Modification of the properties of the metagenomic lipase LipC12 by engineering of the hydrophobic cavity

Abstract Kinetic resolution of racemates with lipases is the preferred method for producing bioactive compounds. One strategy for obtaining highly enantioselective lipases that are stable in the organic media that are often used in these reactions is to improve existing lipases by protein engineering. In this work, we engineered the lipase LipC12, which has good stability in organic media, but only moderate enantioselectivity. Molecular docking with LipC12 identified V261 as a key position influencing, first, enantioselectivity in the transesterification of (RS)-1-phenylethanol and, second, activity in the hydrolysis of p-nitrophenyl octanoate. Variants were then obtained by site-directed mutagenesis, expressed in Escherichia coli, and their performance in these reactions was evaluated. Enzymes immobilized on Immobead 150 were used in the transesterification while free enzyme was used in the hydrolysis reaction. It was not possible to increase the hydrolytic activity and enantioselectivity simultaneously: some variants had increased enantioselectivity but lower hydrolytic activity, and others had increased hydrolytic activity but lower enantioselectivity. The best result for enantioselectivity was obtained for LipC12V261Q, with an increase of the E-value (for (R)-1-phenylethanol) from 46 to 110, however, its hydrolytic activity decreased 4-fold in comparison to LipC12wt. The highest hydrolytic activity was obtained for LipC12V261F, with a value almost 6-fold higher than that of LipC12wt. This variant also had an inverted enantiopreference (i.e. for (S)-1-phenylethanol), but with a very low E-value of only 4.


Introduction
The production of pure enantiomers is of paramount importance for the pharmaceutical and chemical industries, where these compounds are used as chiral building blocks for the synthesis of bioactive drugs and as intermediates for organic synthesis, respectively (Ema 2004;Carvalho et al. 2015).High purity enantiomers can be produced via kinetic resolution of racemates using lipases (EC 3.1.1.3).This process has been used for the resolution of racemic alcohols, amines, esters, and carboxylic acids, through hydrolysis, esterification and transesterification reactions (Miranda et al. 2015;Javed et al. 2018).However, only a few highly enantioselective lipases are commercially available, the most representative ones being lipase B from Candida antarctica (CALB, Novozym 435), and the lipases from Pseudomonas cepacia (Amano lipase PS) and Pseudomonas fluorescens (Amano-P lipase; Carvalho et al. 2015;Miranda et al. 2015).This situation has stimulated researchers to search for new lipases with high enantioselectivity.
Although it is possible to search for new lipases by isolating new lipase-producing organisms or building and screening metagenomic libraries for lipolytic activity (Almeida et al. 2020), the chances of obtaining a lipase with the desired activity are small, such that very large numbers of isolates or clones need to be screened.This has prompted alternative approaches with existing lipases, such as the production of immobilization libraries, in which the properties of the immobilized lipase are varied by altering the immobilization conditions (Arana-Peña et al. 2021), and the use of protein engineering strategies, such as rational design and random mutagenesis, to improve enantioselectivity (Qin et al. 2013).
One promising existing lipase is LipC12, which was originally isolated from a metagenomic library prepared from fat-contaminated soil and is expressed in Escherichia coli.LipC12 has high activity against natural triacylglycerols (1800 U against olive oil) and shows good activity and stability over a wide pH range (4.5-11.0)and in water-miscible solvents such as methanol, ethanol and dimethyl sulfoxide (Glogauer et al. 2011).LipC12 also shows regioselectivity in the deacetylation of glycoderivatives (Alnoch et al. 2015) and good activity for the synthesis of structured lipids (Madalozzo et al. 2016).However, in the transesterification of (RS)-1-phenylethanol with vinyl acetate, a model reaction for the kinetic resolution of secondary alcohols, LipC12 immobilized on Immobead 150 gave an enantiomeric ratio (E) of 46 (Rpreference).This E-value is low compared to the values of over 200 that are obtained for commercial lipases like CALB, CRL (Candida rugosa lipase) and PPL (porcine pancreas lipase; Costa et al. 2013).
The good activity and stability of LipC12 in aqueous media and organic solvents prompted us to attempt to improve its enantioselectivity through protein engineering, guided by molecular docking, a technique that has successfully guided the improvement of enantioselectivity of other lipases (Rotticci et al. 2001;Gu et al. 2011;Dhake et al. 2012;Wu et al. 2013;Yang et al. 2017;Shen et al. 2018).Docking studies for the hydrolysis reaction were done with the transition-state complex in which the tetrahedral intermediate of p-nitrophenyl octanoate is covalently linked to the catalytic serine.The best variants obtained by molecular docking were then expressed in E. coli and characterized experimentally.

