Selectivity of lipases for estolides synthesis

Materials

12-Hydroxystearic acid (99 %), 16-hydroxyhexadecanoic acid (98 %), ricinoleic acid (98 %), Bis(trimethylsilyl)trifluoroacetamide-trimethylchlorosilane (BSTFA + TMCS = 99:1), trans-2-[3-(4-t-butyl-phenyl)-2-methyl-2-propenylidene]malononitrile (DCTB), potassium trifluoroacetate (KTFA), Amano lipase from Pseudomonas fluorescens (P. fluorescens), Pseudomonas cepacia (P. cepacia), Thermomyces lanuginosus (T. lanuginosus), and CLEA Alcalase were purchased from Sigma Aldrich. The lipases from C. rugosa, Aspergillus oryzae (A. oryzae), Candida antarctica A (C. antarctica A), Rhizopus arrhizus (R. arrhizus), Penicillium roqueforti (P. roqueforti), Mucor javanicus (M. javanicus), Candida lipolytica (C. lipolytica), Thermomyces brockii (T. brockii), and hog pancreas were obtained from Fluka. Novozyme 435, Lipozyme TL, Candida antarctica B (C. antarctica B) lipase were purchased from Novozymes. The lipases from Alcaligenes sp (lipase TL), Pseudomonas stutzeri (P. stutzeri) (Lipase TL), were products of Meito Sangyo Co. Ltd., Japan. The CLEA from C. antarctica B and P. stutzeri were purchased from CLEA Technologies. The organic solvents used (heptane, toluene, cyclohexane, methyl ethyl ketone and N,N-dimethyl formamide) were from Merck and their water content (determined by Karl Fischer titration) was below 0.01 %.

Lipases activity assay

The hydrolytic activity of lipases was determined using p-nitro-phenyl palmitate as substrate, as reported elsewhere [15].

Biocatalytic synthesis of estolides

The reactions were performed by adding native and immobilized lipases, 50 U/mmol substrate, to different amounts of substrate (1–5 mM), brought to a final volume of 1 mL in the selected organic solvent. The reactions have been carried out at temperatures between 40 °C and 75 °C, while stirring at 350 rpm for 24 h. At the end of the reaction the enzyme was removed by centrifugation (13 000 rpm, 3 min), and the reaction mixture was analyzed by GC-MS and MALDI-TOF MS.

The HFA conversion was determined based on the GC-MS analysis, after derivatization with BSTFA + TMCS (99:1), at 2:1 reagent: sample ratio (w/w), for 1 h at 95 °C, as previously described [16]. Hexadecane was used as internal standard. The separation was carried out using an Interscience Trace GC Ultra GC+ PTV instrument, equipped with AS3000 II autosampler and Restek Rxi-5ms 30 m × 0.25 mm × 0.25 μm capillary column. The analysis conditions were: oven temperature 100–300 °C with 10 °C/min heating rate, injector temperature 300 °C, carrier gas (helium) flow 1.0 mL/min. Mass spectra were obtained from Interscience Trace DSQ II XL quadrupole mass selective detector (EI, mass range 35–500 Dalton, 150 ms sampling speed), the mass spectrometer being operated at 70 eV. For the calculation of conversions at a defined reaction time, calibration curves were built up for each substrate in the range from 0.05 mM to 3 mM, using hexadecane as internal standard.

MALDI-TOF MS analysis was carried out for composition analysis of estolides, using an Ultraflex Workstation with Flex Control and Flex Analysis software packages for acquisition and processing of the data (Bruker Daltonics, Germany), at an acceleration voltage of 25 kV, using DCTB as matrix and KTFA as ionization agent. 10 μL of the sample was mixed with 10 μL of matrix solution (40 mg/mL DCTB solubilized in THF) and 3 μL of KTFA (5 mg/mL). About 0.3 μL of the mixture have been applied on the plate and measured in the positive mode. Detection was performed in the reflector mode. The number average molecular weight M_n, weight average molecular weight M_w and the polydispersity index PDI = M_w/M_n have been calculated as described elsewhere [17].

