Evaluation of biocompatible ionic liquids for their application in

in phytosterols bioconversion by Mycobacterium sp. resting cells. Jun-Jie Yuan, Yi-Xin Guan*, Shan-Jing Yao. College of Chemical and Biological Engine...
0 downloads 0 Views 1MB Size
Research Article pubs.acs.org/journal/ascecg

Cite This: ACS Sustainable Chem. Eng. 2017, 5, 10702-10709

Evaluation of Biocompatible Ionic Liquids for Their Application in Phytosterols Bioconversion by Mycobacterium sp. Resting Cells Jun-Jie Yuan, Yi-Xin Guan,* and Shan-Jing Yao

Downloaded via UNIV OF CAMBRIDGE on July 8, 2018 at 20:58:18 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

College of Chemical and Biological Engineering, Zhejiang University, Hangzhou 310027, China ABSTRACT: The application of ionic liquids (ILs) as novel solvents in whole-cell bioconversion has a practical significance in innovation of green process engineering for sustainable environment. Therefore, a systematical evaluation of ILs based on their cationic and anionic composition is urgently demanded in the whole-cell system. In this work, production of androst-4ene-3,17-dione (AD) via side-chain cleavage of phytosterols catalyzed by Mycobacterium sp. resting cells was employed to investigate the relationships of ILs structures with their biocompatibility and bioconversion performances. For 13 of biocompatible ILs, the viscosity, hydrophobicity, polarity and hydrogen-bond basicity, as well as IL/water composition at phase equilibrium were measured. Based upon cell membrane integrity and metabolic activity retention, some ILs with anions of [PF6] or [NTf2] showed proper Mycobacterium sp. biocompatibility. Among bioconversion systems containing water-immiscible or water-miscible ILs, [PrMIM][PF6] generated the highest AD production of 2.35 g L−1 after 12 h conversion by Mycobacterium sp. resting cells, indicating the superior of IL/ aqueous biphasic system. Furthermore, IL recovery after bioconversion was conducted to alleviate industrial expense and environmental concerns. This work presented a general procedure of ILs application in a given whole-cell bioconversion process, which can guide the design of novel ILs favorable for bioconversion process. KEYWORDS: Ionic liquid, Whole-cell bioconversion, Biphasic system, Phytosterols, Recovery, Green process



studies appeared in the following years.10−12 Therefore, a systematical evaluation of ILs based on their cationic and anionic structures is highly demanded to push forward the development of ILs in whole-cell bioconversion systems and more successful cases are expected. The microbial conversion of hydrophobic steroids, which are the second large pharmaceuticals in the world, is a key and difficult point catching the attention of many researchers. Androst-4-ene-3,17-dione (AD) is an important steroid intermediate, which will be used for the production of more than one hundred hormone drugs.13 AD can be synthesized by microbial side-chain cleavage of phytosterols that are abundant in oil residues and papermaking waste liquor. This process is a multistep reaction catalyzed by more than 10 enzymes,14 which can only be conducted at whole-cell level up to now. However, low substrate water-solubility and product inhibition are two technical problems existing in this process. In our previous work,15 [PrMIM][PF6] was applied to establish an IL/aqueous biphasic system for phytosterols side-chain cleavage by Mycobacterium sp. growing cells, in which IL acted as a substrate solubilizer and product “reservoir” to solve those two

INTRODUCTION Ionic liquids (ILs) are considered as novel green solvents with advantages of negligible vapor pressure, nonflammability, high thermostability, tunable physicochemical properties,1−4 as well as excellent dissolving power.5 Owing to these characteristics and current environmental constraints, ILs can be good alternatives to traditional volatile and flammable organic solvents in chemical and biological fields. On the other hand, green process is not only based on safe solvents, but also encouraging the use of environmentally benign catalysts.6 Biocatalysts, i.e., enzymes/whole cells, possess high specificity, nontoxicity, and the reactions are achieved under mild conditions. Nowadays, biocatalysis/bioconversion is becoming good substitute for chemical catalysis in production of some fine chemicals, especially pharmaceuticals. For these two aspects, the application of ILs as novel solvents in biocatalysis has a great practical significance in innovation of green process engineering. The early applications of ILs in biocatalysis mainly focused on enzymatic reactions; and the effects of ILs on enzymatic activity and thermostability have been revealed.1,7,8 By contrast, the investigation of whole-cell biocatalysis in the presence of ILs is more challenging due to the negative influence of ILs on cells viability and membrane integrity. The first successful utilization of ILs in whole-cell bioconversion process was reported by Cull et al. in 2000.9 Then, only several relevant © 2017 American Chemical Society

Received: August 4, 2017 Revised: September 23, 2017 Published: September 28, 2017 10702

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering

Compositions of IL/Aqueous Systems at Phase Equilibria. Amounts of IL and water (v/v, 1:10) were mixed for an infinite time. After centrifugation, 500 μL of bottom phase (IL saturated with water) and 5 mL of top phase (water saturated with IL) were transferred to glass weighing bottles individually. The initial weight of every sample bottle was measured as M1. After dried to constant weight in a vacuum drying oven, the datum was recorded as M2. Consequently, the composition in IL/aqueous system can be calculated by weight difference method:

problems. The reaction conditions were optimized and ILs applicability to this reaction was preliminarily explored. In this work, using 13 of biocompatible ILs composed of 4 kinds of cations and 4 kinds of anions, the phytosterols bioconversion was comprehensively investigated to understand the effects of ILs structures and properties on AD production. The physicochemical properties of all ILs were measured, and the ILs biocompatibility with Mycobacterium sp. cells was characterized in detail. Specifically, phytosterols bioconversion in systems containing ILs was catalyzed by Mycobacterium sp. resting cells instead of growing cells because of the stability and flexible reaction conditions required for resting cells. Finally, IL recovery after bioconversion was tentatively carried out to make this process practical and economical.