Strains, plasmids and materials
The strains Escherichia coli TOP10 (Invitrogen, CA, USA) and E. coli BL21 (kDE3; Novagen, MI, USA) and the vector pET-28a(þ) (Novagen, MI, USA) were used as the recombinant protein expression system.Key reagents and materials were DpnI and Phusion DNA polymerase (New England Biolabs, MA, USA), a protein molecular mass marker (Thermo Fisher Scientific.MA, USA) and a His-Trap Chelating HP column (GE Healthcare, Uppsala, Sweden).The support Immobead 150, IPTG (isopropyl b-D thiogalactopyranoside) and the substrates p-nitrophenyl octanoate and (RS)-1-phenylethanol were from Sigma-Aldrich (MO, USA).All other chemicals were of analytical grade.

Molecular docking and structural analyses
The tetrahedral intermediates of the (R) and (S)-isomers of 1-phenylethanol acetate and p-nitrophenyl octanoate were generated using Chimera (Pettersen et al. 2004).The energies of each substrate were minimized to give a low-energy starting conformation with suitable bond distance and angles.Starting from these initial structures, the docking program Autodock Vina was used to explore the conformational space accessible to the substrate covalently bound to the catalytic serine (Ser 83) in a tetrahedral form (Morris et al. 2009;Seeliger and De Groot 2010).The LipC12 model was obtained through homology modelling with PAL (Pseudomonas aeruginosa lipase) as a reference (PDB id: 1EX9) using Modeller (Fiser and Sali 2003) and evaluated with PROCHECK (Laskowski et al. 1993).The docking region covered all amino acids of the catalytic cleft of LipC12.The positions of the amino acid residues were defined as fixed in space using AutoDockTools (Morris et al. 2009).DG binding was calculated for the formation of complexes and the geometry and distance between the tetrahedral intermediate of 1-phenylethanol acetate and the residues of the catalytic triad and the distance between the oxyanion and the hydrogen donors of the oxyanion hole were calculated.Variants of LipC12 were constructed by in silico replacement of the target amino acid (V261).The conformation of the sidechain of the mutated residue was optimized by manually selecting a low-energy conformation from a sidechain rotamer library.Steric clashes (van der Waals overlap) and non-bonded interaction energies (Coulombic) were evaluated for the different sidechain conformations, and the molecular docking simulation was repeated with the new LipC12 variants.

Site-directed mutagenesis
The gene of the variant LipC12 V261A was obtained using the oligonucleotide site-directed mutagenesis method (Ho et al. 1989), using the commercial kit QuikChange II Site-Directed Mutagenesis (Agilent Technologies, CA, USA).Oligoprimers were synthesized by Integrated DNA Technologies (IA, USA).Other genes were also obtained by the same methodology.Incorporation of the desired mutation was confirmed by gene sequencing by the Sanger method (Sanger et al. 1977) using commercial BigDye Terminator v3.1 (Life Technologies, CA, USA) and the 3500XL DNA sequencing system (Applied Biosystems, MA, USA).The recombinant plasmids were then transformed into competent E. coli BL21(kDE3) cells to express variants of LipC12.

