivity of mono-acylation was defined by the following equation (2): M
S e le c tiv ity ( % ) = M
ME
100
ME
M
(2)
DE
Where MME and MDE are the generated amount of mono-ester and di-ester (mole), respectively. The generated mono-ester and di-ester were measured by with a HPLC 18
Page 18 of 32
system (Waters 600 controller) equipped with an evaporative light scattering detector (ELSD) and a 150 mm × 4.6 mm (5 μm) Hypersil ODS2-C18 column. 50 μl samples from the reaction mixture were withdrawn at final reaction times, then filtered and injected into a HPLC. The temperature of drift tube in ELSD was set at 45°C at the pressure of 20 psi (nitrogen gas). The mobile phase system was acetonitrile/water (v/v). The gradient eluted program was used as following profile: 70% acetonitrile at a flow rate of 0.9 ml/min from 0-4 min, followed by 100% acetonitrile from 4-20 min at a flow rate of 1 ml/min. The retention time for lauroyl-D-fructose was 3.4 min, and for 1,6-diacyl-D-fructofuranose was 14 min. Mono-ester and di-ester were isolated and collected by flash column chromatography (elution with petroleum ether and ethyl acetate sequentially). The purity of collected fractions was determined by HPLC. 4.4 Structural characterization of fructose lauric acid esters 13
C NMR spectra were obtained from Varian INOVA 400 (101 MHz for
13
C)
spectrometer in CD3OD. Atmospheric pressure ionization (API) mass spectra were recorded on HP1100LC/MSD spectrometer. Shifts of fructose carbons of four isomers of fructose mono-esters were recorded as follows: (a) 6-lauroyl-β-D-fructofuranose, 13
C NMR (101 MHz, CD3OD): δ = 103.39 (C-2), 80.09 (C-5), 77.43 (C-3), 76.96 (C-4),
66.74 (C-6), 64.34 (C-1). (b) 1-lauroyl-β-D-fructopyranose,
13
C NMR (101MHz,
CD3OD): δ = 98.16 (C-2), 71.39 (C-4), 70.89 (C-5), 69.71 (C-3), 66.58 (C-1), 64.66 (C-6). (c) 6-lauroyl-α-D-fructofuranose,
13
C NMR (101MHz, CD3OD): δ = 106.00
(C-2), 84.06 (C-3), 80.42 (C-5), 79.00 (C-4), 65.42 (C-6), 64.92 (C-1). (d) 1 -lauroyl-β-D-fructofuranose,
13
C NMR (101MHz, CD3OD): δ = 101.44 (C-2), 83.31 19
Page 19 of 32
(C-5), 78.43 (C-3), 76.33 (C-4), 65.56 (C-1), 63.98 (C-6). Shifts of groups of four lauric acid isomers: δ = 175.35, 175.29, 175.26, 174.89 (C=O, 4 isomers), 25.70-34.89 (CH2), 14.45 (CH3). API-MS [M + Na] + m/z: 385. Shifts
of
the
two
isomers
1,6-dilauroyl-β-D-fructofuranose, 80.15
(C-5),
77.96
(C-3),
1,6-dilauroyl-α-D-fructofuranose,
of
13
fructose
di-esters
are
as
follows:
(a)
C NMR (101MHz, CD3OD) δ = 101.67 (C-2),
77.06 13
(C-4),
66.72
(C-1),
65.30
(C-6).