wtIL% = (M 2 − M 0)/(M1 − M 0)

in which wtIL % denotes the weight percentage of IL in either phase, and M0 is the weight of blank bottle. Microorganism Cultivation and Bioconversion. Mycobacterium sp. MB 3683 strain was preserved and activated as described by Yuan et al.,15 then inoculated into a complex medium (composed of glucose 8 g L−1, citric acid 2 g L−1, ferric ammonium citrate 0.05 g L−1, K2HPO4 0.5 g L−1, (NH4)2HPO4 6 g L−1, MgSO4 0.5 g L−1, phytosterols 1 g L−1, tween 80 0.05% (v/v), adjusted to pH 7.0). After cultured on a gyratory incubator shaker at 30 °C and 200 rpm for 60 h, cells were harvested by centrifugation and washed with pH 7.0 phosphate buffer for three times. The wet cell paste (roughly 200 mg dry cell weight per gram) was stored at −20 °C for the following experiments. Aerobic bioconversion was performed in 100 mL shake flask that was especially designed with two baffles on both sides of the bottom. About 0.5 g wet cell paste was added to 10 mL IL/buffer solution (Tris−HCl 0.1 M pH 7.5, phase ratio 1:20 (v/v)) containing 3 g L−1 phytosterols. The reaction proceeded on a gyratory incubator shaker at 30 °C and 200 rpm for 12 h. In monophasic systems, the bioconversion was carried out under the same conditions but with 5% (v/v) water-miscible ILs or with no ILs. At the end of bioconversion, AD production was analyzed by HPLC as described by Yuan et al.15 Determination of Cell Membrane Integrity and Glucose Metabolic Activity Retention. LIVE/DEAD BacLight Bacterial Viability Kit L7007 was used to determine the cell membrane integrity (MI).19,20 Mycobacterium sp. was cultivated in the above complex medium exclusive of phytosterols for 36 h, and harvested as wet cell paste. 10 g L−1 cell paste was suspended in 0.85% NaCl solution containing 0.4% or 5% (v/v) ILs at 30 °C for 12 h. Then the cells suspension was diluted (OD670 = 0.05), and stained with 0.3% (v/v) of the mixed dye (SYTO 9 and PI, 1:1). The samples were detected for the fluorescence emission spectra (excitation at 470 nm, emission at 485−700 nm) using fluorescence spectrophotometer (F-4500, Hitachi, Japan). The cell MI was defined as

EXPERIMENTAL SECTION

Materials. Mycobacterium sp. MB 3683 was obtained from China General Microbiological Culture Collection Center (Beijing, China). ILs including 11 of water-immiscible ([PF6] or [NTf2] anion; and imidazolium, pyrrolidinium, quaternaryphosphonium, or quaternaryammonium cation) and 2 of water-miscible ([BF4] or [Lac] anion; and imidazolium as cation) were supplied by Center for Greenchemistry and Catalysis, LICP, CAS (Gansu, China). Androst-4-ene3,17-dione (AD) standard with the purity ≥98% was purchased from Dibo Chemical Technology Co. Ltd. (Shanghai, China). Phytosterols, which is composed of 45% of β-sitosterol, 26.2% of stigmasterol, 23.5% of campesterol and 3.2% of brassicasterol, were provided by Xi’an Healthful Biotechnology Co. Ltd. (Shanxi, China). LIVE/DEAD BacLight Bacterial Viability Kit L7007 was purchased from Thermo Fisher Scientific Inc. (USA). Reichardt’s dye (2,6-diphenyl-4-(2,4,6triphenyl-1-pyridinio)phenolate) was from Sigma-Aldrich Chemical Inc. (USA). All other reagents were of analytical grade and available commercially. Methods. Hydrophobicity of ILs. The hydrophobicity of ILs was designated as the logarithm of octanol−water partition coefficients (log P). Concentrations of imidazolium ILs were detected as described by Kaar et al.16 based on the characteristic absorption of imidazole group at 211 nm, and the values of other ILs were determined by weighing the residual ILs after evaporating the solvents. Log P can be calculated using eq 1:

log P = log[C IL(octanol phase)/ C IL(aqueous phase)]

(1)

in which CIL denotes the concentration of per IL in either phase. Polarity of ILs. The polarity of ILs was investigated using solvatochromic probe Reichardt’s dye.17 Appropriate amount of Reichardt’s dye was added to ILs gradually until ILs were dyed to different colors. Subsequently, the visible spectra scan for these dyed ILs were conducted using a UV/visible spectrophotometer (Ultrospec 3300 pro, GE Healthcare, USA), and the maximum absorption wavelength was recorded as λmax (nm). The empirical solvent polarity was denoted by ET(30):

E T(30) (kcal · mol−1) = 28591/λmax Then, the normalized style of ET(30) was defined as

EN T = [E T(30) − 30.7]/32.4

MI = (G /R )/(G0 /R 0) × 100%

(6)

in which G and R denote the integral intensity of spectra in 510−540 nm (green) and 620−650 nm (red), respectively; G0 and R0 denote the values of cells suspended in 0.85% NaCl solution with the absence of IL. The glucose metabolic activity retention was determined by measuring the glucose consumption rate in the presence of ILs. 20 g L−1 cell paste was suspended in 10 mL Tris−HCl solution (0.1 M, pH 7.5) containing 0.4% or 5% (v/v) ILs and 2 g L−1 glucose. The blank control contained no ILs. After incubating at 30 °C and 200 rpm for 4 h, glucose consumption was measured by DNS chromotest as described by Mao et al.21 ILs Recovery. After bioconversion, the IL phase was handled by liquid extraction (ethanol, 40% (v/v) ethanol solution, n-hexane, or 1butanol). The extract was dried in a vacuum drying oven to let the solvent evaporate, and the crystallized AD product was obtained. Meanwhile, IL in the raffinate can be reused for the next batch bioconversion. In this work, the distribution coefficient of AD between extract phase and IL phase was calculated as the quotient of AD concentrations in both phases, which were measured by HPLC. The ILs recovery percentages in the process were determined by the dry weight of ILs before and after extraction. HPLC Analysis. HPLC analysis was performed by Agilent 1100 (Agilent Technologies, USA) with a hypersil ODS-2 column (250 ×