Expression and purification of recombinant LipC12 and its variants
Wild-type LipC12 (LipC12 wt ) and its variants were expressed according to Glogauer et al. (2011), using E. coli BL21(kDE3) transformed with pET-28a(þ) and grown in lysogenic broth with kanamycin (50 mg mL À1 ), in an orbital shaker at 180 rpm and 37 C, until an OD 600 of 0.5.The culture was then induced by the addition of 0.5 mM IPTG.The induced culture was incubated for a further 16 h at 20 C before harvesting of the cells by centrifugation (10,000Âg for 10 min) at 4 C.The cell pellet was re-suspended in 30 mL of lysis buffer (50 mM Tris-HCl pH 7.5, 500 mM NaCl) and disrupted by ultrasonication in an ice bath (15 cycles of 20 s pulses, 90 W, with 15 s intervals), using a SONICATOR V R XL 2020 (Heat Systems-Ultrasonics Inc., NY, USA).The crude extract was centrifuged at 12,000Âg for 30 min at 4 C to pellet the cell debris and the supernatant containing protein was then loaded onto a HiTrap Chelating HP column previously loaded with Ni 2þ and equilibrated with lysis buffer.Elution was stepwise, with the lysis buffer and increasing concentrations (from 10 to 500 mM) of imidazole.The fractions containing purified LipC12 were detected by 12% (w/v) SDS-PAGE using 10 mg of protein per well and then pooled.The pooled fractions were then dialyzed in 50 mM Tris-HCl pH 8.0, 150 mM NaCl, and 10 mM CaCl 2 and stored at 4 C until use.

Protein determination and electrophoresis
Protein content was determined by the method of Bradford (1976), using the Quick Start Bradford Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA), with bovine serum albumin as the standard.Electrophoresis of protein samples was done by 12% (w/v) SDS-PAGE according to Laemmli (1970).The gels were stained with Coomassie Brilliant Blue R-250 and destained with methanol/acetic acid/water (5/1/4 v/v/v).

Lipase activity assay
The hydrolytic activity of free LipC12 and its variants was determined using p-nitrophenyl butyrate, p-nitrophenyl octanoate, p-nitrophenyl dodecanoate and pnitrophenyl palmitate, with monitoring of the release of p-nitrophenol at 410 nm, 25 C, in an iMark TM Microplate Absorbance Reader (Bio-Rad, CA, USA; Winkler and Stuckmann 1979).The free enzyme was added to a 250 mL reaction mixture containing 50 mM Tris-HCl buffer, pH 7.5, 1 mM CaCl 2 , 0.6% (v/v) Triton X-100 and 1 mM of the substrate.The final concentration of free enzyme in the assay was 5 nM.All experiments were carried out in triplicate and the results were obtained using a calibration curve for p-nitrophenol.Initial reaction velocities were determined by linear regression using Microsoft Excel.Results are reported as specific activities ± the sample standard deviation (both in U mg À1 ).A unit of activity represents 1 mmol of p-nitrophenol produced per minute under the assay conditions.

Lipase immobilization
For the transesterification of 1-phenylethanol, it was necessary to immobilize LipC12 wt and its variants as free LipC12 does not have activity when added directly to the reaction medium.Based on previous studies of the immobilization of LipC12, Immobead 150 was chosen as the support as it gave the best results for LipC12 in organic media (Madalozzo et al. 2015;S anchez et al. 2021).LipC12 wt and its variants were immobilized covalently onto Immobead 150, according to Madalozzo et al. (2015).Protein loadings of 20 mg g À1 of support were added in 10 mL of buffer (Tris-HCl buffer 25 mM pH 7.0) and the mixture was incubated in an orbital shaker at 150 rpm and 4 C for 12 h.Aliquots of the supernatant were removed at the end of the immobilization process and assayed for lipase activity using the pNP method.The immobilization efficiency was 100% (i.e.all activity disappeared from the supernatant).

