Introduction

α-Lipoic acid, which has been isolated [1] from natural sources and identified as the R configuration [2], is a growth factor for a variety of microorganisms and a cofactor involved in many enzyme-catalyzed reactions, particularly in decarboxylation of α-keto acids [1, 3–5]. Since α-lipoic acid exhibits a high level of biological activity, its use in the treatment of various diseases has been investigated. For example, it shows effects in diabetes mellitus [5] and hepatic diseases [6, 7] as well as anti-oxidative [8, 9], anti-inflammatory [10] and immunological activity [11]. Recently, it has also been reported that α-lipoic acid and its derivatives are highly active as anti-HIV [12, 13] and anti-tumor agents [14].

Generally, the (R)-enantiomer is much more active than the (S)-enantiomer [15]. Therefore, great attention has been focused on the stereoselective synthesis of pure (R)-enantiomer. There are two main specific routes for the synthesis of (R)- and (S)-α-lipoic acid by chemical methods, including asymmetric synthesis [16–18] and a strategy starting from ‘chiral pool’ material [2, 19–23]. Various methods of enzyme catalysis, such as enzyme-catalyzed reactions of key intermediates or precursors [24–28] and kinetic resolution of racemic material [29], have also been developed. Lipase-catalyzed kinetic resolution of α-lipoic acid involves esterifying the carboxylic group located four carbon atoms away from the stereogenic center. Although several examples of enzymatic resolution of key compounds with remote stereocenters have been reported [30–34], in practice, enzymatic or classical resolution involving a remote chiral center remains a difficult task. To date, no enzymes have been reported to enable kinetic resolution of racemic α-lipoic acid except for the commercially available Candida rugosa lipase, which shows enantioselectivity towards the (S)-enantiomer. However, results obtained with this latter enzyme are unsatisfactory, with a low enantiomeric excess of 23.8% towards (R)-α-lipoic acid [29]. In this study, lipase from Aspergillus oryzae WZ007 was used to catalyze kinetic resolution of α-lipoic acid. We present evidence to suggest that this enzyme can enantioselectively esterify (S)-α-lipoic acid to the corresponding (S)-ester (Scheme 1), and that the residual (R)-α-lipoic acid can be recovered by extraction with an enantiomeric excess of 92.5%. Various reaction parameters affecting the conversion rate and enantioselectivity were investigated.

Scheme 1
scheme 1

Kinetic resolution of racemic α-lipoic acid by lipase from Aspergillus oryzae WZ007

Full size image

Materials and methods

Chemicals and reagents

α-Lipoic acid was purchased from Fluka BioChemika (Buchs, Switzerland). Porcine pancreas lipase (Type ΙΙ) was purchased from Sigma (St. Louis, MO). Lipases from Penicillium expansum and Aspergillus niger were purchased from Shenzhen Leveking Bio-engineering (Shenzhen, Guangdong, China). All the other chemicals were obtained from commercial sources and were of analytical reagent grade.

Microorganisms and culture conditions

The strain of Aspergillus oryzae WZ007 was newly isolated from lipase-producing strains and deposited with the China Center for Type Culture Collection with the accession number of CCTCC No. M206105. The strain was maintained on slant medium consisting of potato 200 g/l, glucose 20 g/l and agar 17 g/l. The culture was grown aerobically at 30°C and 200 rpm for 48 h in cell growth medium consisting of glucose 10 g/l, peptone 5 g/l, KH2PO4 1 g/l, MgSO4 ⋅ 7H2O 0.5 g/l, FeSO4 ⋅ 7H2O 0.01 g/l, KCl 0.5 g/l and olive oil 10 ml/l. After harvesting by centrifugation, the mycelium was thoroughly washed with distilled water and 100 mM Tris-HCl buffer (pH 7.0) and then dried using a freeze-drying system (Christ, Osterode am Harz, Germany).

Esterification reaction

In a typical experiment, α-lipoic acid (206 mg, 1 mmol), n-octanol (0.79 ml, 5 mmol), heptane (20 ml) and lyophilized microbial cells (200 mg) were placed in a conical flask (100 ml). The reaction mixtures were incubated at 200 rpm and 37°C for 48 h on a shaker. Reaction mixtures without microbial cells were also run to exclude any possible spontaneous chemical reaction. The esterifying reaction was quenched by removing mycelium or enzyme powder through centrifugation. Unreacted α-lipoic acid was extracted with 40 ml 0.5% (w/v) sodium bicarbonate and recovered after acidification with 20% (v/v) hydrochloric acid and extraction with dichloromethane. Dichloromethane was removed by vacuum distillation and the recovered α-lipoic acid was dissolved in acetonitrile for subsequent high performance liquid chromatography (HPLC) analysis.