(2) ENT:16 (3)

ENT

ranged from 0 (tetramethylsilane as a nonpolar solvent) in which to 1 (water as a most polar solvent). H-Bond Basicity of ILs. The H-bond basicity (β) was measured by 4-nitroaniline/N,N-diethyl-4-nitroaniline solvatochromic method described by Jin et al.:18

β = [1.305ν(2)max − ν(1)max + 2.64]/2.80

(5)

(4)

in which ν(1)max and ν(2)max were the wave numbers for the maximum absorption of IL samples containing 4-nitroaniline and N,N-diethyl-4nitroaniline, respectively. 10703

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering

Table 1. Density (ρ), Viscosity (ν), Hydrophobicity (log P), Reichardt’s Dye Polarity (ENT) and H-Bond Basicity (β) for ILs Used in This Work at 30 °C ILs [PrMIM][PF6] [BMIM][PF6] [HMIM][PF6] [BMIM][NTf2] [HMIM][NTf2] [BMPL][NTf2] [HMPL][NTf2] [N6,2,2,2][NTf2] [N1,4,4,4][NTf2] [N10,4,4,4][NTf2] [P10,4,4,4][NTf2] [BMIM][BF4]d [BMIM][Lac]d a

ρ (g cm−3)

Full names 1-Propyl-3-methylimidazolium hexafluorophosphate 1-Butyl-3-methylimidazolium hexafluorophosphate 1-Hexyl-3-methylimidazolium hexafluorophosphate 1-Butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide 1-Hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide 1-Butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide 1-Hexyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide Triethylhexylammonium bis(trifluoromethylsulfonyl)imide Tributylmethylammonium bis(trifluoromethylsulfonyl)imide Tributyldecylammonium bis(trifluoromethylsulfonyl)imide Tributyldecylphosphonium bis(trifluoromethylsulfonyl)imide 1-Butyl-3-methylimidazolium tetrafluoroborate 1-Butyl-3-methylimidazolium lactate

a

1.41 1.36a 1.29a 1.43a 1.37a 1.39a 1.33a 1.28a 1.26a n.d.b n.d. 1.20a 1.10a

ν (cP) a,c

39 202a 357a 41a 56a 62a 87a 138a 369a n.d. n.d. 85a 684a

log P

ENT

β

−1.92 −1.61 −0.87 −0.37 0.15 0.03 0.68 0.78 1.31 2.38 2.55 −2.18 −2.70

0.70 0.67 0.67 0.75 0.67 0.58 0.73 0.84 0.81 0.79 0.49 0.70 0.64

0.10 0.19 0.09 0.24 0.13 0.19 n.d. 0.21 0.28 n.d. 0.41 0.36 0.86

NIST Standard Reference Database. bNot detected. cIL saturated with water. dWater-miscible IL.

4.6 mm, 5 μm) (Thermo Fisher Scientific, USA). Sample of 20 μL was loaded to column with a mobile phase composed of methanol and water (8:2) at a flow rate of 1.0 mL min−1, and detected at 254 nm. AD was eluted at 4.3 min.

Polarity. The subject of IL polarity has been addressed by several methodologies, e.g., maximum absorption wavelength of a solvatochromic dye, keto−enol equilibrium, and the dielectric constants measurement.22 In this work, the ENT polarity scale is based on the shift of the charge-transfer absorption band of solvatochromic dye due to hydrogen bonding between the solvent and the oxygen atom in dye. High ENT (polarity) means strong hydrogen bonding force of IL with oxygen atom, thus have potential to interfere with enzyme structure in biocatalytic system.16 Tested by this method, the ENT value of typically used [BMIM][PF6] was 0.67, which is consistent with previous researches.16,17 By comparing with more than 300 organic solvents for the ENT values described by Reichardt,23 the ILs in this work showed higher polarity than most of traditional organic solvents apart from alcohols and a few special solvents; while the frequently used alkylimidazolium ILs have similar polarity to ethanol. As shown in Table 1, all 7 of alkylimidazolium ILs independent of anionic type were with ENT values between 0.64 and 0.75, which slightly decreased with the elongation of cationic alkyl chain; the similar rule can be seen for quaternary ammonium ILs. Yet the polarity of alkylpyrrolidinium ILs increased by a large margin with the extension of cationic alkyl chain. Abnormally, the quaternary phosphonium IL showed extremely low polarity. Therefore, IL polarity depends strongly on cationic type, and the relatively high polarity of ILs suggests their potential effects on enzyme structure. Furthermore, the polarity always contacts with solvating power. It is noteworthy that the dissolving capacity of ILs cannot be simply explained by the “like dissolves like” law of traditional solvents, because ILs may behave like a polar solvent when encountered with polar solutes but display nonpolar character with nonpolar solutes, just like nanostructured materials.22 This is amazing for ILs to process some poorly soluble substances, e.g., hydrophobic steroids with multiring structure. In our previous work,15 the high solubility of steroid AD with weak polarity in ILs was just the case; besides, for the same type of ILs only with varying alkyl chain length, AD solubility increased with the decrease of IL polarity, which indeed supported the rule. H-Bond Basicity. H-bond binding is an important force for ILs as solvents. H-bond basicity is the ability of a solvent to form H-bond force as a H-bond acceptor.24 Jin et al.18,25



