Lipase-Catalyzed Synthesis of Citronellyl Esters by Different Acylating Agents: A Kinetic Comparison Study

Abstract

Citronellyl esters present broad applications in the cosmetic, pharmaceutical, and food industries due to their organoleptic and biological properties. Lipase-catalyzed acylation of citronellol has been a suitable way to obtain these natural esters on a high scale. This work describes the lipase-catalyzed acylation of citronellol through esterification and transesterification with linear acylating agents (C₃-C₁₈) using Novozym^®435 as a biocatalyst. The results show that all esters achieved not less than 91% conversion from an esterification model, showing a good reaction profile. Citronellyl propionate and laurate were selected for study in a transesterification reaction model, and it resulted in a lower reaction time for ester conversions. As an example, we highlight citronellyl propionate production, which approximately 10 g L^-1 of the product was obtained after 1.5 h of esterification compared to 2 min by transesterification. The final esters were purified with high purity (≥ 90%) and identified by gas chromatography-mass spectrometry (GC-MS), and their structures were elucidated by ¹H and ¹³C nuclear magnetic resonance (NMR). Enzyme recycling was evaluated and Novozym^®435 was able to be applied for at least 20 cycles. Results demonstrated the efficiency of acylation reactions catalyzed by this enzyme and showed the potential of the methodology for the production and purification of flavor esters.

Keywords: citronellol; citronellyl esters; biocatalysis; trans/esterification; Novozym435^®

Introduction

Terpenes are secondary metabolites found in the essential oils of various plants. This class of bioactive compounds is widely used as components of flavors and fragrances in the cosmetics industry and personal care formulations,¹,² and in the food and pharmaceutical industries.³-⁵ Among the various monoterpenes of industrial and biotechnological importance, citronellol and its esters have gained prominence in recent years. Present in the essential oil of plants of the Cymbopogon genus, these compounds have long-standing organoleptic properties similar to the rose.⁶ They are a flavoring agent in perfumes, foods, and drinks to obtain fruity-floral nuances.⁷,⁸

Terpene esters are widespread, the most common being ethyl derivatives, which are highly volatile and degradable when exposed to extreme temperature and humidity conditions. In addition, they have a low yield per mass of plant raw material, making their extraction difficult for industrial-scale processes and raising the price of the products that use them.⁹,¹⁰

To meet the demand for terpene esters, most of these compounds have been obtained by traditional chemical routes.¹¹ However, the protocols applied have limitations such as reagents, catalysts, and solvents, where extreme temperature and pressure conditions are required, generating various by-products and the need for subsequent exhaustive purification steps,¹² losing their natural classification. According to American¹³ and European¹⁴ legislation, a natural flavor includes only that obtained from natural sources, including direct extraction from plants or those obtained biocatalytically from substrates isolated from nature.

Biocatalysis has become essential for obtaining esters with high regio-, chemoand enantioselectivity and under mild reaction conditions.¹⁵,¹⁶ In this trend, lipases (triacylglycerol hydrolases, EC 3.1.1.3) are the most popular enzymes for this type of reaction, including for synthesizing monoterpene esters.¹⁷ These enzymes naturally catalyze the hydrolysis of triacylglycerols in aqueous media. However, in water-restricted environments, lipases can also carry out the reverse reaction.¹⁸ In addition, these enzymes show high activity in organic solvents, accept a wide range of substrates, and catalyze esterification, transesterification, amidation, and kinetic resolution reactions with high enantio-, regioand chemoselectivities without the need for co-factors.¹⁹ These characteristics make lipases one of the most widely used enzymes in the biosynthesis of various industrially essential compounds. Some lipases have a hydrophobic lid that covers the catalytic site. To express activity, they need to perform the phenomenon of interfacial activation, wherein the presence of an emulsion, the lid is distorted, and the catalytic site is exposed.¹⁹,²⁰ Some lipases have a small lid and do not need this phenomenon. In contrast, others may not have a lid and are active in different environments, such as lipase B from Pseudozyma antarctica (Cal B), explored in this work. To be active in extreme environments of temperature, water deficiency, or even polarity, some lipases have been applied in their immobilized form, where the association with the support can occur by various mechanisms, resulting in greater retention of catalytic activities, stability and the possibility of reuse in various protocols.²¹ Immobilization also makes it possible to apply lipases that have a lid, since they can be immobilized in a distorted conformation.²²

Considering the need to investigate new systems for obtaining citronellol esters, this study investigated the comparison of obtaining compounds derived from citronellol using different acyl donors with linear carbon chains (ranging from C₃ to C₁₈) in esterification (citronellol and fatty acids) and transesterification (fatty acid esters) reactions, using immobilized lipase B from Pseudozyma antarctica (Novozym^®435) as a biocatalyst, with subsequent kinetic study and structural elucidation.

Experimental

Chemicals

Citronellol (95% purity), acylating agents ((propionic, butyric, valeric, hexanoic, octanoic, lauric, myristic, stearic, and oleic acids) and (vinyl propionate and vinyl laurate)), and 3 Å molecular sieves were purchased from Sigma Aldrich (São Paulo, Brazil). The solvent n-heptane (reagent grade) was purchased from Isofar (Rio de Janeiro, Brazil), while acetonitrile, hexane and ethyl acetate were purchased from Tedia (Fairfield, USA).

Immobilized Pseudozyma (Candida) antarctica lipase B (Novozym^®435, N435) was purchased from Novozymes^® (Araucária, Brazil). It is a commercial preparation containing lipase B from Pseudozyma (Candida) antarctica immobilized in cationic ion exchange acrylic resin (immobilization by ion adsorption), which presents a final activity of 5 U mg^-1.