Analytical methods

Conversion rate and enantiomeric excess of α-lipoic acid were assayed by HPLC 1100 (Agilent, Wilmington, DE) with a Chiralpak AS-H column (250 mm × 4.6 mm, 5 μm, Daicel, Hyogo, Japan). The mobile phase was composed of hexane/2-propanol/trifluoroacetic acid at a ratio of 97/3/0.1. The flow rate was 0.8 ml/min. Absorbance of column effluents was monitored at 220 nm. Enantioselectivities (E values) were calculated from conversion rate and enantiomeric excess according to the following equation [35].

$$ E = \frac{{\ln [(1 - c)(1 - ee_{\text{S}} )]}}{{\ln [(1 - c)(1 + ee_{\text{S}} )]}} $$
$$ c = \frac{{c_{0} - c_{\text{e}} }}{{c_{0} }} \times 100\% $$
$$ ee_{\text{s}} = \frac{[R] - [S]}{[R] + [S]} \times 100\% $$

where c is the conversion ratio of reaction, c 0 the initial amount of racemic α-lipoic acid, c e the amount of racemic α-lipoic acid at the end of reaction, and ee s is the enantiomeric excess of the residual α-lipoic acid; [R] and [S] are the peak areas corresponding to the (R)-α-lipoic acid isomer and (S)-α-lipoic acid isomer, respectively.

Results and discussion

Screening of lipase

Lipases from Penicillium expansum, Aspergillus niger, porcine pancreas and Aspergillus oryzae WZ007 were used in this study. Lipase powder from P. expansum, A. niger and porcine pancreas were commercially available. Lipase from A. oryzae WZ007 was produced by fermentation. A strain of A. oryzae WZ007 with high activity and enantioselectivity towards biotin intermediate 1,3-dibenzyl-5-(hydroxymethyl)-2-oxo-4-imidazolidinecarboxylic acid was newly isolated from soil samples by our team in a previous study [36]. Esterification reactions of racemic α-lipoic acid using the four lipases demonstrated that only the lipase from Aspergillus oryzae WZ007 exhibited high esterification ability and enantioselectivity towards the (S)-enantiomer. The other three lipases showed the opposite enantioselectivity. Hydrolysis of the (R)-α-lipoic acid esters formed is thus also a potential alternative route to produce (R)-α-lipoic acid. α-Lipoic acid could not be esterified for the reaction system in the absence of enzyme. In this work, lipase from A. oryzae WZ007 was chosen for further studies.

Optimization of esterification conditions

Effect of alcohol chain length

To investigate the effect of alcohols with different chain length on the conversion rate and enantioselectivity, esterification reactions of α-lipoic acid in heptane were carried out with n-butanol (C4), n-pentanol (C5), n-hexanol (C6), n-octanol (C8), n-decanol (C10) and n-dodecanol (C12), respectively. The results in Table 1 indicate that there was a marked tendency for the enantioselectivity, together with the conversion rate, to rise as the alcohol chain length increased up to n-octanol. Use of alcohols with longer chains, such as n-decanol and n-dodecanol, did not increase the resolution of α-lipoic acid further. Thus, n-octanol was chosen for further study. Similarly, longer alcohols, such as n-octanol and n-decanol, resulted in a higher initial rate and enantioselectivity than n-hexanol when enantioselective esterification of ibuprofen was catalyzed by Candida rugosa lipase in isooctane [37]. The results suggested a mode of binding of the acyl donor (acid) in a hairpin conformation that leaves the active site tunnel empty to accommodate the acyl acceptor (alcohol) [38]. The exact location of the alcohol molecule in the tunnel is crucial for the enantioselectivity of the lipase, and the optimal carbon chain length of alcohol would therefore fit the active site tunnel of the lipase to provide both high enzyme activity and enantioselectivity [39].