(b)
C NMR(101MHz, CD3OD) δ = 104.90 (C-2),
83.73 (C-3), 80.54 (C-5), 78.55 (C-4), 65.92 (C-1), 65.30 (C-6). Shifts of acyl skeleton: δ = 175.26, 174.77 (C=O, β-anomer), 175.14, 175.06 (C=O, α-anomer), 23.73-34.92 (CH2), 14.48 (CH3). API-MS [M + Na] + m/z: 567. 4.5 Measurement of fructose solubility Fructose (0.222 mmol) was added to glass vials containing 0.6 ml 2M2B, ILs or ILs/2M2B, respectively. The suspension was stirred at 200 rpm 50°C After 12 h, the supernatant was removed after centrifugation of the suspension. Then 1 ml deionized water was added to the vial. The residual fructose was dissolved to determine sugar content. The fructose concentration was confirmed by DNS method with a fructose standard. 4.6 Determination of kinetic constants in co-solvent system Initial reaction rates, expressed as micromole produced mono-ester per minute and per gram of enzyme, were determined by the amount of lauroyl-D-fructose generated within 50 min via HPLC. To avoid fructose dissolution rate complicating the analysis of enzyme kinetics, different solvents with saturated fructose prior to the reaction were 20
Page 20 of 32
incubated and the concentration of fructose were kept constant. The effect of lauric acid concentration on the reaction rate was investigated by esterifying fixed initial quantities of fructose with different initial concentrations of lauric acid (20 mM to 600 mM). In all experiments it was used an input amount of 25 mg immobilized CALB. In order to get the parameters of the kinetic model, initial rates were fitted to the proposed reaction rate equation by non-linear regression analysis with the computer program polymath 5.1. 4.7 Conformation analysis of CALB by Fourier Transform Infrared (FTIR) spectrometry Aided by attenuated total reflectance (ATR) - FTIR spectroscopy, the secondary structure of immobilized CALB can be directly measured, which caused by the characteristic peptide group vibrations, amide I band (1700-1600 cm-1) that arises from the C=O stretching vibrations32. Prior to the essay, 50 mg immobilized CALB were incubated for 30 min with 46.30 mM dissolved fructose or lauric acid in 2M2B, 60% [BMIM][TfO]/2M2B mixture and 20% [BMIM][BF4]/2M2B mixture, respectively. Then experiments were performed without substrates under same conditions as control groups. Besides, for subtracting the absorption of immobilized enzyme carrier (Lewatit VP OC 1600 resin), the carrier was strictly treated and collected under same conditions. FTIR absorption spectra from 4000 to 640 cm-1 were collected using a Nicolet 6700 FTIR spectrometer equipped with a nitrogen-cooled, mercury-cadmium-tellurium detector at a resolution of 4 cm-1. Spectroscopic data were recorded and analyzed with 21
Page 21 of 32
Thermo Scientific OMNICTM software (OMNIC 8.2.387, Thermo Fisher Scientific Inc.). The spectra were smoothed with a binomial 5 points smoothing function, then second derivative spectra were obtained following the Savitsky–Golay method (3rd grade polynomial)33. The center position for each amide I sub-peak from 1700 to 1600 cm−1 was determined on the basis of previous assignments34-36 and confirmed by using second-derivative peak analysis. Fourier-self-deconvolution of the amide I band was performed by using an enhancement factor K=2.4 and half-bandwidth=20 cm−1. Curve-fitting process was accomplished with Gaussian function by employing OMNIC software. 4.8 Statistical analysis Statistical analyses were performed using computer program OriginPro (version 8.5). A one-way analysis of variance (ANOVA) with Tukey test was applied to determine significant differences among means. The significance value was set at p < 0.05.
ACKNOWLEDGEMENT This work was supported by the Major Basic Research Program of China (Grant No. 2009CB724700). We thank Novozymes (Beijing, China) for kindly supplying immobilized lipase and LANXESS Chemical (Shanghai, China) for their generous 22
Page 22 of 32
donation of macroporous resin. We thank Limei Luo (Analysis Center, School of Chemical Engineering) and Kun Jin (State Key Laboratory of Fine Chemicals) for their assistances in FTIR spectroscopy and NMR measurement.
REFERENCES 1.
Watanabe, T.; Katayama, S.; Matsubara, M.; Honda, Y.; Kuwahara, M. Curr Microbiol 2000, 41, 210-213.
2.
Pyo, S. H., and D. G. Hayes. . In Biobased Surfactants and Detergents: 23
Page 23 of 32
Synthesis, Properties, and Applications; D.G. Hayes, D. K., D.K.Y. Solaiman, and R.D. Ashby Ed.; AOCS Champaign, IL, 2009; pp. 323-350. 3.
Gandhi, N. N.; Patil, N. S.; Sawant, S. B.; Joshi, J. B.; Wangikar, P. P.; Mukesh, D. Catalysis Reviews 2000, 42, 439-480.
4.
Cao, L.; Fischer, A.; Bornscheuer, U. T.; Schmid, R. D. Biocatal Biotransform 1996, 14, 269-283.
5.
Šabeder, S.; Habulin, M.; Knez, Ž. J Food Eng 2006, 77, 880-886.
6.
Zakrzewska, M. E.; Bogel-Łukasik, E.; Bogel-Łukasik, R. Energy Fuels 2010, 24, 737-745.
7.
Ganske, F.; Bornscheuer, U. T. Org Lett 2005, 7, 3097-3098.
8.
Kim, M.-J.; Choi, M. Y.; Lee, J. K.; Ahn, Y. J Mol Catal B: Enzym 2003, 26, 115-118.
9.
Klähn, M.; Lim, G. S.; Seduraman, A.; Wu, P. PCCP 2011, 13, 1649-1662.