Determination of the enantiomeric ratio
The E-values of immobilized LipC12 wt and its variants were determined for the transesterification of (RS)-1phenylethanol (0.1 mmol) with vinyl acetate (0.4 mmol) in n-hexane (2.0 mL), using 100 mg of each immobilized preparation, in 24 h reactions (30 C; Alnoch et al. 2015).Samples were collected and analysed by gas chromatography, using a GC-2010 chromatograph (Shimadzu Co., Kyoto, Japan), equipped with a flame ionization detector (FID) and a chiral column CP Chirasil-DEX CB (25 m Â 0.25 mm diameter, 0.25 mm film thickness).A sample of 0.5 mL was injected using N 2 as the carrier gas (1 mL min À1 ).The injector and detector were set at 220 C. The oven program was as follows: initial value of 110 C, with heating at 1 C min À1 to 120 C. The peaks were identified based on a control reaction catalyzed by CALB, which has a known R-preference in the resolution of (RS)-1-phenylethanol (Habulin and Knez 2009).Retention times were: ester-(R) 13.49 min, ester-(S) 13.79 min, alcohol-(S) 14.38 min and alcohol-(R) 15.09 min, The peak areas obtained for the R-enantiomer (S R ) and the S-enantiomer (S S ) of the substrate were used to calculate the enantiomeric excess of the substrate (ee s , %): Likewise, the peak areas obtained for the R-enantiomer (P R ) and the S-enantiomer (P S ) of the product were used to calculate the enantiomeric excess of the product (ee p , %).
These enantiomeric excesses were then used to calculate the degree of conversion of the reaction (c, %).When the enzyme showed R-preference, c was calculated as When the enzyme showed S-preference, c was calculated as Finally, the enantiomeric excess of substrate and c were used to calculate the enantiomeric ratio (E) using the equation of Chen et al. (1982).When the enzyme showed R-preference, E was calculated as When the enzyme showed S-preference, E was calculated as

Results and discussion
3.1.In silico analysis of interactions and transition-state geometry During the synthesis of 1-phenylethanol acetate from vinyl acetate and 1-phenylethanol, the 1-phenylethanol attacks the acetylated serine of LipC12, forming a tetrahedral intermediate (Figure 1 polypeptide backbone of residues L16 and Q84); and (iii) the distance between the catalytic histidine (H260) and the oxygen of the alcohol moiety of the substrate and the oxygen of the catalytic serine.
The distance between the oxygen of the catalytic serine and the catalytic histidine is 3.24 Å for the (S)isomer and 3.16 Å for the (R)-isomer; the distance between the oxygen of the alcohol moiety of the substrate and the catalytic histidine is 2.70 Å for the (S)isomer and 2.52 Å for the (R)-isomer (Figure 1(b,c)).The smaller distance for the (R)-isomer complex suggest that proton transfer could be easier for the (R)isomer, allowing faster reaction, which is in accordance with experimental results (Markle et al. 2011).
Distance between the amino acid sidechains of the active site and the substrate are similar for the (R)-isomer and the (S)-isomer, except for L240 and V261: there are significant differences in the distance from these sidechains to the two isomers.L240 is at a distance greater than 4.0 Å from the (S)-isomer in Figure 1(b), but less than 4.0 Å from the (R)-isomer in Figure 1(c).Conversely, V261 is closer to the (S)-isomer than it is to the (R)-isomer.
To improve the enantioselectivity of lipases for alcohols, residues in the hydrophobic and hydrophilic cavities should be mutated (Figure 2 Other possible positions for mutations would be L16 (part of the oxyanion hole) and H82 (in the hydrophilic cavity), which are near the bound 1-phenylethanol acetate moieties.However, due to their proximity to the catalytic serine (S83), mutations at these positions could inactivate LipC12.
As a position analogous to V261 in LipC12 has been found to be important for the enantioselectivity of lipases (Durmaz et al. 2013), this position was chosen for the site-directed mutation experiments.

Using docking to evaluate the importance of position 261 in LipC12
To confirm the importance of position 261 for the enantioselectivity of LipC12, we replaced the valine residue with alanine, which is also nonpolar, but less bulky.Docking shows that, compared to LipC12 wt , this replacement does not affect the distance between the oxyanion and the hydrogen donors of the oxyanion hole, suggesting that the geometry of the transitionstate complex is not affected.However, the replacement increases the distance between the aromatic ring of the 1-phenylethanol moiety and the hydrophobic cavity for both the (S)-isomer (V261 ¼ 3.76 Å and A261 ¼ 5.65 Å) and the (R)-isomer (V261 ¼ 7.70 Å and A261 ¼ 9.48 Å; Figure 3).In LipC12 V261A , the increase in the distance decreases van der Waals interactions between the aromatic ring of the 1-phenylethanol moiety and the hydrophobic cavity.This could facilitate the catalysis of the reaction with the (S)-isomer and decrease the enantioselectivity of this variant.
The identity of the amino acid at position 261 is predicted to affect the hydrolytic activity of LipC12.Figure 4 shows the tetrahedral intermediate of p-nitrophenyl octanoate, in which the position of the aromatic ring (of the p-nitrophenyl moiety) is similar to the position of the aromatic ring of the tetrahedral intermediate of (R)-1-phenylethanol acetate.The mutation V261A increases the distance between the aromatic ring of p-nitrophenyl octanoate and the sidechain at position 261: for LipC12 wt it is 6.19 Å and for LipC12 V261A it is 8.94 Å.

