Microbial Synthesis and Evaluation of Fungistatic Activity of 3-Butyl-3-hydroxyphthalide, the Mammalian Metabolite of 3-n-Butylidenephthalide

Phthalides are bioactive compounds that naturally occur in the family Apiaceae. Considering their potentially versatile applications, it is desirable to determine their physical properties, activity and metabolic pathways. This study aimed to examine the utility of whole-cell biocatalysts for obtaining 3-butyl-3-hydroxyphthalide, which is the metabolite formulated during mammalian metabolism of 3-n-butylidenephthalide. We performed transformations using 10 strains of fungi, five of which efficiently produced 3-butyl-3-hydroxyphthalide. The product yield, determined by high-performance liquid chromatography, reached 97.6% when Aspergillus candidus AM 386 was used as the biocatalyst. Increasing the scale of the process resulted in isolation yields of 29–45% after purification via reversed-phase thin layer chromatography, depending on the strain of the microorganism used. We proposed different mechanisms for product formation; however, hydration of 3-n-butylidenephthalide seems to be the most probable. Additionally, all phthalides were tested against clinical strains of Candida albicans using the microdilution method. Two phthalides showed a minimum inhibitory concentration, required to inhibit the growth of 50% of organisms, below 50 µg/mL. The 3-n-butylidenephthalide metabolite was generally inactive, and this feature in combination with its low lipophilicity suggests its involvement in the detoxification pathway. The log P value of tested compounds was in the range of 2.09–3.38.

Phthalides have also been examined as potential fungistatic agents. 3-n-butylidenephthalide appeared to be more active against dermatophytes than the antifungals itraconazole and ketoconazole [12]. Significantly, similarly structured 3-n-butylphthalide inhibited the growth of four clinical strains of Candida albicans, while the reference drug, fluconazole, was inactive at much higher concentrations [13]. Candida species are one of the most common causes of fungal infections worldwide. Even though these microorganisms are ubiquitous (Figures S1-S16, Supplementary Materials). Next, we evaluated the lipophilicity of compounds using reversed-phase high-performance liquid chromatography (RP-HPLC) and tested their antimicrobial properties against selected Candida yeasts. We also aimed to establish whether a metabolite with a different polarity also exhibits bioactive potential or is suggestive of xenobiotic detoxification.