General acylation procedures

Esterification reactions were conducted using propionic, butyric, valeric, hexanoic, octanoic, lauric, myristic, stearic, and oleic acids as acyl donors while transesterification reactions were conducted with vinyl propionate and vinyl laurate. Each reaction, according to Figure 1, was conducted in triplicate, in a conical flask from the mixture of citronellol and the acyl donor, in a 1:1 molar ratio, 100 mM (15.67 g L^-1) of citronellol in n-heptane, 3 Å molecular sieves (10 mg, which corresponds to 5 spheres), and 5 U of commercial lipase Novozym^®435 per mL of reaction, under 150 rpm of orbital stirring at 70 °C. Samples were taken until 6 h of reaction.

Figure 1
Reactional scheme of citronellol enzymatic acylation.

Ester conversion was determined concerning citronellol decay by gas chromatography-mass spectrometry (GC MS), model GC-QP2010 SE (Shimadzu, Barueri, Brazil) equipped with a DB-5MS column (30.0 m × 0.25 mm, 0.25 μm film thickness), from Agilent Technologies (Barueri, Brazil). Injector temperature was set at 270 °C, with helium (99.999%) as carrier gas at a flow rate of 1 mL min^-1, and 1 µL injection sample (diluted 1:7 in acetonitrile) at a split ratio of 50. The column temperature was held at 100 °C for 1 min, increased to 300 °C at 10 °C min^-1, and kept for 2 min, totalizing a 23 min run. The equipment was operated by electron impact ionization mode at 70 eV and scanned at the range of 40 to 500 m/z. The mass spectrum ion source temperature was 260 °C, the interface temperature was 300 °C, and the identification started after 3 min of the run.

Purification and structural elucidation of the esterified products

In order to obtain higher amounts of monoterpene esters, the best reaction time was selected by GC-MS analysis and repeated scaling up three times, adopting higher proportional amounts of substrates and enzymes, by the same operational conditions described above. Products purification was performed by preparative thin-layer chromatography (25 cm × 25 cm silica gel plate with CaSO₄ (254 UV), Macherey-Nagel, Germany), using hexane:ethyl acetate (9:1) as eluent.

The solvent was evaporated, and the products were analyzed by GC-MS. Subsequently, the esters structures were elucidated by ¹H and ¹³C nuclear magnetic resonance (NMR), recorded at 500 and 150 MHz, respectively, using Avance II HD spectrometer (Bruker Co., Switzerland) with tetramethylsilane (TMS) as an internal reference standard. The chemical shift (δ) is given in ppm. The multiplicity is indicated by the following abbreviations: s (singlet), d (doublet), t (triplet), dd (double of doublets), ddd (doublet of doublet of doublet), dddd (doublet of doublet of doublet of doublet), ddt (double of doublets of triplet), td (triplet of doublet), m (multiplet), q (quartet). All coupling constants (J values) were given in Hz.

Enzyme recycling

The transesterification reaction of citronellol and vinyl propionate was selected for enzyme recycling analysis. The reaction was executed in triplicate at the same conditions described above, by 30 min time. After each batch, aliquots were taken from the reactional medium and analyzed by GC-MS to determine the conversion rate. The immobilized enzymes were washed with 30 mL of heptane (3 × 10 mL) and then dried for further reuse.

Statistical analysis

Statistical analysis was performed using GraphPad Prism version 7.0 for Windows.²³ Kinetic parameters were determined from the non-linear regression with the application of the Michaelis-Menten model, while the differences between the samples were compared by one-way analysis of variance (ANOVA) and Tukey’s multiple comparison test, with 95% confidence (α = 0.05).

Results

Kinetic of citronellyl esters production by esterification and transesterification

We began this work by evaluating the production of citronellol esters via esterification of alcohol with different fatty acids, catalyzed by lipase N435 up to 6 h reaction time. The main goal of this step was to investigate the influence of the acylation agent on the ester products through a kinetic study. Figure 2 shows that up to 6 h reaction all esters seemed to achieve conversions higher than 95%. More precisely, in a 2 h reaction time, ester conversions varied from about 71% to about 97%, being the last one correspondent to citronellyl butyrate, followed by citronellyl hexanoate, as the second higher conversion, with about 96%. Both showed a plateau after a 2 h reaction time, with higher than 98% conversion. In addition, it is also important to highlight those kinetic parameters such as V₀ (initial velociy); V_max (maximal velocity) and K_m (Michaelis-Menten constant), which were determined for each reaction with the non-linear regression of the data, with the application of the Michaelis-Menten model (Supplementary Information (SI) section, Table S1). It was possible to observe that although lipase N435 catalyzed all the reactions with longand short-chain fatty acids, the lowest K_m values were obtained for long-chain acyl donors. This reflects the fact that Pseudozyma antarctica lipase B (CaLB) has a greater affinity for long-chain fatty acids, both for hydrolysis and synthesis reactions, as well as having esterase activity, where it can accommodate short-chain fatty acids in its catalytic site. In addition, the commercial preparation N435 contains the CaLB lipase adsorbed to the surface in a stable and distorted form, where its small lid is in its open conformation, with no need for interfacial activation to exert activity, corroborating the results obtained so far. Concerning the volumetric production, it is possible to check in Figure S1 (SI section) that the esterification method provided a minimum of 15 g L^-1 in esterified products in 2 h of reaction, with all the acyl donors tested, reaching over 35 g L^-1 in case of stearate and oleate esters, composed by the higher molecular weight acyl donors. They were followed by myristate and laurate, which achieved close to 35 and almost 30 g L^-1 production, respectively.