RESULTS AND DISCUSSION Physicochemical Properties of Selected ILs. The physicochemical property of an ideal solvent plays an important role in a biocatalytic process. In many situations, bioconversion performance can be partially predicted via the physicochemical property of the solvent used. However, relevant data of some ILs are exceedingly scarce, and an overview of physicochemical properties is necessary prior to ILs application. In this work, 13 of commercially used ILs, which have been reported to possess relatively excellent biocompatibility, are illustrated in Table 1 to discuss their physicochemical properties relevant to biocatalytic process combining with ILs structures. Density and Viscosity. The density and viscosity data of ILs used in this work were collected from NIST Standard Reference Database. As shown in Table 1, the density values of ILs ranged from 1.10 to 1.43 g cm−3, and slightly decreased with the extension of the cationic alkyl chain. The high density of ILs is beneficial for their separation from aqueous cellcontaining phase,2 especially for the water-immiscible ILs. The viscosity of pure ILs in Table 1 ranged from 41 to 684 cP, and increased with the elongation of the cationic alkyl chain; the values were greatly higher than that of pure water (approximately 1 cP). This is undoubtedly a disadvantage toward the biocatalytic process due to increased mass transfer resistance in viscous fluid. Surprisingly, it was demonstrated that the IL viscosity dropped greatly in the presence of little water.2 For example, the viscosity of [PrMIM][PF6] saturated with water was only 39 cP. Additionally, when the [PF6] anion was replaced by [NTf2], the viscosity decreased obviously. Hydrophobicity. The hydrophobicity of ILs determines whether IL/aqueous biphasic system can form spontaneously. It was indicated by logarithm of octanol−water partition coefficient (log P). As shown in Table 1, log P of tested ILs varied from −2.70 to +2.55, and the hydrophobicity strengthened with the increase of cationic alkyl chain length. Besides, the hydrophobicity sequence of different anions was [NTf2] > [PF6] > [BF4] > [Lac], thereof the last two ILs with log P < −2 cannot constitute biphasic system. 10704

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering reported the long-chain carboxylate ILs with strong H-bond basicity showed extremely high solubility for cholesterol and αtocopherol. In this work, the H-bond basicity (β) of ILs with different cations and anions was detected and shown in Table 1. The values ranged from 0.09 to 0.86. The correlationship of β with phytosterols solubility (detected in our previous work15) in these ILs is presented in Figure 1. On the whole, the

Figure 2. Weight percentages of ILs in bottom phase (IL saturated with water) and top phase (water saturated with IL) for IL/aqueous biphasic systems at phase equilibria. (Phase ratio of IL/aqueous 1:10, temperature 30 °C.)

were well consistent with those reported using UV spectroscopy.27 Biocompatibility of ILs. The biocompatibility of ILs with cells is a big concern and should be evaluated prior to their application in bioconversion. In previous reports, studies of ILs biocompatibility mainly focused on Escherichia coli, Pichia pastoris, etc.,12,28,29 few researches about Mycobacteria were referred to. In this work, the biocompatibility of ILs with Mycobacterium sp. cells was characterized by considering both cell membrane integrity and glucose metabolic activity retention in the presence of ILs. Cell Membrane Integrity. In whole-cell bioconversion processes, ILs, especially hydrophobic ILs, tend to decrease cell membrane integrity (MI), i.e., increase cell membrane permeability. As a result, substrates and products are allowed to diffuse more quickly into and out of cells, leading to the acceleration of biocatalytic reaction. However, excessive loss of MI may incur cell death.12,30 Therefore, it is significant to investigate the cell MI of Mycobacterium sp. in various ILs before bioconversion process. As shown in Figure 3a, among the 4 of ILs with the same cation [BMIM], the most hydrophobic anion [NTf2] led to the minimum level of MI, followed by [PF6]; while the maximum MIs were generated by hydrophilic [BF4] and [Lac], which were even higher than that in the single buffer system (100%). These were greatly consistent with the molecular dynamics simulations results obtained by Yoo et al.31 They showed that the hydrophobic parts of ILs tended to spontaneously insert into the lipid bilayer, and roughened the surface of the bilayer; this could be a precursor for bilayer disruption.31 On the other hand, for the cations, the elongation of alkyl chain enhanced the MIs in the mass, in accordance with the result obtained from testing on E. coli by Weuster-Botz.28 Particularly, among the 8 of [NTf2] ILs, only [P10,4,4,4][NTf2] presented a relatively high MI, implying the key role of cationic species. From Figure 3a, it can be seen that higher 5% of IL content brought about lower MIs, thus the IL addition quantity is an important parameter in bioconversion process. Glucose Metabolic Activity Retention. Glucose metabolic activity retention (MAR) is an easy indicator of cells viability,