10.
Park, S.; Kazlauskas, R. J. J Org Chem 2001, 66, 8395-8401.
11.
Katsoura, M.; Polydera, A.; Katapodis, P.; Kolisis, F.; Stamatis, H. Process Biochem 2007, 42, 1326-1334.
12.
Lee, S. H.; Dang, D. T.; Ha, S. H.; Chang, W. J.; Koo, Y. M. Biotechnol Bioeng 2008, 99, 1-8.
13.
Li, L.; Ji, F.; Wang, J.; Li, Y.; Bao, Y. Enzyme Microb Technol 2015, 69, 46-53.
14.
Scheckermann, C.; Schlotterbeck, A.; Schmidt, M.; Wray, V.; Lang, S. Enzyme Microb Technol 1995, 17, 157-162. 24
Page 24 of 32
15.
Arcos, J. A.; Bernabé, M.; Otero, C. Enzyme Microb Technol 1998, 22, 27-35.
16.
Devi, B.; Guo, Z.; Xu, X. AICHE J 2011, 57, 1628-1637.
17.
Zhao, H.; Baker, G. A.; Song, Z.; Olubajo, O.; Zanders, L.; Campbell, S. M. J Mol Catal B: Enzym 2009, 57, 149-157.
18.
Lozano, P.; de Diego, T.; Guegan, J. P.; Vaultier, M.; Iborra, J. L. Biotechnol Bioeng 2001, 75, 563-569.
19.
Van Rantwijk, F.; Sheldon, R. A. Chemical reviews 2007, 107, 2757-2785.
20.
Husson, E.; Garcia-Matilla, V.; Humeau, C.; Chevalot, I.; Fournier, F.; Marc, I. Enzyme Microb Technol 2010, 46, 338-346.
21.
Janssen, A. E.; Vaidya, A. M.; Halling, P. J. Enzyme Microb Technol 1996, 18, 340-346.
22.
Goto, M.; Kamiya, N.; Miyata, M.; Nakashio, F. Biotechnol Prog 1994, 10, 263-268.
23.
Romero, M.; Calvo, L.; Alba, C.; Daneshfar, A. J Biotechnol 2007, 127, 269-277.
24.
Liu, Y.; Zhang, X.; Tan, H.; Yan, Y.; Hameed, B. H. Process Biochem 2010, 45, 1176-1180.
25.
Li, X.; Xu, L.; Wang, G.; Zhang, H.; Yan, Y. Process Biochem 2013, 48, 1905-1913.
26.
Lau, R. M.; Sorgedrager, M. J.; Carrea, G.; van Rantwijk, F.; Secundo, F.; Sheldon, R. A. Green Chem 2004, 6, 483-487.
27.
De Diego, T.; Lozano, P.; Gmouh, S.; Vaultier, M.; Iborra, J. L. 25
Page 25 of 32
Biomacromolecules 2005, 6, 1457-1464. 28.
Liu, Y.; Chen, D.; Yan, Y.; Peng, C.; Xu, L. Bioresource Technol 2011, 102, 10414-10418.
29.
Burney, P. R.; Pfaendtner, J. J Phys Chem B 2013, 117, 2662-2670.
30.
Jin, Z.; Liang, S.; Zhang, X.; Han, S.; Ren, C.; Lin, Y.; Zheng, S. Biotechnol Bioproc E 2013, 18, 365-374.
31.
Kaar, J. L.; Jesionowski, A. M.; Berberich, J. A.; Moulton, R.; Russell, A. J Am Chem Soc 2003, 125, 4125-4131.
32.
Byler, D. M.; Susi, H. Biopolymers 1986, 25, 469-487.
33.
Natalello, A.; Ami, D.; Brocca, S.; Lotti, M.; Doglia, S. Biochem. J 2005, 385, 511-517.
34.
Jackson, M.; Mantsch, H. H. Crit Rev Biochem Mol Biol 1995, 30, 95-120.
35.
Kong, J.; Yu, S. Acta Biochim Biophy Sin 2007, 39, 549-559.
36.
Haris, P. I.; Severcan, F. J Mol Catal B: Enzym 1999, 7, 207-221.
37.
Izutsu, K. Electrochemistry in nonaqueous solutions; John Wiley & Sons, 2009.
38.
Viswanath, D. S.; Ghosh, T. K.; Prasad, D. H.; Dutt, N. V.; Rani, K. Y. Viscosity of liquids; Springer, 2007.