Figure 2
Kinetic of citronellyl ester production by the esterification reaction (molar ratio 1:1, 100 mM of citronellol, 5 U mL^-1 of Novozym^®435, heptane as a solvent, 150 rpm, and 70 °C). Reactions were conducted in triplicate, and the average values are shown. Reactions without enzymes did not result in ester products.

To observe the kinetic profile in transesterification reactions, the product obtained with the lowest K_m (citronellyl propionate) and highest K_m (citronellyl laurate) in the esterification step was investigated for obtaining via transesterification, with vinyl propionate and vinyl laurate as acyl donors (Figure 3).

Figure 3
Conversion (%) to citronellyl esters by transesterification reaction (molar ratio 1:1, 100 mM of citronellol, 5 U mL^-1 of Novozym^®435, heptane as a solvent, 150 rpm, and 70 °C). Reactions were conducted in triplicate, and the average values are shown. Reactions without enzymes did not result in ester products.

As a result, in comparison with the 2 h esterification reaction, the transesterification protocol resulted in the same ester conversion rate (%), however, only in 5 and 8 min reaction time for propionate (approximately 70%) and laurate (approximately 80%), respectively (Figure 4, Table S4, SI section). Comparing the ester production, it is possible to see that transesterification takes place much faster than esterification. As an example, we highlight citronellyl propionate production, which was obtained about 10 g L^-1 of the product after 1.5 h of esterification (Figure S1, SI section) compared to 2 min by transesterification (Figure S2, SI section). We also realize that for citronellyl laurate almost 25 g L^-1 of the product was formed after 1.5 h of esterification (Figure S1) compared to 5-8 min by transesterification (Figure S2). In addition, it is possible to see that the initial rate of this transesterification reaction is almost 10 folds higher than esterification (Table S1).

Figure 4
Citronellyl propionate chemical structure and HMBC spectra (¹H NMR 500 MHz, CDCl₃ and ¹³C NMR 150 MHz, CDCl₃).

Purification and products structural elucidation

In order to demonstrate a simple protocol for purifying and separating the esters obtained in this work, the samples for the 9 esters previously obtained by esterification and transesterification were subjected to preparative thin-layer chromatography (TLC) on silica gel, with subsequent recovery of the products by filtration and submission to structural elucidation by ¹³C and ¹H NMR. In general, products were obtained with over 95% purity, except for citronellyl oleate, which presented 90% purity (Table 1) showing that the purification protocol adopted was successful. The preparative TLC is an efficient method of purification and separation since TLC contains a high number of theoretical plates, promoting a good separation of the substances, and being also a fast, stable, and cost-efficient chromatographic separation technique.²⁴,²⁵