15

Figure 1. Correlationship between phytosterols solubility in ILs and the H-bond basicity (β) of ILs. (Temperature 30 °C)

increased H-bond basicity resulted in higher substrate solubility, because the H-bond basicity of ILs strengthened their interactions with hydroxyl groups of phytosterols which act as H-bond donors. Compositions of IL/Aqueous Systems at Phase Equilibria. Generally, ILs with [PF6] or [NTf2] anion are regarded as water-immiscible solvents, and can be used to construct IL/aqueous biphasic systems. Practically in this biphasic system, a small amount of IL may dissolve into water, and some water will also enter into IL phase, thus impacting the volume and composition of each phase. In order to establish and investigate IL/aqueous system accurately, the compositions of both phases must be determined. It was reported that the water solubility data of alkylimidazolium ILs were usually measured using UV spectroscopy.26,27 However, this method is only applicable for imidazolium ILs due to the characteristic absorption of imidazole group around 211 nm. In this work, by weight difference method, all of the mutual solubility data between ILs and water were feasibly measured. As shown in Figure 2, [PrMIM][PF6] and [BMIM][PF6] (−2 < log P < −1) presented relatively high water solubility, 3.16% of [PrMIM][PF6] and 2.00% of [BMIM][PF6] can dissolve into water at 30 °C. For other ILs (log P > −1), the values were between 0.02% and 1%. Likewise, in IL enriched phases, only [PrMIM][PF6] and [BMIM][PF6] were saturated with more than 2% of water, the others ranged from 0.7% to 1.7%. Obviously, the values were always inversely related to log P, a lower log P suggested a higher mutual dissolution. Consequently, the mutual dissolution of water-immiscible ILs with water should be taken into account when constructing IL/ aqueous systems, especially for the popular [PF6] ILs with short-chain alkyl imidazolium, which possessed the log P between −2 and −1. In this work, the results of alkylimidazolium ILs were also used as references, which 10705

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering

[Lac] at 5% of IL content. This is greatly beneficial for us in the application of IL/aqueous biphasic system. As far as we know, log P have also been used to predict biocatalyst activity in traditional organic solvents. Usually, organic solvents with a log P > 4 are considered suitable for biocatalytic systems; in contrast, solvents having a small log P may distort the essential water−enzyme interactions for their hydrophilic characteristics, thereby inactivating the biocatalyst.16,34 In this work, the ILs were all with a log P < 3 and seemed not applicable in biocatalytic systems according to above assessment method in organic solvents systems. However, a lot of successful applications of these ILs, e.g., [BMIM][PF6] and [PrMIM][PF6], in biocatalysis conflicted with this hypothesis.10,21,32 Combing with the hydrophobicity and biocompatibility of ILs, we can conclude that ILs with a −2 < log P < 3 should be well enough for biocatalytic systems compared with organic solvents. All in all, the biocompatibilities of ILs characterized by these two indicators should be taken into account comprehensively. Most of all, the biocompatibility might be substantially influenced by IL content, exposure time, indicating organism and biomass.30 Therefore, prior to the application of IL in bioconversion, both of cell MI and MAR in the presence of IL should be carefully investigated under a specific situation. Bioconversion of Phytosterols to AD in Systems Containing ILs. Side-chain cleavage of phytosterols catalyzed by Mycobacterium sp. resting cells was examined with all 13 of the selected ILs. As shown in Figure 4, the highest AD

Figure 3. Biocompatibility of ILs. (a) Membrane integrity of Mycobacterium sp. MB 3683 resting cells after 12 h of incubation in the presence of 0.4% or 5% (v/v) ILs. (Wet cell concentration 10 g L−1, temperature 30 °C, buffer set as 100%.) (b) Glucose consumption by Mycobacterium sp. MB 3683 resting cells in the presence of 0.4% or 5% (v/v) ILs. (Glucose concentration 2 g L−1, wet cell concentration 20 g L−1, rotational speed 200 rpm, temperature 30 °C, buffer as a control.)

exhibiting the cells tolerance to ILs.30 In this work, MAR was determined by measuring the glucose consumption rate in buffers containing ILs. As shown in Figure 3b, the effect of cationic alkyl chain length on MAR was complicated and depended on both of cationic and anionic types. For the typically used ILs [BMIM][PF6] and [HMIM][PF6], the MARs decreased with the elongation of alkyl chain, consistent with the result obtained from testing on fungus by Mao et al.32 The same results were also gained for quaternary ammonium ILs, but the contrary trend for pyrrolidinium and imidazolium [NTf2] ILs was observed, which was also reported by Wood et al.33 when characterizing E. coli cells by inhibition zone method. Therefore, the rules are greatly dependent on ILs species. Additionally, different from MIs, the MARs were not significantly affected by the quantity of hydrophobic ILs, the values changed less or even improved with the increase of IL contents. But for hydrophilic ILs, i.e., [BMIM][BF4], and [BMIM][Lac], the MARs decreased remarkably with the increase of IL concentration. It should be noted that the MAR order was just the same to that of ILs hydrophobicity (log P): [NTf2] > [PF6] > [BF4] >

Figure 4. Bioconversion of phytosterols to AD in different IL/aqueous systems by Mycobacterium sp. MB 3683 resting cells. (Phase ratio 1:20 or with 5% (v/v) water-miscible ILs, substrate concentration 3 g L−1, wet cell concentration 50 g L−1, conversion time 12 h.)

production was obtained by [PrMIM][PF6], similar to our previous research conducted by Mycobacterium sp. growing cells.15 Compared with the 2.23 g L−1 AD production by growing cells within 120 h of bioconversion, the conversion time in this work was greatly shortened to 12 h, and the AD production was improved to 2.35 g L−1. Generally, the whole-cell bioconversion process can be substantially influenced by biocompatibility between solvents and cells; a more biocompatible environment tends to generate 10706

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering

Table 2. Distribution Coefficients of AD between Liquid Extract and ILs Phase and Recovery Percentages of ILs by Reextraction (AD initial concentration in IL 20 g L−1, volume ratio of liquid extract/IL 1:1, temperature 30 °C) Extractants Ethanol ILs [PrMIM][PF6] [BMIM][PF6] [HMIM][PF6] [BMIM][NTf2] [HMIM][NTf2] [BMPL][NTf2] [HMPL][NTf2] [N6,2,2,2][NTf2] [N1,4,4,4][NTf2] [N10,4,4,4][NTf2] [P10,4,4,4][NTf2]

Da

Rb (%)

0.708 0.472

D

R (%)

D

c

99.2 99.0 n.f.d n.f. n.f. n.f. n.f. n.f. n.f. n.f. n.f.

n-Hexane

40% Ethanol n.d. n.d. 0.046 0.016 0.011 0.027 0.016 0.024 0.026 0.023 0.037

1-Butanol R (%)

D

n.d. n.d. 94.9 99.5 99.6 99.6 92.9 99.1 96.8 96.5 85.3

0.011 0.009 0.010 0.013 0.009 0.012 0.013 0.011 0.013

R (%) n.d. n.d.