Specificity of lipases in the synthesis of estolides

The substrate specificity study of lipases was carried out to establish the differences in terms of product formationand substrate selectivity of enzymes from different sources. The esterification reactions of HFA were performed by using the same enzyme activity related to the substrate, i.e., 50 units/mmol substrate, in every reaction. In order to monitor the monolactone formation, the reactions have been performed at 1 mM substrate concentration, since previous works showed that lipase-catalyzed synthesis of cyclic esters is favored by lower substrate concentrations [11, 20]. The temperature was set at 40 °C, the optimum temperature for most lipases, using toluene as reaction media. After 24 h reaction time, the processes were stopped by centrifugation of the enzyme and the samples were derivatized and analyzed by GC-MS, as described in the Methods section. The GC-MS results together with the qualitative analysis performed by MALDI-TOF MS are presented in Table 1. Unexpectedly, monolactones were not identified in all reaction products (Table 1). The formation of the monolactone (easily monitored by GC-MS) was observed for all three tested substrates only when the lipases from P. fluorescens and P. stutzeri were used as catalyst. For the substrate bearing primary OH group, 16HHDA the monolactone was formed also by P. cepacia and Alcaligenes PL catalysis. The chain elongation for 16HHDA was favored by the lipases from P. cepacia, C. antarctica B, hog pancreas and P. stuzeri when the presence of the species as trimer and tetramer were identified in the MALDI-TOF MS spectra.

The highest conversion was achieved for 16HHDA. These results are in accordance with the results reported by Makita et al. [11], and can be explained by the regioselectivity of lipases for the primary alcohol groups, compared to the secondary alcohol groups due to the size of the active site and the orientation of the active site Ser in the middle. Secondary alcohols need to be bend to fit into the active site resulting in less degrees of freedom, and therefore more difficult binding. There were 5 lipases that led to singular products, the most interesting were the lipases from R. arrhizus and T. lanuginosus, that were able to form the dilactone with conversions higher than 40 %. The native lipase from C. antarctica B also showed to be promising for estolide synthesis, particularly from 16HHDA.

In the case of C18 substrates (12HSA and RCA), important differences were noticed. Even though each of these substrates contains a secondary OH group in the same position (C12), several lipases, such as C. rugosa, A. oryzae, M. javanicus, C. antarctica A, R. arrhizus, P. roqueforti, P. cepacia, T. langinosus and hog pancreas were not at all active against the saturated substrate (12HSA), but led to conversions up to 35 % for ricinoleic acid. Moreover, there were lipases that presented no activity for RCA, namely C. lipolytica, Alcaligenes PL and T. brockii, but led to 12HSA based monoestolide (referred as dimer D in Table 1) formation. Another important conclusion for these two substrates is that monoestolide was synthesized as unique product by all tested lipases, excepting those from Pseudomonas.

Even though such substrates have been already tested, especially ricinoleic acid, alone [10] or in combination with different polyols or fatty acids [21], there are no reports concerning the selectivity of several commercially available lipases, as we performed in this study. Moreover, in previous reports substrate conversion or chain elongation have been monitored solely by titration [9] or thin layer chromatography [14]. In our study, the MALDI-TOF technique allowed the identification of all the species present in the reaction mixture.

Effect of temperature and substrate concentration

To investigate the effect of temperature and substrate concentration on conversion, the concentration of the substrate has been increased 5 times, up to 5 mM, and the temperature raised up to 75 °C.

As preliminary experiments for the thermal stability study several lipases, selected according to their capability to favor the chain elongation reaction and/or higher conversion, were incubated up to 24 h, at temperatures up to 80 °C (Figure 1S). Excepting the lipase from T. lanuginosus the thermal stabilities were excellent, as they preserved more than 85 % of the initial hydrolytic activities.