Table 1
Citronellyl esters purity and their structural elucidation

Substance	Puritya / %	m/z b	1H NMR (500 MHz, CDCl3)
Citronellyl propionate	95	253, 213, 183, 155, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.84 [(d, J 6.6 Hz, 1H) 8], 1.07 [(t, 7.6 Hz, 1H) 3’], 1.37 [(ddd, 13.6, 7.4, 3.9 Hz, 1H) 4], 1.48 [(dd, 13.2, 6.6 Hz, 1H) 5], 1.53 [(s, 3H) 9,10], [1.63-1.57 (m, 4H) 6], 1.91 [(td, 15.6, 7.1 Hz, 2H) 3], 2.25 [(q, 7.6 Hz, 2H) 2’], 4.08-3.99 [(m, 2H) 7], 5.02 [(t, 6.9 Hz, 1H) 2]
Citronellyl butyrate	95	282, 253, 207, 183, 155, 153, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.91 [(d, 6.6 Hz, 3H) 8], 0.94 [(t, 7.4 Hz, 3H) 4’], 1.38-130 [(m, 1H) 4], 1.43 [(dt, 13.7, 6.7 Hz, 1H) 5], 1.54 [(dd, 13.1, 6.5 Hz, 1H) 6], 1.60 [(s, 3H) 9,10], 1.69-1.61 [(m, 6H) 3’], 1.97 [(td, 15.2, 7.1 Hz, 2H) 3], 2.27 [(t, 7.4 Hz, 2H) 2’], 4.15-4.05 [(m, 2H) 7], 5.08 [(t, 7.0 Hz, 1H) 2]
Citronellyl valerate	98	377, 281, 253, 207, 183, 167, 155, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.90 [(dd, J 7.0, 2.0 Hz, 4H) 8], 0.92 [(s, 1H) 5’], 1.17 [(dddd, J 13.5, 9.4, 7.7, 6.0 Hz, 1H) 4’], 1.36-131 [(m, 1H) 4], 1.42 [(dd, J 13.8, 6.3 Hz, 1H) 5], 1.59 [(s, 1H) 9,10], 1.65-1.62 [(m, 1H) 3’], 1.67 [(d, J 5.2 Hz, 1H) 6], 1.97 [(td, J 15.7, 7.1 Hz, 1H) 3], 2.26-2.30 [(m, 1H) 2’], 4.14‑4.04 [(m, 1H) 7], 5.08 [(t, J 7.0 Hz, 1H) 2]
Citronellyl hexanoate	98	390, 355, 281, 255, 211, 181, 155, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.89 [(d, J 7.1 Hz, 2H) 8], 0.90-0.92 [(m, 3H) 6’], 1.21-1.15 [(m, 1H) 4’], 1.30 [(dt, J 11.8, 3.8 Hz, 3H) 5’], 1.44 [(dt, J 13.7, 6.7 Hz, 1H) 4], 1.55 [(dd, J 13.1, 6.5 Hz, 1H) 5], 1.60 [(s, 3H) 9,10], 1.67-1.61 [(m, 2H) 3’], 1.68 [(t, J 2.5 Hz, 3H) 6], 1.98 [(td, J 15.6, 7.1 Hz, 2H) 3], 2.28 [(t, J 7.5 Hz, 2H) 2’], 4.13-4.07 [(m, 2H) 7], 5.09 [(t, J 7.0 Hz, 1H) 2]
Citronellyl octanoate	98	403, 370, 283, 239, 209, 171, 155, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.88 [(t, J 7.0 Hz, 3H) 8’], 0.91 [(d, J 6.6 Hz, 3H) 8], [1.16 (dt, J 9.4, 3.9 Hz, 1H), 1.21 [(dd, J 5.5, 3.8 Hz, 1H), 1.27 (d, J 2.3 Hz, 2H), 1.29 (dd, J 6.5, 4.0 Hz, 4H) 4’-7’], 1.36 [(dd, J 3.7, 2.2 Hz, 1H) 4], 1.43 [(ddd, J 13.6, 10.5, 6.7 Hz, 1H) 5], 1.54 [(dt, J 11.9, 5.9 Hz, 1H) 3’], 1.60 [(s, 3H) 9,10], 1.68 [(t, J 2.6 Hz, 4H) 6], 1.98 [(td, J 15.9, 7.1 Hz, 2H) 3], 2.28 [(t, J 7.6 Hz, 2H) 2’], 4.14-4.05 [(m, 2H) 7], 5.09 [(t, J 7.1 Hz, 1H) 2]
Citronellyl laurate	98	419, 402, 339, 281, 265, 227, 202, 183, 163, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.88 [(t, J 6.9 Hz, 2H) 12’], 0.91 [(d, J 6.6 Hz, 2H) 8], [1.16 (dd, J 5.6, 3.8 Hz, 1H), 1.19-1.17 (m, 1H), 1.21 (dd, J 3.3, 1.5 Hz, 1H), 1.28-1.24 (m, 9H) 6’-10’], 1.29 [(d, J 4.7 Hz, 3H) 4], [1.34-1.32 (m, 1H), 1.36-1.35 (m, 1H), 1.38 (d, J 6.3 Hz, 1H) 4’,5’,11’], 1.42 [(ddd, J 13.6, 7.6, 6.3 Hz, 1H) 5], 1.53 [(dt, J 19.6, 6.5 Hz, 1H) 3’], 1.60 [(s, 3H) 9,10], 1.68 [(t, J 2.6 Hz, 3H) 6], 1.98 [(td, J 15.9, 7.1 Hz, 2H) 3], 2.28 [(t, J 7.6 Hz, 2H) 2’], 4.14-4.05 [(m, 2H) 7], 5.09 [(t, J 7.1 Hz, 1H) 2]
Citronellyl myristate	99	429, 405, 367, 293, 255, 229, 211, 191, 155, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.87 [(t, J 6.9 Hz, 3H) 14’], 0.90 [(d, J 6.6 Hz, 3H) 8], [1.16-1.13 (m, 1H), 1.18-1.16 (m, 1H), 1.20-1.18 (m, 1H), 1.24-1.22 (m, 1H), 1.27‑1.23 (m, 16H), 1.30 (dd, J 12.4, 5.6 Hz, 6H) 4’-13’], 1.37-1.34 [(m, 1H) 4], 1.43 [(dt, J 13.7, 6.8 Hz, 1H) 5], 1.53 [(dd, J 4.5, 3.2 Hz, 1H) 3’], 1.59 [(s, 3H) 9,10], 1.67 [(dd, J 4.5, 1.2 Hz, 4H) 6], 1.97 [(td, J 15.8, 6.8 Hz, 2H) 3, 2.27 [(t, J 7.5 Hz, 2H) 2’], 4.13-4.04 [(m, 2H) 7], 5.08 [(t, J 7.0 Hz, 1H) 2]
Citronellyl stearate	99	476, 430, 415, 355, 281, 265, 252, 227, 211, 183, 163, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.88 [(t, J 6.9 Hz, 2H) 18’], 0.91 [(d, J 6.6 Hz, 2H) 8], [1.17-1.14 (m, 1H), 1.19 (ddt, J 9.8, 4.1, 2.0 Hz, 1H), 1.30-1.21 (m, 1H), 1.35-1.33 (m, 1H), 1.37-1.35 (m, 1H), 1.40 (dd, J 10.3, 4.3 Hz, 1H), 1.43 (dd, J 6.1, 1.5 Hz, 1H) 4’-17’], 1.45 [(dd, J 5.3, 4.2 Hz, 1H) 4], 1.54 [(dt, J 6.5, 6.0 Hz, 1H) 5], 1.59 [(d, J 5.8 Hz, 1H) 3’], 1.60 [(s, 3H) 9,10], 1.68 [(dd, J 3.9, 1.5 Hz, 3H) 6], 1.98 [(td, J 15.6, 7.1 Hz, 2H) 3], 2.28 [(t, J 7.5 Hz, 2H) 2’], 4.13-4.06 [(m, 2H) 7], 5.09 [(t, J 6.9 Hz, 1H) 2]
Citronellyl oleate	90	475, 420, 347, 309, 280, 265, 247, 221, 207, 179, 163, 138, 123, 109, 95, 82, 81, 69, 67, 57, 55, 41	δ 0.88 [(t, J 6.8 Hz, 2H) 18’], 0.91 [(d, J 6.6 Hz, 2H) 8], 1.21-1.15 [(m, 1H) 4], [1.24-1.21 (m, 1H), 1.25 (dd, J 6.9, 4.8 Hz, 6H), 1.33-1.28 (m, 8H), 1.41-1.37 (m, 1H), 1.43 (ddd, J 13.6, 7.3, 3.8 Hz, 1H), 1.53 (dt, J 13.1, 6.5 Hz, 1H) 4’-7’, 12’-17’, 16’, 17’], 1.59 [(s, 3H) 9,10], 1.60 [(t, J 4.0 Hz, 3H) 5], 1.66-1.62 [(m, 1H) 3’], 1.70-1.66 [(m, 3H) 6], 1.94 [(dd, J 15.4, 8.1 Hz, 1H) 8’,11’], 2.01 [(dd, J 12.8, 6.8 Hz, 3H) 3], 2.28 [(t, J 7.5 Hz, 1H) 2’], 4.14-4.05 [(m, 2H) 7], 5.08 [(t, J 7.1 Hz, 1H) 2], 5.34 [(dt, J 5.8, 3.6 Hz, 1H) 9’,10’]