99.8 99.9 100 100 100 99.9 100 99.9 99.8

0.416 0.241 0.200 0.373 0.252 0.329 0.432

99.6 100 100 99.6 99.9 97.4 96.0 n.f. n.f.

a

Distribution coefficient of AD between liquid extract and IL phase. bRecovery percentage of IL in an extraction process. cNot detected. dWithout two phase formed.

In practice, after bioconversion, the AD in [PrMIM][PF6] phase was re-extracted with 2-fold ethanol for three times, the extract was dried and the crystallized AD was obtained from aqueous solution. Meanwhile [PrMIM][PF6] in raffinate could be recovered. Both IL recovery and AD extraction yield were above 91%, and the residue AD in IL was only 1.42 mg g−1. The recovered [PrMIM][PF6] was used for phytosterols bioconversion for batches. As shown in Figure 5, the bioconversion

higher conversion efficiency. By comparing Figure 4 with Figure 3 in detail, for the regularly used [BMIM][PF6] and [HMIM][PF6], the biocompatibility and bioconversion efficiency both decreased with the elongation of cationic alkyl chain, corresponding to the general rule and many previous researches.21,32 Nevertheless, when the experimental subjects expanded, the bioconversion will not always keep pace with biocompatibility. In fact, ILs can influence the whole-cell bioconversion efficiency by varying both cell membrane permeability and metabolic activity. On one hand, too low MI (e.g., most of [NTf2] ILs) or metabolic activity (e.g., [BMIM][Lac]) is not appropriate for biocatalysis. On the other hand, in the process of phytosterols bioconversion, both substrate and product with multiring structure are very hydrophobic and easy to dissolve out, so a moderately permeable cell membrane (MI about 40%) was a requisite for mass transfer across the membrane, which was a limiting factor than MAR.35 This may be the reason for the better performance of [PrMIM][PF6] and [BMIM][PF6] than [P10,4,4,4][NTf2] with a high MI and MAR. In addition, the biphasic systems behaved better than monophasic system containing water-miscible [BMIM][BF4] or [BMIM][Lac], suggesting the advantage of biphasic systems in production of hydrophobic and cell-inhibitory substances. Recycle of ILs. The two main obstacles against the use of ILs are the high cost and potential ecotoxicity they may bring about. In order to solve these two problems, ILs must be recovered and recycled after bioconversion. Dennewald et al. have reported the recycling of [HMPL][NTf2] in biphasic whole-cell biocatalysis, the product (R)-2-octanol was separated through distillation.36 Unlike it, in this work, the product of AD with high boiling point was separated by re-extraction. The distribution coefficients of AD between second phase and ILs phase were measured to characterize the extraction efficiency. As shown in Table 2, the extraction of AD from [PrMIM][PF6] and [BMIM][PF6] phase with ethanol possessed considerable extraction efficiency. Besides, the recovery percentages of these two ILs in an isometric extraction process were both above 99%. Unfortunately, ethanol could not be used to extract AD from other ILs owing to their intermiscibility. After that, the 40% ethanol, n-hexane, and 1-butanol were tried, and 1-butanol exhibited extracting ability for these 7 of ILs.

Figure 5. Bioconversion of phytosterols to AD with recovered [PrMIM][PF6]. (Phase ratio 1:20, substrate concentration 3 g L−1, wet cell concentration 50 g L−1, conversion time 12 h.)

efficiency kept nearly constant during five cycles. The successful recycle of IL combined with the low cost and cleanness of ethanol greatly conforms to the green concept in current world. General Evaluation Procedure for ILs Application in a Given Whole-Cell System. Combining with the concept of green process engineering, 13 of commercially used ILs including water-miscible and water-immiscible ones were evaluated for their application in phytosterols bioconversion. This work also takes an example for a general procedure of ILs applied in the whole-cell bioconversion process. First, the physicochemical properties of ILs should be determined. Among them, the hydrophobicity (log P) not only determines the suitability as a second solvent phase (log P > −2) or cosolvent but also greatly influences cell MI and MAR. The phase equilibria values of IL/water that are always inversely related to log P are essential in constructing biphasic systems, 10707

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering especially for [PF6] ILs with a log P between −2 and −1. The polarity and H-bond basicity (β) are relevant with substrate/ product solubility, and a strong H-bond basicity suggests a high solubility of H-bond donor substances (e.g., sterols) in ILs. Second, the ILs will be evaluated under the predetermined reaction with the substrate/product and cells involved. The solubility of substrate/product in ILs, and distribution coefficients (D) of substrate/product in IL/aqueous biphasic systems should be measured, especially for hydrophobic and cytotoxic substrate/product (shown in our previous report).15 Generally, a high substrate/product solubility and a log D greater than 2.0 are preferable.37 Furthermore, the biocompatibility between ILs and cells can be characterized by cell MI and MAR; an appropriate cell membrane permeability (MI was about 40% in this work) is crucial in bioconversion involving hydrophobic substrate and product. On the basis of this knowledge, ILs are tried to construct ILcontaining monophasic systems or IL/aqueous biphasic systems for the whole-cell bioconversion process (a high AD production was achieved in [PrMIM][PF6]/aqueous biphasic system using Mycobacterium sp. resting cells in this work). Ultimately, after the successful bioconversion, it is necessary to consider the economic and environmental factors of the selected IL, IL recycle using clean ethanol is significant to deal with these concerns. The evaluation procedure of ILs application established in whole-cell bioconversion provides a reference for more comprehensive and in-depth researches on ILs application and novel ILs design.