The results presented in Fig. 2 indicate that the conversions were about 2-fold higher when the temperature and substrate concentration have been increased for 12HSA and RCA (excepting CalB). Nevertheless, the chain elongation for the C18 substrates was favored only by the lipases from Ps. species (data not shown). Because the lipase from P. fluorescens proved to be efficient for all three substrates, further experiments were performed with this lipase.

Fig. 2

Influence of temperature and substrate concentration

1 mM 40 °C,

5 mM 75 °C on the conversion of HFA catalyzed by lipases from P. fluorescens, P. stutzeri and C. antarctica B.

Influence of the reaction medium

Solvents with different polarity spanning logP [22] values between –1.0 and 4.3 were investigated to disclose the influence of the reaction medium polarity on the formation of estolides. The reactions were performed at 75 °C, 5 mM substrate concentration, using as catalyst 50 U/mmol substrate lipase from P. fluorescens. The results presented in Table 2 indicate that the substrate conversion and estolide formation were favored by the increase of the logP value of the solvent, meaning decreasing polarity. The results are in concordance with previous reports, where apolar solvents have been successfully used for the synthesis of monolactones [11].

Products were not detected when the most polar solvents, N,N-dimethyl formamide and methyl ethyl ketone, were used as reaction medium. The highest conversion values were achieved, as expected, for the 16HHDA substrate, due to the high selectivity of the lipase for the primary OH group, as previously discussed. Even though the chain elongation was favored by heptane, leading to oligomers with polymerization degrees up to 6, identified in the MALDI-TOF MS spectrum, the solubilization of the substrate into heptane was not complete. Consequently, a real conversion in this solvent could not be calculated. The higher boiling point of toluene allows more elevated reaction temperatures, which can increase the reaction rate. Considering both conversion values and formation of products with higher chain elongation, toluene can be considered the best reaction medium to produce estolides.

Synthesis of estolides by immobilized lipases

In order to compare the catalytic efficiency and selectivity for the chain elongation reaction of different immobilized enzymes in estolides synthesis, several commercially available immobilized lipases and one protease have been tested. The reactions were performed as previously described, using 5 mM substrate concentration, toluene as solvent and 24 h reaction time at 75 °C. CLEA lipases from P. stutzeri (CLEA-P. stutzeri) C. antarctica B (CLEA CalB), CLEA Alcalase, as well as T. lanuginosus lipase immobilized on granulated silica (Lipozyme TL) and C. antarctica immobilized on acrylic resin (Novozyme 435) were tested. M_n and M_w values were calculated based on the MALDI-TOF MS spectra, as described elsewhere [17]. The results indicate that although conversions higher than 70 % were obtained for 16HHDA, the higher polymerization degree (up to 10) was achieved when RCA was used as substrate and CLEA- P. stutzeri lipase as catalyst (Table 3).

The formation of estolides with polymerization degree ranging from 2 to 10 could be observed in the MALDI-TOF MS spectrum of the RCA-based reaction product (as potassium adduct), as presented in Fig. 3. Lower molecular weight estolides were previously reported for ricinoleic acid as substrate when the immobilized lipase from Staphylococcus xylosus has been used [23].

Fig. 3

MALDI-TOF MS spectrum of ricinoleic-based estolides catalyzed by CLEA- P. stutzeri lipase.

CLEA P. stutzeri lipase was the most efficient lipase with respect to chain elongation, for all three substrates tested. In the case of 12HSA substrate, formation of the monoestolide was observed in most of the cases, while formation of the monoestolide together with the dilactone was favored by the CLEA- P. stutzeri lipase. However, considering the highest selectivity of lipase for the dimer formation without any by-product and the usefulness of the monoestolide for cosmetic applications [24], this synthesis route is very promising, particularly in the perspective of further development for reuse of the catalyst and operation in continuous systems.

Interesting results have been obtained with CLEA Alcalase as well, proving the ability of this enzyme to use HFA as substrates for estolides synthesis in good yields, even though it is a protease.