• ^a
  Determined by GC-MS from the analysis of the peak area;
• ^b
  determined by GC-MS with 70 eV of MS source ionization energy.

The products were also analyzed by GC-MS for mass fragmentation analysis (Table 1) and further comparison of their mass spectra with the NIST library (NIST 05). The results confirmed that all products corresponded to the previewed citronellyl ester derivatives.

To confirm the structure of citronellyl esters, nuclear magnetic resonance mono and bidimensional analysis have proceeded. In the ¹H NMR spectra, there were signals in the region of δ 0.84-5.09 ppm for all hydrogen atoms (Table 1) related to methyl, methylene, and methine hydrogens corresponding to the citronellyl portion, and methyl and methylene hydrogens from the acyl skeleton. In addition to those, the sign in δ 5.34 ppm could be seen in the citronellyl oleate spectrum related to the unsaturation of the acyl portion, which sets it apart from other straight-chain esters. From the ¹³C NMR spectrum, there were visualized signals between δ 9 and 175 ppm, highlighting a sign in the region of δ 175 ppm related to the carbon from the carbonyl of the ester.

From the heteronuclear multiple bond correlation (HMBC) spectra (Figure 4), it was possible to confirm that all the products were esters, which was established by a correlation between the carbonyl carbon from the acyl portion (position 1’ in Figure 4) and the methylene hydrogen from the citronellyl portion (position 7 in Figure 4), as shown for citronellyl propionate. This analysis confirmed that all the products were citronellyl esters.

Enzyme recycling

Many factors involved in the bioproduction of esters catalyzed by lipases are relevant from an industrial point of view, especially when considering the recycling of the immobilized enzyme used in the process. Stability, enzyme leaching, and maintenance of activity are primordial factors that influence this choice. The enzyme recycling assay performed in this work on the production of citronellyl propionate could infer that this enzyme can be successfully used for more than 20 cycles (Figure 5), with the preservation of conversion to the desired product without a significant difference (α = 0.05) between the batches (Table S23, SI section).

Figure 5
Recycling study of Novozym^®435 on the production of citronellyl propionate by transesterification reaction (molar ratio 1:1, 100 mM of citronellol, 5 U mL^-1 of Novozym^®435, heptane as a solvent, 150 rpm, and 70 °C). Reactions were conducted in triplicate, and the average values are shown. Reactions without enzymes did not result in ester products.

Discussion

In this work, the acylation of citronellol catalyzed by lipase N435 through esterification and transesterification with different acyl donors was investigated, where the effect of the acyl donor and the process conditions were analyzed and discussed. In general, the two strategies studied generated conversions of more than 95% in esters in up to 6 h of reaction (Figure 2), with high conversion rates. Regarding volumetric production (Figure S1), values were generally higher than a minimum of 15 g L^-1 of acylated products in 2 h of reaction. However, citronellyl propionate exhibited lower productivity over time, as well as citronellyl butyrate, which presented the second higher V_max, higher initial rate (V₀), and conversion (%). A similar behavior was observed by Al-Saadi et al.,²⁶ in the production of biodiesel, where they obtained 95% of the esterified product in 3 h of transesterification and 71.4% of the product in 6 h of esterification, using SrOeZnO/Al₂O₃ as a catalyst. In addition, Melchiorre et al.,²⁷ in the synthesis of benzyl levulinate using iron(III) catalyst, achieved 97% of the transesterification product in 18 h compared to 82% in esterification. This behavior can be explained by the interference of the size of the fatty acid carbon chain in the reaction. Short-chain fatty acids (smaller than 5 carbons) have a smaller molecular volume, which can interact more easily with the catalytic site of the lipases, so the esterification reaction can take place more quickly.²⁸ However, it is known that CaLB lipase acts by the Ping Pong Bibi mechanism, where a group is transferred by the enzyme from an electron donor to an electron acceptor. Its active site, like a standard α,β-hydrolase, is formed by a catalytic triad of Ser-His-Asp amino acids.²⁹