Experimental results of selected variants
The variant LipC12 V261A , obtained by site-directed mutagenesis, showed a decrease in enantioselectivity for (RS)-1-phenylethanol (E ¼ 5, compared to E ¼ 46 for LipC12 wt ) in the transesterification reaction with vinyl acetate and an almost 2-fold increase in activity for the hydrolysis of p-nitrophenyl octanoate.These results indicate that position 261 is key for both enantioselectivity and hydrolytic activity of LipC12 and a suitable target for protein engineering.
We obtained several variants of LipC12 in which the valine at position 261 was mutated to a different amino acid.In LipC12 V261Q and LipC12 V261K , V261 was mutated to glutamine and lysine, respectively.In both cases, the aim was to increase the polarity of the hydrophobic cavity.In LipC12 V261F , LipC12 V261Y and LipC12 V261W , V261 was mutated to phenylalanine, tyrosine and tryptophan, respectively.The bulky aromatic sidechains of these amino acids affect the geometry of the transition state and the distance from the oxygen of the substrate to the residues of the oxyanion hole and also to the catalytic histidine.In LipC12 V261I , V261 was mutated to isoleucine, for which docking studies showed the greatest positive DDG binding between the (R) and (S)-isomers of 1-phenylethanol acetate, and which had a DG binding close to that of LipC12 for pnitrophenyl octanoate (Figure S1).The amino acid at position 261 affected both the enantioselectivity and the hydrolytic activity of LipC12 (Table 1).In terms of the enantioselectivity in the kinetic resolution of 1-phenylethanol, the best variant was LipC12 V261Q , for which the E-value increased 2.4fold.However, this variant gave a very low conversion (3% in 24 h), suggesting that the change also affected its transesterification activity.There was a similar decrease in its hydrolytic activity (40 U mg À1 ) compared to LipC12 wt (173 U mg À1 ).On the other hand, the best variant in terms of hydrolytic activity was LipC12 V261F , which had an almost 6-fold higher hydrolytic activity than did LipC12 wt .Other variants with increased hydrolytic activity were LipC12 V261I and LipC12 V261A , both of which showed a significant decrease in enantioselectivity.
Potentially, the differences in activity and stereoselectivity in the kinetic resolution, which was done with immobilized preparations, could be due not only to direct effects of the mutations on the shape of the active site but also to effects of the mutations on the immobilization of the variants on Immobead 150.The effects are most likely direct effects on the active site, since the modifications are limited to a single amino 156 ± 19 a When the enzyme shows R-preference, ee S s is calculated using Equation (1a) and ee R p is calculated using Equation (2b), the conversion (c) is calculated using Equation (3a) and E R is calculated using Equation (4a).When the enzyme shows S-preference, ee R s is calculated using Equation (1b) and ee S p is calculated using Equation (2a), c is calculated using Equation (3b) and E S is calculated using Equation (4b).b The subscript indicates the preferred enantiomer.Transesterification of (RS)-1-phenylethanol (0.1 mmol) with vinyl acetate (0.4 mmol) in n-hexane (2.0 mL) using 100 mg of immobilized preparation in a 24 h reaction (35 C). c Hydrolytic activity was measured using p-nitrophenyl octanoate, with monitoring of the released p-nitrophenol at 410 nm (25 C).The free enzyme was added to the 250 mL reaction mixture (50 mM Tris-HCl buffer, pH 7.5, 1 mM CaCl 2 , 0.6% (v/v) Triton X-100, 1 mM of p-nitrophenyl octanoate), to give a final enzyme concentration of 5 nM.Results are reported as units of enzyme activity (U) and errors represent the sample standard deviation.1 U is defined as 1 mmol of p-nitrophenol produced per minute under the assay conditions.acid in the hydrophobic cavity of the active site, and the active site is unlikely to be involved directly in the formation of covalent bonds with Immobead 150.In other words, it is unlikely that LipC12 wt and the variants immobilized differently on Immobead 150.However, this is a question that deserves further attention.
None of the variants showed both high enantioselectivity and high hydrolytic activity.Similar trade-offs between enantioselectivity and activity have been reported previously.For example, in the directed evolution of an esterase of Rhodobacter sphaeroides, Ma et al. (2013) obtained mutant enzymes with either high enantioselectivity or high specific activity, but not both.In a similar manner, Dorau et al. (2018) engineered subtilisin Carlsberg for improved stereoselectivity in the transesterification of 1-phenylethanol, but the activity was low.More recently, after site-specific mutagenesis on lipase A from Bacillus subtilis, Li et al. (2021) obtained a double mutant with a significantly improved enantioselectivity for the hydrolysis of 1phenylpropyl acetate, but the catalytic constant decreased by a factor of three in relation to the wildtype enzyme.Such results are not limited to stereoselectivity: Quaglia et al. (2019) engineered CAL-A (Candida antarctica lipase A) to change its chainlength selectivity, but, although the mutant enzyme was more selective for the hydrolysis of pNP esters of long-chain fatty acids, it had lower activity on these long-chain fatty acids than did the wild-type.
Further protein engineering has the potential to overcome the initial trade-off between selectivity and activity.For example, Guo et al. (2013) carried out sitedirected saturation mutagenesis on four residues of one of the variants of the esterase of Rhodobacter sphaeroides obtained by Ma et al. (2013), obtaining a mutant with high enantioselectivity and activity.Likewise, Dorau et al. (2018) introduced a second mutation into subtilisin Carlsberg, obtaining a variant with good activity and very high stereoselectivity.Molecular dynamics, docking and deconvolution are useful tools to guide such studies (Li and Reetz 2016).
In the variant with the highest enantioselectivity, LipC12 V261Q , the mutation increased both the steric hindrance (Gln has a bigger sidechain than Val) and the polarity of the hydrophobic cavity.Of the variants that presented higher hydrolytic activities than LipC12 wt , two have nonpolar aliphatic residues (Ile and Ala) in position 261, while the one with the highest hydrolytic activity, LipC12 V261F , has a large aromatic residue (Phe) in this position.Interestingly, LipC12 V261F had enantiopreference for (S)-1-phenylethanol, whereas LipC12 wt and all other variants had Renantiopreference.
Changes in enantiopreference of lipases have been reported and usually occur when a smaller residue is replaced with a significantly larger one or vice versa (Funke et al. 2005;Li et al. 2021).In our case, the increase of the hydrophobicity of the hydrophobic cavity with the introduction of Phe allows more interactions with the (S)-isomer.
Due to the low conversions obtained with the variants LipC12 V261Q , LipC12 V261Y and LipC12 V261W , the kinetic resolutions were repeated for these variants with 200 mg of immobilized preparation (instead of 100 mg) and 72 h reaction time (instead of 24 h).The E-values were the same as those shown in Table 1, as would be expected.The conversions increased, reaching 11%, 5% and 6%, respectively.However, these values are still low and demonstrate the need for further work to improve the activity of these variants.