(7) Zhao, H. Protein stabilization and enzyme activation in ionic liquids: specific ion effects. J. Chem. Technol. Biotechnol. 2016, 91 (1), 25−50. (8) Lee, S. Y.; Vicente, F. A.; E Silva, F. A.; Sintra, T. E.; Taha, M.; Khoiroh, I.; Coutinho, J. A. P.; Show, P. L.; Ventura, S. P. M. Evaluating self-buffering ionic liquids for biotechnological applications. ACS Sustainable Chem. Eng. 2015, 3 (12), 3420−3428. (9) Cull, S. G.; Holbrey, J. D.; Vargas-Mora, V.; Seddon, K. R.; Lye, G. J. Room-temperature ionic liquids as replacements for organic solvents in multiphase bioprocess operations. Biotechnol. Bioeng. 2000, 69 (2), 227−233. (10) Wu, D. X.; Guan, Y. X.; Wang, H. Q.; Yao, S. J. 11αHydroxylation of 16α,17-epoxyprogesterone by Rhizopus nigricans in a biphasic ionic liquid aqueous system. Bioresour. Technol. 2011, 102 (20), 9368−9373. (11) Cornmell, R. J.; Winder, C. L.; Schuler, S.; Goodacre, R.; Stephens, G. Using a biphasic ionic liquid/water reaction system to improve oxygenase-catalysed biotransformation with whole cells. Green Chem. 2008, 10 (6), 685−691. (12) Wang, N. Q.; Li, J.; Sun, J.; Huang, J.; Wang, P. Bioreduction of 3,5-bis(trifluoromethyl)acetophenone using ionic liquid as a co-solvent catalyzed by recombinant Escherichia coli cells. Biochem. Eng. J. 2015, 101, 119−125. (13) Xu, Y. G.; Guan, Y. X.; Wang, H. Q.; Yao, S. J. Microbial sidechain cleavage of phytosterols by Mycobacteria in vegetable oil/ aqueous two-phase system. Appl. Biochem. Biotechnol. 2014, 174 (2), 522−533. (14) Szentirmai, A. Microbial physiology of sidechain degradation of sterols. J. Ind. Microbiol. 1990, 6 (2), 101−115. (15) Yuan, J. J.; Guan, Y. X.; Wang, Y. T.; Wang, H. Q.; Yao, S. J. Side-chain cleavage of phytosterols by Mycobacterium sp. MB 3683 in a biphasic ionic liquid/aqueous system. J. Chem. Technol. Biotechnol. 2016, 91 (10), 2631−2637. (16) Kaar, J. L.; Jesionowski, A. M.; Berberich, J. A.; Moulton, R.; Russell, A. J. Impact of ionic liquid physical properties on lipase activity and stability. J. Am. Chem. Soc. 2003, 125 (14), 4125−4131. (17) Fletcher, K. A.; Storey, I. A.; Hendricks, A. E.; Pandey, S.; Pandey, S. Behavior of the solvatochromic probes Reichardt’s dye, pyrene, dansylamide, Nile Red and 1-pyrenecarbaldehyde within the room-temperature ionic liquid bmimPF6. Green Chem. 2001, 3 (5), 210−215. (18) Jin, W. B.; Yang, Q. W.; Huang, B. B.; Bao, Z. B.; Su, B. G.; Ren, Q. L.; Yang, Y. W.; Xing, H. B. Enhanced solubilization and extraction of hydrophobic bioactive compounds using water/ionic liquid mixtures. Green Chem. 2016, 18 (12), 3549−3557. (19) Boulos, L.; Prevost, M.; Barbeau, B.; Coallier, J.; Desjardins, R. LIVE/DEAD® BacLight: application of a new rapid staining method for direct enumeration of viable and total bacteria in drinking water. J. Microbiol. Methods 1999, 37 (1), 77−86. (20) Mao, S. H.; Yu, L.; Ji, S. X.; Liu, X. G.; Lu, F. P. Evaluation of deep eutectic solvents as co-solvent for steroids 1-en-dehydrogenation biotransformation by Arthrobacter simplex. J. Chem. Technol. Biotechnol. 2016, 91 (4), 1099−1104. (21) Mao, S. H.; Hua, B. Y.; Wang, N.; Hu, X. J.; Ge, Z. J.; Li, Y. Q.; Liu, S.; Lu, F. P. 11α Hydroxylation of 16α, 17-epoxyprogesterone in biphasic ionic liquid/water system by Aspergillus ochraceus. J. Chem. Technol. Biotechnol. 2013, 88 (2), 287−292. (22) Van Rantwijk, F.; Sheldon, R. A. Biocatalysis in ionic liquids. Chem. Rev. 2007, 107 (6), 2757−2785. (23) Reichardt, C. Solvatochromic dyes as solvent polarity indicators. Chem. Rev. 1994, 94, 2319−2358. (24) Yang, Q. W.; Xu, D.; Zhang, J. Z.; Zhu, Y. M.; Zhang, Z. G.; Qian, C.; Ren, Q. L.; Xing, H. B. Long-chain fatty acid-based phosphonium ionic liquids with strong hydrogen-bond basicity and good lipophilicity: synthesis, characterization, and application in extraction. ACS Sustainable Chem. Eng. 2015, 3 (2), 309−316. (25) Jin, W. B.; Yang, Q. W.; Zhang, Z.; Bao, Z. B.; Ren, Q. L.; Yang, Y. W.; Xing, H. B. Self-assembly induced solubilization of drug-like