Initially, the carbon atom of the carbonyl group of the substrate acid is attacked by the oxygen of the hydroxyl group of serine, which has received electrons from the amino acid histidine, present in the hydrophobic lid of the lipase.³⁰ By transforming the C=O bond of the substrate into a single bond, the oxygen atom acquires a negative charge. An intermediate transition complex with a tetrahedral geometry is thus formed around this carbon atom through hydrogen bonds between the oxyanion cavity of the enzyme and two NH groups on the main chains. As a result of the formation of this complex, a proton is transferred from serine to histidine. Aspartate, which is deprotonated, partially neutralizes the charge that develops in the transition complex.³¹ As the protonated form of histidine loses the proton, the carbonyl bond is repaired and the tetrahedral intermediate is broken up, releasing water as the first product. The alkyl chain of the first substrate remains covalently linked to the serine, forming the enzyme acyl intermediate, and completing the first stage of acylation. In the subsequent step, deacylation occurs similarly, but in reverse order. The alcohol nucleophilically attacks the enzyme acyl intermediate, forming once again, with the help of histidine, a new tetrahedral intermediate that collapses releasing the second product, the ester. Consequently, the enzyme returns to its initial state.³² Short-chain fatty acids can compete with the alcohol in the second reaction step, forming dead-end complexes and displacing the lipase, which is a type of competitive inhibition.²⁸ It is also possible, initially, that the size of the carbon chain increases the speed of the esterification reaction, and consequently the formation of products, and the latter, when they acquire critical concentrations in the system, also cause inhibitions in the enzyme, displacing the biocatalyst and reducing the maximum reaction speeds. Figures S1 and S2 also show that increasing the size of the carbon chains had a general influence on the conversion rates into citronellyl esters. Statistical analysis showed a significant difference (α = 0.05) as the carbon chain increased (Table S4, SI section), which can be explained by the difference in density, viscosity, evaporation rate and unsaturation in the acyl donor. Similar results were obtained by Kroutil et al.,³³ who observed that increasing the chain length and adding unsaturation to the acyl donor caused a decrease in the biocatalytic esterification of streptol using Candida rugosa lipase as a biocatalyst. In addition, the bond angles of the acyl-enzyme intermediate can be altered due to the increase in the acyl donor portion, which makes bonding with citronellol in the active site difficult. The reactions carried out with acyl donors with a carbon chain between 3 and 18 atoms showed a difference in conversion rates, showing that this enzyme can catalyze reactions with different acylating agents, but increasing the substrate’s carbon chain compromises its active site and catalytic activity. An additional and relevant factor is the competitive secondary hydrolysis reaction that occurs in esterification, as water can act as a nucleophile for the hydrolysis of the ester, producing the alcohol and acid initially used as a substrate, as shown in Figure 5. Therefore, at critical water concentrations, there is a competition between hydrolysis and esterification, which affects the reaction rate, equilibrium and final conversions. There are a limited number of publications on the biocatalytic production of citronellyl esters from direct acylation. One of them is described by Lozano et al.,³⁴ who carried out reactions with ionic liquid and lipase B from P. antarctica (Novozym^®435) as a biocatalyst to obtain citronellyl acetate, butyrate and valerate in 4 h of reaction. Melo et al.¹¹ worked on the production of citronellyl butyrate and valerate using a lipase from Rhizopus sp. and obtained up to 81% conversion in 48 h of reaction. Another study was carried out by Abdullah¹⁰ on the synthesis of citronellyl laurate using immobilized lipase from Candida rugosa. Corrêa et al.³⁵ carried out reactions between citronellol and stearic and oleic acids to obtain esters using immobilized lipases from P. antarctica (Novozym^®435) and Thermomyces lanuginosus (Lipozyme^®TLIM) with conversion rates of 93.9 and 89.9% for citronellyl oleate, respectively.

In order to investigate other mechanisms for obtaining citronellol esters and understand the performance of the N435 lipase, transesterification tests were also carried out to obtain citronellyl propionate and laurate, with corresponding vinyl esters, since these products exhibited the highest and lowest V_max, respectively, in the esterification step, as well as representing derivatives of small (3C) and large (12C) carbon chains. In this type of reaction, water is not formed as a reaction product, but rather the corresponding alcohol, which also affects the equilibrium of the system. As a result, higher rates of ester formation were observed than in the esterification stage, which can be explained by the characteristic keto-enolic tautomerism of the vinyl radical: the tautomerization reaction of the vinyl alcohol, obtained as a secondary product of transesterification, contributes to an equilibrium that tends to form an aldehyde,³⁶-³⁸ making transesterification irreversible. As no water is formed in the system, there is no possibility of hydrolysis in this type of reaction. Although in this work we used molecular sieves to adsorb any water that might be present in the system, previously published work by Xiong et al.³⁹ shows that, although water plays a fundamental role in maintaining the spatial conformation of the lipase, by non-covalent forces, greater amounts of water present in the system during transesterification end up reducing reaction speeds by stimulating the formation of other non-active lipase conformations, displacing the reaction, or by causing an adsorption effect on the enzyme surface, which limits enzyme-substrate interaction. Even though there is a higher reaction speed compared to esterification, the formation of citronellol esters by transesterification also follows a double ping-pong reaction mechanism inhibited by the end product. Still, the authors³⁹ demonstrated that, citronellyl acetate, vinyl acetate and the free enzyme complex form a non-covalent enzyme-ester complex, which is converted into the acyl-enzyme intermediate after isomerization. At the same time, the first product and the enol are released to produce the modified enzyme. Citronellol reacts with the activated enzyme to produce another complex, which produces the ester-enzyme complex after isomerization. The enzyme-product complex finally decomposes into citronellyl acetate and the free enzyme.