The effect of the length of the fatty acid carbon chain on the activity of LipC12 wt and its variants
The hydrolytic activities of LipC12 wt and its variants were tested against several p-nitrophenyl esters with fatty acid chains of different lengths (Figure 5) to verify the effect of mutations in the hydrophobic cavity.
The profile for LipC12 V261K was quite similar to that of LipC12 wt .The profile for LipC12 V261A was also similar, with the exception that the activity against p-nitrophenyl octanoate was 2-fold higher (Figure 5).For both LipC12 V261I and LipC12 V261F , all activities were higher than the corresponding activities of the wild type, especially the activity against p-nitrophenyl octanoate.
Changes in specific activity against fatty acids of different chain lengths have been reported in engineered lipases (Scheib et al. 1999;Durmaz et al. 2013).Usually, site-directed mutagenesis decreases the hydrolytic activity when it increases steric hindrance (i.e. when the introduction of larger amino acids decreases the space available inside the catalytic cleft).This could explain the lower activities presented by LipC12 V261Y and LipC12 V261W , against all substrates, with a decrease of activity of 80% compared to LipC12 wt (Figure 5).However, the highest activity against the p-nitrophenyl fatty acid esters was found in LipC12 V261F , in which Val was replaced with the bulkier Phe.Importantly, Phe (XLogP3 ¼ À1.5) is more hydrophobic than Val (XLogP3 ¼ À2.3; Kim et al. 2021) and therefore increases the hydrophobicity of the hydrophobic cavity, which could facilitate the interaction of the substrate with the catalytic cleft (Petersen et al. 2001).
Positions analogous to V261 of LipC12 have been described as important for the activity and enantioselectivity of lipases.Durmaz et al. (2013) produced a variant of Geobacillus thermocatenulatus lipase (BTL2), BTL2 L360F , in which the mutation is analogous to that of LipC12 V261F (Figure 6).Compared to the wild type, BTL2 L360F had the same hydrolytic activity against tributyrin, but a 3-fold higher activity against tricaprylin.The results suggest that the increased hydrophobicity of the hydrophobic cavity could facilitate the access of more hydrophobic substrates while it repelled the charged product more efficiently.