Results and Discussion
Screening biotransformations involving fungal strains were performed to select the most effective biocatalysts for producing 3-butyl-3-hydroxyphthalide (2). This approach allowed the resulting microorganisms to be used in scale-up processes to isolate the metabolite and confirm its structure by nuclear magnetic resonance (NMR) spectroscopy (Figures S1-S16, Supplementary Materials). Next, we evaluated the lipophilicity of compounds using reversed-phase high-performance liquid chromatography (RP-HPLC) and tested their antimicrobial properties against selected Candida yeasts. We also aimed to establish whether a metabolite with a different polarity also exhibits bioactive potential or is suggestive of xenobiotic detoxification.
Biotransformation of 3-n-butylidenephthalide (1) as a mixture of (E) and (Z) isomers (9:1) at a concentration of 0.27 mg/mL was carried out for 14 days at 25 °C with shaking culture in Sabouraud medium. Samples were taken every 2 days, extracted with ethyl acetate and analyzed via thin layer chromatography (TLC) and HPLC (Figures S14-16, Supplementary Materials).
The highest product (2) yield of 97.61% was obtained using A. candidus AM 386 on day 4, as determined by HPLC ( Figure 2). However, in contrast to that in other strains, A. candidus AM 386 further metabolized compound (2) to unidentified derivatives. The other tested microorganisms, A. cylindrospora AM 336, M. spinosus AM 398, C. indicum AM 158 and P. resinae AM 50, produced yields between 25.13% and 72.40% by day 14. We also focused on the pH of the cultures over the course of the biotransformations (Figure 3). In cultures with M. spinosus AM 398 and P. resinae AM 50, the conditions were slightly acidic (pH 5-6), whereas the A. cylindrospora AM 336 and C. indicum AM 158 strains subtly alkalized the culture (pH 7.4-8.5). In the A. candidus AM 386 culture, the pH was in the range of 5.5-5.0 in the first few days but subsequently increased to 8.5. The sudden change in pH value presumably occurred due to the change in nutrient availability. During culture growth, depletion of the carbon source Initially, screening-scale biotransformation of 3-n-butylidenephthalide (1) was conducted using 10 fungal strains, which were analyzed for their catalytic potential. These strains included Absidia cylindrospora AM 336, Ascosphaera apis AM 496, Aspergillus candidus AM 386, Chaetomium indicum AM 158, Fusarium culmorum AM 9, Fusicoccum amygdali AM 258, Laetisporus sulphurens AM 515, Mucor spinosus AM 398, Penicillium chrysogeum AM 112 and Pycnidiella resinae AM 50 (data not shown). Of these, the five strains that exhibited the most efficient product (2) formation were Aspergillus candidus AM 386, Absidia cylindrospora AM 336, Mucor spinosus AM 398, Chaetomium indicum AM 158 and Pycnidiella resinae AM 50. These five strains were chosen for further screening.
Biotransformation of 3-n-butylidenephthalide (1) as a mixture of (E) and (Z) isomers (9:1) at a concentration of 0.27 mg/mL was carried out for 14 days at 25 • C with shaking culture in Sabouraud medium. Samples were taken every 2 days, extracted with ethyl acetate and analyzed via thin layer chromatography (TLC) and HPLC (Figures S14-16, Supplementary Materials).
The highest product (2) yield of 97.61% was obtained using A. candidus AM 386 on day 4, as determined by HPLC ( Figure 2). However, in contrast to that in other strains, A. candidus AM 386 further metabolized compound (2) to unidentified derivatives. The other tested microorganisms, A. cylindrospora AM 336, M. spinosus AM 398, C. indicum AM 158 and P. resinae AM 50, produced yields between 25.13% and 72.40% by day 14. We also focused on the pH of the cultures over the course of the biotransformations (Figure 3). In cultures with M. spinosus AM 398 and P. resinae AM 50, the conditions were slightly acidic (pH 5-6), whereas the A. cylindrospora AM 336 and C. indicum AM 158 strains subtly alkalized the culture (pH 7.4-8.5). In the A. candidus AM 386 culture, the pH was in the range of 5.5-5.0 in the first few days but subsequently increased to 8.5. The sudden change in pH value presumably occurred due to the change in nutrient availability. During culture growth, depletion of the carbon source causes stress for the microbial cells, resulting in extracellular release of ammonia [34]. The stability of the compounds under extreme pH conditions were examined using a mixture of substrates and products. causes stress for the microbial cells, resulting in extracellular release of ammonia [34]. The stability of the compounds under extreme pH conditions were examined using a mixture of substrates and products.  In order to confirm the structure of compound (2) and determine its mechanism of action, we scaled up the biotransformation with the following strains: A. candidus AM 386, A. cylindospora AM 336 and C. indicum AM 158. We decided to use three strains on a preparative scale instead of focusing on only one of them, as the isolation yields of product (2) might have been diversified. This issue may result from different metabolite concentrations in particular strains, which are not visible during screening-scale transformation while the biotransformation is monitored by HPLC. We also sought to determine the stereoselectivity of the process for particular strains. Biotransformations were performed in 2000-mL shaken flasks with 3-n-butylidenephthalide (1) at a concentration of 150 mg/500 mL at 25 °C. The products were analyzed using gas chromatography (GC), HPLC and RP-TLC. All three microorganisms sufficiently transformed 3-n-butylidenephthalide (1) (Figure 4). The conversion of the substrate was slower in C. indicum AM 158, which is consistent with the results of the small-scale experiments. The biotransformations continued until day 14 with the exception of A. candidus AM 386, which was only continued to day 8 because of an observed decrease in product yield. Content of product (2) during the screening-scale biotransformations of 3-nbutylidenephthalide (1) determined using a high-performance liquid chromatography-diode array detector. causes stress for the microbial cells, resulting in extracellular release of ammonia [34]. The stability of the compounds under extreme pH conditions were examined using a mixture of substrates and products.  In order to confirm the structure of compound (2) and determine its mechanism of action, we scaled up the biotransformation with the following strains: A. candidus AM 386, A. cylindospora AM 336 and C. indicum AM 158. We decided to use three strains on a preparative scale instead of focusing on only one of them, as the isolation yields of product (2) might have been diversified. This issue may result from different metabolite concentrations in particular strains, which are not visible during screening-scale transformation while the biotransformation is monitored by HPLC. We also sought to determine the stereoselectivity of the process for particular strains. Biotransformations were performed in 2000-mL shaken flasks with 3-n-butylidenephthalide (1) at a concentration of 150 mg/500 mL at 25 °C. The products were analyzed using gas chromatography (GC), HPLC and RP-TLC. All three microorganisms sufficiently transformed 3-n-butylidenephthalide (1) (Figure 4). The conversion of the substrate was slower in C. indicum AM 158, which is consistent with the results of the small-scale experiments. The biotransformations continued until day 14 with the exception of A. candidus AM 386, which was only continued to day 8 because of an observed decrease in product yield. In order to confirm the structure of compound (2) and determine its mechanism of action, we scaled up the biotransformation with the following strains: A. candidus AM 386, A. cylindospora AM 336 and C. indicum AM 158. We decided to use three strains on a preparative scale instead of focusing on only one of them, as the isolation yields of product (2) might have been diversified. This issue may result from different metabolite concentrations in particular strains, which are not visible during screening-scale transformation while the biotransformation is monitored by HPLC. We also sought to determine the stereoselectivity of the process for particular strains. Biotransformations were performed in 2000-mL shaken flasks with 3-n-butylidenephthalide (1) at a concentration of 150 mg/500 mL at 25 • C. The products were analyzed using gas chromatography (GC), HPLC and RP-TLC. All three microorganisms sufficiently transformed 3-n-butylidenephthalide (1) (Figure 4). The conversion of the substrate was slower in C. indicum AM 158, which is consistent with the results of the small-scale experiments. The biotransformations continued until day 14 with the exception of A. candidus AM 386, which was only continued to day 8 because of an observed decrease in product yield. The pH of the culture media was in the range of 7-8.25 when using fungi from the genera Chaetomium and Absidia. Meanwhile, the pH of cultures of fungi from the genus Aspergillus was acidic and remained stable at pH 5 during biotransformation. Monitoring the pH value allowed us to determine whether it was necessary to acidify the samples to extract the product from the organic phase. The biotransformation mixtures were extracted by ethyl acetate and purified on RP-TLC plates ( Figure S12, Supplementary Materials).
Using NMR spectra, we confirmed that the main isolated product of the biotransformation was (-)-3-butyl-3-hydroxyphthalide (2). The assignment of aromatic signals of 3butyl-3-hydroxyphthalide (2) in the range of δ = 7.56-7.81 ppm was mainly allowed by Correlation Spectroscopy (COSY) (coupling of H-5 with H-6 and H-4 protons). This was also facilitated via Nuclear Multiple Bond Coherence (HMBC) spectrum coupling of C-13 with protons H-5 and H-7 without correlation with H-4, as well as a coupling of C-12 with H-4 and H-6 without correlation with H-7 ( Figure 5). The presence of a hydroxyl group was confirmed by the broad signal in the range of δ = 3.70-4.50 ppm ( Figure S3, Supplementary Materials). The characteristic signal for the proton H-8 in the substrate structure at δ = 5.64 was absent in the spectra of this derivative. The signal for the carbon atom with the hydroxyl group (C-3) in the 3-butyl-3-hydroxyphthalide (2) spectrum was shifted upfield (δ = 107.9) when compared to that of the unsubstituted carbon atom (δ = 145.9) in the precursor compound (1) ( Figure S4, Supplementary Materials). Overall, the yield of the product (2) in the biotransformation mixture at the end of the process was in the range 57-82% depending on the strain used, as determined by HPLC. The yield after isolation was approximately 29-45% (Table 1)