FIRURE LEGENDS Figure 1. Enzymatic synthesis scheme of fructose monolaurate and dilaurate. Figure 2. Effects of volume ratios on esterification in ILs/2M2B co-solvent system. Different ratios of [BMIM][BF4](A), [BMIM][TfO](B) as co-solvent in 2M2B (v/v). 26
Page 26 of 32
Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50°C, 200 rpm (orbital stirrer), 12 h. Asterisks indicate significant differences (P < 0.05). Figure 3. Effects of rotating speed on esterification in ILs/2M2B co-solvent system. Reaction media, 20% [BMIM][BF4]/2M2B mixture (A), 60% [BMIM][TfO]/2M2B mixture (B). Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50°C, 200 rpm (orbital stirrer), 12 h. Figure 4. Time course of CALB-catalyzed production of fructose laurate in ILs/2M2B co-solvent system. Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm (orbital stirrer). Figure 5. Schematic representation of Ping Pong Bi−Bi mechanism. Figure 6. The effect of lauric acid concentration on the reaction rate. The concentration of lauric acid was varied from 20 mM to 600 mM with fixed fructose amount (0.222 mmol). Reaction conditions: 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm (orbital stirrer), 50 min. Figure 7. Operational stability of CALB in ILs/2M2B co-solvent system. Reaction media: 20% [BMIM][BF4]/2M2B mixture (A), 60% [BMIM][TfO]/2M2B mixture (B). Values in each column with different letters present significant differences (P < 0.05). Reaction conditions: 0.222 mmol fructose, 0.222 mmol lauric acid, 0.6 ml reaction media, 25 mg immobilized CALB, 100 mg molecular sieve (3 Å), 50 °C, 200 rpm 27
Page 27 of 32
(orbital stirrer). After each batch, enzymes were filtered, then washed with equal volume of 2M2B for three times and finally were used for next reaction cycle with new reactants, molecular sieves and reaction media. The reaction solvent mixture and three washed 2M2B were pooled for mono-ester quantitative detection.
TABLES Table 1. Solubility of fructose in different solvents and viscosity of different solvents
28
Page 28 of 32
Solubility of fructosea (g/l, Mean% ± SD)
Visocosity Range (cP)
2M2B
16.28 ± 1.41
0.57 – 13.80c
[BMIM][PF6]
2.87 ± 1.45
173 – 450b
[BMIM][BF4]
16.74 ± 0.46
92 – 219b
20% [BMIM][BF4]/2M2B
16.50 ± 0.73
-
[BMIM][TfO]
45.31 ± 1.20
90b
60% [BMIM][TfO]/2M2B
34.52 ± 1.06
-
Solvent
a
Detected by DNS method, data represent the mean values of three experiments ± SD.
SD= Standard deviation; bRef.37; cRef.38
Table 2. Kinetic constants of the enzymatic esterification of fructose with lauric acid in 2M2B. 29
Page 29 of 32
Ping Pong Bi−Bi mechanism without substrate inhibition
Ping Pong Bi−Bi mechanism with fructose inhibition
Value
95% confidence
Value
95% confidence
Vmax
0.172
1.612×10-4
0.187
1.819×10-6
KmLA
695.277
1.526
2.184
4.957×10-5
KmF
390.441
0.735
434.392
0.008
-
-
0.261
5.940×10-6
Constant
Ki
Table 3. Kinetic constants of the enzymatic esterification of fructose with lauric acid in ILs/2M2B co-solvent system. 30
Page 30 of 32
Constant
2M2B
20% [BMIM][BF4]
60%[BMIM][TfO]
Vmax
0.187
0.194
0.230
KmLA
2.184
3.611
9.878
KmF
434.392
317.207
303.998
0.261
0.406
1.339
Ki
Table 4. Contents of secondary structure elements of CALB in ILs/2M2B co-solvent systems by FTIR analysis at Amide I region. 31
Page 31 of 32
2M2B Secondary Structure
Controla
Lauric acidb
20% [BMIM][BF4]/2M2B Fructosec
Controla
Lauric acidb
60% [BMIM][TfO]/2M2B
Fructosec
Controla
Lauric acidb
Fructosec
α-Helix (%)
15.4
9.5
10.6
10.6
7.5
10.7
9.4
8.4
8.1
β-Sheet (%)
42.1
39.1
34.7
37.3
41.9
33.6
40.9
42.8
39.6
β-Turn (%)
22.5
22.9
24.2
24.3
22.9
24.6
27.1
25.0
29.8
Random coil (%)
12.0
22.7
27.1
24.3
18.9
26.2
18.1
17.0
18.2
a
Control: binding no substrate; bCALB binding lauric acid; cCALB binding fructose.
32
Page 32 of 32