Concerning the recycling of the enzymes, it was observed that the system showed operational stability, generating conversions of between 90 and 95% on average during the reuses, indicating that the enzyme probably did not leach from the medium or lose activity during the processes. At the end of the cycles, an average activity of 4.9 U mg^-1 was found, demonstrating that the enzyme remains active. This result is in line with those published by Marín-Suárez et al.,⁴⁰ regarding the recycling of Novozym^®435 for more than ten cycles (35 °C and 8 h) in the transesterification reaction of fish oil used to produce biodiesel. In addition, Talukder et al.⁴¹ determined that Novozym^®435 can perform more than 15 cycles (50 °C and 2 h) in the esterification reaction to produce biodiesel. However, this was higher than that described by Yadav and Lathi,⁴² who observed a 20% loss in activity in the third reaction cycle (30 °C and 15 min) when using the enzyme as a catalyst for the transesterification of methyl acetoacetate with n-butanol. This can be explained by the greater thermostability and the low loss of enzyme from the support material in organic solvents.⁴³-⁴⁵

Conclusions

Citronellyl esters were synthesized with good experimental yield, showing over 15 g L^-1 in the esterified products in 2 h of esterification or 5 min of transesterification, with different linear acylating agents, showing that transesterification is preferable when the time is evaluated. Carbon chain size had a significant influence on the biocatalytic synthesis of citronellyl esters, in view that the increase in the acyl donor carbon chain leads to a decrease in ester production. The reaction products were purified with high purity (≥ 90%), characterized, and structurally elucidated as citronellyl esters. The enzyme recycling assay showed that the same amount of enzyme can be used for more than 20 cycles of transesterification reaction (at the studied conditions) and still maintain the catalytic power.

Supplementary Information

Supplementary data are available free of charge at http://jbcs.sbq.org.br as PDF file.

Acknowledgments

The authors are thankful for the support from the “Central Analítica do DPNA”, Faculty of Pharmacy, Federal University of Rio de Janeiro.