Table 1 Effects of alcohol chain length on the conversion rate and enantioselectivity of esterification of α-lipoic acid
Full size table

Effect of solvents

It has been demonstrated that the activity and enantioselectivity of lipases is greatly affected by the nature of the non-aqueous solvent used when catalysis occurs in a nearly anhydrous environment [40, 41]. The effects of various organic solvents, ranging from hydrophobic to hydrophilic, were tested in the typical esterification reaction experiment, replacing heptane with the other organic solvents. The log P-value of the solvents was the usual parameter taken to express solvent hydrophobicity and its possible effects on enzyme activity in the non-aqueous phase [41]. The results are shown in Table 2. In hydrophobic solvents, the conversion rate increased with increase in log P values, reaching a maximum value of 71.8% in hexane (log P = 3.5). Enantioselectivity increased with increasing log P values up to 4.3 in heptane. No conversion occurred in hydrophilic solvents such as tetrahydrofuran, acetonitrile and acetone. The highest enantiomeric excess of (R)-α-lipoic acid (ee s = 74.6%), enantioselectivity (E = 4.3) was obtained using heptane. The activity and enantioselectivity of an enzyme decreases as the hydrophobicity of the solvent decreases, and hydrophilic solvents may alter or denature enzymes by stripping off essential water from the enzyme [42]. More hydrophilic organic solvents can result in conformational changes in the enzyme that affect the affinity of the substrate-binding site for its ligand and the enantioselectivity of the enzyme [39].

Table 2 Effects of different organic solvents on the conversion rate and enantioselectivity of esterification of α-lipoic acid
Full size table

Effect of molar ratio of n-octanol to acid

In enzymatic stereoselective esterification reactions, alcohols act as nucleophiles and their concentration is known to affect reaction rate and enantioselectivity. Enantio-selective esterification reactions were carried out at five different molar ratios of n-octanol to acid, i.e., 4:1, 5:1, 6:1, 7:1, and 8:1, respectively. As shown in Fig. 1, upon increasing n-octanol concentration in the esterification reaction, the conversion rate and enantioselectivity increase up to an optimum of 68.0% and E = 4.3 (ratios 5:1 or 6:1), respectively. However, when the n-octanol concentration increased further (ratios 7:1 and 8:1), enantioselectivity dropped. A good explanation for the observed change in enantioselectivity is that the two diastereomeric acyl-enzymes have different K m values for the alcohol. A typical case of this would be a situation where both acyl-enzymes are saturated with alcohol at the high concentration, whereas at the low concentration only the S-acyl-enzyme is saturated, leading to a higher E-value [43]. This is also explained at a molecular level by a molecular modeling study. In contrast to the fast-reacting enantiomer, the slow-reacting R-enantiomer leaves the tunnel empty. It is thus suggested that the alcohol coordinates to the tunnel and inhibits the fast-reacting but not the slow-reacting enantiomer. An increased alcohol concentration would therefore decrease the enantioselectivity [38].

Fig. 1
figure 1

Effect of the molar ratio of n-octanol to racemic α-lipoic acid on the enantioselective esterification of racemic α-lipoic acid. Reaction conditions: 1 mmol racemic α-lipoic acid, different molar ratios of n-octanol to racemic α -lipoic acid and 200 mg microbial cells in 20 ml heptane, reaction time = 48 h, reaction temperature = 37°C, shaking at 200 rpm. △ Conversion, □ ee s, ○ enantioselectivity

Full size image

Effect of temperature

The esterification reaction of α-lipoic acid and n-octanol catalyzed by whole-cell lipase from A. oryzae WZ007 was carried out in heptane at temperatures ranging from 32 to 55°C. The results (shown in Fig. 2), demonstrate that both the conversion rate and enantioselectivity increase with increasing reaction temperature up to 50°C. At higher temperatures, the conversion rate and enantioselectivity decreased gradually. A reaction temperature of 50°C was selected as an optimum temperature for the reaction as it exhibited the highest enantioselectivity and appropriate conversion of 75.2% with an ee s of 92.5% at a reaction time of 48 h.

Fig. 2
figure 2

Effect of reaction temperature on the enantioselective esterification of racemic α-lipoic acid. Reaction conditions: 1 mmol racemic α-lipoic acid, 5:1 molar ratio of n-octanol to racemic α-lipoic acid and 200 mg microbial cells in 20 ml heptane, reaction temperature ranging from 32 to 55°C, reaction time = 48 h, shaking at 200 rpm. △ Conversion, □ ee s, ○ enantioselectivity

Full size image

Conclusions

Lipase from A. oryzae WZ007 can enantioselectively esterify (S)-α-lipoic acid, leaving the target product (R)-α-lipoic acid in the unreacted form, although the reaction center is four carbon atoms away from the stereogenic center. The effects of various reaction parameters on the performance of lipase from A. oryzae WZ007 in the preparation of enantiopure (R)-α-lipoic acid were studied. When α-lipoic acid was esterified with n-octanol at 50°C in heptane for 48 h with a molar ratio of alcohol to acid of 5:1, the enantiomeric excess of (R)-α-lipoic acid could reach as high as 92.5%, with a conversion rate of 75.2%.