Conclusion
Molecular docking of transition-state complexes identified position V261 in LipC12 wt as important for enantioselectivity in the transesterification of (RS)-1-  phenylethanol and for activity in the hydrolysis of pnitrophenyl octanoate.Site-directed mutagenesis of LipC12 in this position produced three variants with higher hydrolytic activities, with the best one, LipC12 V261F , having an activity 5.8-fold higher than that of LipC12 wt .The best result for enantioselectivity was obtained for the variant LipC12 V261Q , which presented an E-value 2.4-fold higher than LipC12 wt .Improvements in activity and enantioselectivity were inversely correlated: compared to the wild type, variants with high activities had lower enantiomeric ratios and the variant with increased enantioselectivity (LipC12 V261Q ) had a 4-fold lower hydrolytic activity.Even though activity and enantioselectivity were inversely correlated in the variants, this study serves as a basis for other mutations in LipC12 wt and its variants with the aim of compensating the trade-off of enantioselectivity and activity.
Figure 1.Transition states of LipC12 wt with 1-phenylethanol acetate enantiomers.(a) Scheme of the tetrahedral intermediate showing the catalytic triad and the residues of the oxyanion hole.(b) Complex with (S)-1-phenylethanol acetate.(c) Complex with (R)-1-phenylethanol acetate.The red arrows indicate the positions of the residues L240 and V261.
(a); Maldonado et al. 2021).Positions L240 and V261 are good candidates for mutation since they are both in the hydrophobic cavity, with V261 near the hydrophilic cavity and L240 near the acyl-binding pocket (Figure 2(b)).

Figure 2 .
Figure 2. Tetrahedral intermediates formed with 1-phenylethanol acetate isomers complexed with wild-type LipC12.(a) View of the catalytic cleft with both enantiomers (which are shown using the ball and stick representation).(b) Focus on the hydrophobic cavity and possible targets for site-directed mutagenesis.

Figure 4 .
Figure 4. Tetrahedral intermediate of p-nitrophenyl octanoate.(a) Scheme of the tetrahedral intermediate with the residues of the catalytic triad and the oxyanion hole.(b) Complex with wild-type LipC12.(c) Complex with LipC12 V261A .The red arrow indicates the position where LipC12 was mutated.

Figure 5 .
Figure 5. Activity of free LipC12 wt and its variants against p-nitrophenyl esters with different carbon chain lengths.The specific hydrolytic activities (U mg À1 ) are reported as the means of triplicates, with the error bars representing the standard deviation of the sample. 1 U was defined as 1 mmol of p-nitrophenol produced per minute under the assay conditions.

Figure 6 .
Figure 6.Superimposed structures of Geobacillus thermocatenulatus lipase and LipC12.(a) Structure of LipC12 wt lipase in light sea green and structure of G. thermocatenulatus wt lipase in light gray; (b) mutations at positions L360 for G. thermocatenulatus lipase and V261 for LipC12.

Table 1 .
Effect of mutations at position 261 on the enantioselectivity and on the hydrolytic activity of LipC12.