References

• ¹ Sá, A. G. A.; de Meneses, A. C.; de Araújo, P. H. H.; de Oliveira, D.; Trends Food Sci. Technol. 2017, 69, 95. [Crossref]
  » Crossref
• ² Akacha, N. B.; Gargouri, M.; Food Bioprod. Process. 2015, 94, 675. [Crossref]
  » Crossref
• ³ Yadav, G. D.; Lathi, P. S.; J. Mol. Catal. B: Enzym. 2004, 27, 113. [Crossref]
  » Crossref
• ⁴ Dhake, K. P.; Deshmukh, K. M.; Patil, Y. P.; Singhal, R. S.; Bhanage, B. M.; J. Biotechnol. 2011, 156, 46. [Crossref]
  » Crossref
• ⁵ Serri, N. A.; Kamaruddin, A. H.; Long, W. S.; Bioprocess Biosyst. Eng. 2006, 29, 253. [Crossref]
  » Crossref
• ⁶ Santos, P. L.; Matos, J. P. S. C. F.; Picot, L.; Almeida, J. R. G. S.; Quintans, J. S. S.; Quintans-Júnior, L. J.; Food Chem. Toxicol. 2019, 123, 459. [Crossref]
  » Crossref
• ⁷ Mohamad, N. R.; Buang, N. A.; Mahat, N. A.; Lok, Y. Y.; Huyop, F.; Aboul-Enein, H. Y.; Abdul Wahab, R.; Enzyme Microb. Technol. 2015, 72, 49. [Crossref]
  » Crossref
• ⁸ Li, Y.; Sun, R.; Kong, Y.; Cai, X.; Jiang, T.; Cheng, S.; Yang, H.; Song, L.; Lü, X.; Wang, X.; Shi, C.; Ultrason. Sonochem 2024, 102, 106752. [Crossref]
  » Crossref
• ⁹ Liaquat, M.; Apenten, R. K. O.; J. Food Sci. 2000, 65, 295. [Crossref]
  » Crossref
• ¹⁰ Abdullah, A. Z.; Sulaiman, N. S.; Kamaruddin, A. H.; Biochem. Eng. J. 2009, 44, 263. [Crossref]
  » Crossref
• ¹¹ Melo, L. L. M. M.; Pastore, G. M.; Macedo, G. A.; Process Biochem. 2005, 40, 3181. [Crossref]
  » Crossref
• ¹² You, P.; Su, E.; Yang, X.; Mao, D.; Wei, D.; J. Mol. Catal. B: Enzym. 2011, 71, 152. [Crossref]
  » Crossref
• ¹³ United States Food and Drug Administration (FDA), US Code of Federal Regulations, 21 CFR, 101.22, https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?CFRPart=101, accessed in September 2024.
  » https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?CFRPart=101
• ¹⁴ Europe - Council Directive, The Council of the European Communities, 88/388, https://eur-lex.europa.eu/legal-content/PT/TXT/HTML/?uri=CELEX:31988L0388, accessed in September 2024.
  » https://eur-lex.europa.eu/legal-content/PT/TXT/HTML/?uri=CELEX:31988L0388
• ¹⁵ Pellis, A.; Cantone, S.; Ebert, C.; Gardossi, L.; New Biotechnol. 2018, 40, 154. [Crossref]
  » Crossref
• ¹⁶ Fryszkowska, A.; Devine, P. N.; Curr. Opin. Chem. Biol. 2020, 55, 151. [Crossref]
  » Crossref
• ¹⁷ Staudt, A.; Brack, Y.; Itabaiana Jr, I.; Leal, I. C. R.; Mol. Catal. 2022, 528, 112464. [Crossref]
  » Crossref
• ¹⁸ Damnjanović, J. J.; Žuža, M. G.; Savanović, J. K.; Bezbradica, D. I.; Mijin, D. Ž.; Bošković Vragolović, N.; Knežević-Jugović, Z. D.; J. Mol. Catal. B: Enzym. 2012, 75, 50. [Crossref]
  » Crossref
• ¹⁹ Silva, N. C. A.; Miranda, J. S.; Bolina, I. C. A.; Silva, W. C.; Hirata, D. B.; de Castro, H. F.; Mendes, A. A.; Biochem. Eng. J. 2014, 82, 139. [Crossref]
  » Crossref
• ²⁰ do Nascimento, M. A.; Leão, R. A. C.; Froidevaux, R.; Wojcieszak, R.; de Souza, R. O. M. A.; Itabaiana, I.; Biochem. Eng. J. 2023, 195, 108915. [Crossref]
  » Crossref
• ²¹ Punekar, N. S.; Enzymes: Catalysis, Kinetics and Mechanisms; Springer: Singapore, 2018. [Crossref]
  » Crossref
• ²² Lopresto, C. G.; Calabrò, V.; Woodley, J. M.; Tufvesson, P.; J. Mol. Catal. B: Enzym 2014, 110, 64. [Crossref]
  » Crossref
• ²³ GraphPad Prism, version 7.0; GraphPad Software, Boston, Massachusetts, USA, 2016.
• ²⁴ Sharma, R.; Chisti, Y.; Banerjee, U. C.; Biotechnol. Adv. 2001, 19, 627. [Crossref]
  » Crossref
• ²⁵ Claon, P. A.; Akoh, C. C.; Biotechnol. Lett. 1994, 16, 235. [Crossref]
  » Crossref
• ²⁶ Al-Saadi, A.; Mathan, B.; He, Y.; Renewable Energy 2020, 158, 388. [Crossref]
  » Crossref
• ²⁷ Melchiorre, M.; Amendola, R.; Benessere, V.; Cucciolito, M. E.; Ruffo, F.; Esposito, R.; Mol. Catal. 2020, 483, 110777. [Crossref]
  » Crossref
• ²⁸ Zhang, C.; Liang, X.; Abdo, A. A. A.; Kaddour, B.; Li, X.; Teng, C.; Wan, C.; Biotechnol. Biotechnol. Equip. 2021, 35, 255. [Crossref]
  » Crossref
• ²⁹ Athawale, V.; Manjrekar, N.; Athawale, M.; Tetrahedron Lett. 2002, 43, 4797. [Crossref]
  » Crossref
• ³⁰ Murcia, M. D.; Gómez, M.; Gómez, E.; Gómez, J. L.; Hidalgo, A. M.; Sánchez, A.; Vergara, P.; Chem. Eng. Res. Des. 2018, 138, 135. [Crossref]
  » Crossref
• ³¹ Liang, S.; Zhuang, W.; Wang, Z.; Process Biochem 2024, 141, 1. [Crossref]
  » Crossref
• ³² Remonatto, D.; Fantatto, R. R.; Pietro, R. C. L. R.; Monti, R.; Oliveira, J. V.; De Paula, A. V.; Bassan, J. C.; Process Biochem 2022, 120, 287. [Crossref]
  » Crossref
• ³³ Kroutil, W.; Hagmann, L.; Schuez, T. C.; Jungmann, V.; Pachlatko, J. P.; J. Mol. Catal. B: Enzym. 2005, 32, 247. [Crossref]
  » Crossref
• ³⁴ Lozano, P.; Bernal, J. M.; Navarro, A.; Green Chem 2012, 14, 3026. [Crossref]
  » Crossref
• ³⁵ Corrêa, L. S.; Henriques, R. O.; Rios, J. V.; Lerin, L. A.; de Oliveira, D.; Furigo, A.; Appl. Biochem. Biotechnol. 2020, 190, 574. [Crossref]
  » Crossref
• ³⁶ Cederstav, A. K.; Novak, B. M.; J. Am. Chem. Soc. 1994, 116, 4073. [Crossref]
  » Crossref
• ³⁷ Degueil-Castaing, M.; De Jeso, B.; Drouillard, S.; Maillard, B.; Tetrahedron Lett. 1987, 28, 953. [Crossref]
  » Crossref
• ³⁸ Pedersen, N. R.; Halling, P. J.; Pedersen, L. H.; Wimmer, R.; Mathiesen, R.; Veltman, O. R.; FEBS Lett. 2002, 519, 181. [Crossref]
  » Crossref
• ³⁹ Xiong, J.; Sun, W.; Xu, H.; Bian, J.; Song, Y.; Pan, S.; Gao, J.; Food Sci. Technol 2022, 42, e91922. [Crossref]
  » Crossref
• ⁴⁰ Marín-Suárez, M.; Méndez-Mateos, D.; Guadix, A.; Guadix, E. M.; Renewable Energy 2019, 140, 1. [Crossref]
  » Crossref
• ⁴¹ Talukder, M. M. R.; Wu, J. C.; Lau, S. K.; Cui, L. C.; Shimin, G.; Lim, A.; Energy Fuels 2009, 23, 1. [Crossref]
  » Crossref
• ⁴² Yadav, G. D.; Lathi, P. S.; J. Mol. Catal. B: Enzym. 2005, 32, 107. [Crossref]
  » Crossref
• ⁴³ Mathias, E. V.; Halkar, U. P.; J. Anal. Appl. Pyrolysis 2004, 71, 515. [Crossref]
  » Crossref
• ⁴⁴ Zhou, D.; Li, Y.; Huang, L.; Qian, M.; Li, D.; Sun, G.; Yang, B.; Food Chem. 2020, 303, 125399. [Crossref]
  » Crossref
• ⁴⁵ Hari Krishna, S.; Karanth, N. G.; Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 2001, 1547, 262. [Crossref]
  » Crossref

Edited by

• Editor handled this article:
  Albertina Moglioni (Associate)

Publication Dates

• Publication in this collection
  08 Nov 2024
• Date of issue
  2025

History

• Received
  02 Feb 2024
• Accepted
  04 Oct 2024