AUTHOR INFORMATION

Corresponding Author

*Prof. Yi-Xin Guan. Tel: +86-571-87951982. Fax: +86-57187951982. E-mail: [email protected]. ORCID

Yi-Xin Guan: 0000-0001-6983-7843 Shan-Jing Yao: 0000-0003-3199-3044 Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China. REFERENCES

(1) Ghaedizadeh, S.; Emamzadeh, R.; Nazari, M.; Rasa, S. M. M.; Zarkesh-Esfahani, S. H.; Yousefi, M. Understanding the molecular behaviour of Renilla luciferase in imidazolium-based ionic liquids, a new model for the α/β fold collapse. Biochem. Eng. J. 2016, 105, 505− 513. (2) Quijano, G.; Couvert, A.; Amrane, A. Ionic liquids: Applications and future trends in bioreactor technology. Bioresour. Technol. 2010, 101 (23), 8923−8930. (3) Amarasekara, A. S. Acidic ionic liquids. Chem. Rev. 2016, 116 (10), 6133−6183. (4) Marques, M. P. C.; Carvalho, F.; de Carvalho, C. C. C. R.; Cabral, J. M. S.; Fernandes, P. Steroid bioconversion: towards green processes. Food Bioprod. Process. 2010, 88 (1), 12−20. (5) Lou, W. Y.; Zong, M. H.; Smith, T. J. Use of ionic liquids to improve whole-cell biocatalytic asymmetric reduction of acetyltrimethylsilane for efficient synthesis of enantiopure (S)-1-trimethylsilylethanol. Green Chem. 2006, 8 (2), 147−155. (6) Lozano, P.; Bernal, J. M.; Gómez, C.; García-Verdugo, E.; Isabel Burguete, M.; Sánchez, G.; Vaultier, M.; Luis, S. V. Green bioprocesses in sponge-like ionic liquids. Catal. Today 2015, 255, 54−59. 10708

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709

Research Article

ACS Sustainable Chemistry & Engineering molecules in nanostructured ionic liquids. Chem. Commun. 2015, 51 (67), 13170−13173. (26) Neves, C. M. S. S.; Batista, M. L. S.; Cláudio, A. F. M.; Santos, L. M. N. B.; Marrucho, I. M.; Freire, M. G.; Coutinho, J. A. P. Thermophysical properties and water saturation of [PF6]-based ionic liquids. J. Chem. Eng. Data 2010, 55 (11), 5065−5073. (27) McFarlane, J.; Ridenour, W. B.; Luo, H.; Hunt, R. D.; DePaoli, D. W.; Ren, R. X. Room temperature ionic liquids for separating organics from produced water. Sep. Sci. Technol. 2005, 40 (6), 1245− 1265. (28) Weuster-Botz, D. Process intensification of whole-cell biocatalysis with ionic liquids. Chem. Rec. 2007, 7 (6), 334−340. (29) Ganske, F.; Bornscheuer, U. T. Growth of Escherichia coli, Pichia pastoris and Bacillus cereus in the presence of the ionic liquids [BMIM][BF4] and [BMIM][PF6] and organic solvents. Biotechnol. Lett. 2006, 28 (7), 465−469. (30) Xu, P.; Zheng, G. W.; Du, P. X.; Zong, M. H.; Lou, W. Y. Whole-cell biocatalytic processes with ionic liquids. ACS Sustainable Chem. Eng. 2016, 4 (2), 371−386. (31) Yoo, B.; Shah, J. K.; Zhu, Y.; Maginn, E. J. Amphiphilic interactions of ionic liquids with lipid biomembranes: a molecular simulation study. Soft Matter 2014, 10 (43), 8641−8651. (32) Mao, S. H.; Hu, X. J.; Hua, B. Y.; Wang, N.; Liu, X. G.; Lu, F. P. 15α-Hydroxylation of a steroid (13-ethyl-gon-4-en-3,17-dione) by Penicillium raistrickii in an ionic liquid/aqueous biphasic system. Biotechnol. Lett. 2012, 34 (11), 2113−2117. (33) Wood, N.; Ferguson, J. L.; Gunaratne, H. Q. N.; Seddon, K. R.; Goodacre, R.; Stephens, G. M. Screening ionic liquids for use in biotransformations with whole microbial cells. Green Chem. 2011, 13 (7), 1843. (34) Laane, C.; Boeren, S.; Vos, K.; Veeger, C. Rules for optimization of biocatalysis in organic solvents. Biotechnol. Bioeng. 1987, 30 (1), 81−87. (35) Shen, Y.; Wang, L.; Liang, J.; Tang, R.; Wang, M. Effects of two kinds of imidazolium-based ionic liquids on the characteristics of steroid-transformation Arthrobacter simplex. Microb. Cell Fact. 2016, 15 (1), 118−127. (36) Dennewald, D.; Pitner, W.; Weuster-Botz, D. Recycling of the ionic liquid phase in process integrated biphasic whole-cell biocatalysis. Process Biochem. 2011, 46 (5), 1132−1137. (37) Pfruender, H.; Jones, R.; Weuster-Botz, D. Water immiscible ionic liquids as solvents for whole cell biocatalysis. J. Biotechnol. 2006, 124 (1), 182−190.

10709

DOI: 10.1021/acssuschemeng.7b02681 ACS Sustainable Chem. Eng. 2017, 5, 10702−10709