Chaetomium indicum AM158
Aspergillus candidus AM 386 The pH of the culture media was in the range of 7-8.25 when using fungi from the genera Chaetomium and Absidia. Meanwhile, the pH of cultures of fungi from the genus Aspergillus was acidic and remained stable at pH 5 during biotransformation. Monitoring the pH value allowed us to determine whether it was necessary to acidify the samples to extract the product from the organic phase. The biotransformation mixtures were extracted by ethyl acetate and purified on RP-TLC plates ( Figure S12, Supplementary Materials).
Using NMR spectra, we confirmed that the main isolated product of the biotransformation was (-)-3-butyl-3-hydroxyphthalide (2). The assignment of aromatic signals of 3-butyl-3-hydroxyphthalide (2)   The pH of the culture media was in the range of 7-8.25 when using fungi from the genera Chaetomium and Absidia. Meanwhile, the pH of cultures of fungi from the genus Aspergillus was acidic and remained stable at pH 5 during biotransformation. Monitoring the pH value allowed us to determine whether it was necessary to acidify the samples to extract the product from the organic phase. The biotransformation mixtures were extracted by ethyl acetate and purified on RP-TLC plates ( Figure S12, Supplementary Materials).
Using NMR spectra, we confirmed that the main isolated product of the biotransformation was (-)-3-butyl-3-hydroxyphthalide (2). The assignment of aromatic signals of 3butyl-3-hydroxyphthalide (2)   Overall, the yield of the product (2) in the biotransformation mixture at the end of the process was in the range 57-82% depending on the strain used, as determined by HPLC. The yield after isolation was approximately 29-45% (Table 1)

Chaetomium indicum AM158
Aspergillus candidus AM 386 Overall, the yield of the product (2) in the biotransformation mixture at the end of the process was in the range 57-82% depending on the strain used, as determined by HPLC. The yield after isolation was approximately 29-45% (Table 1). The specific rotations

Proposed Pathways for 3-Butyl-3-hydroxyphthalide (2) Formation
Hydroxylation frequently occurs via cytochrome P450 enzymes involved in xenobiotic metabolism [35]. To examine the possibility of 3-butyl-3-hydroxyphthalide (2) formation through hydroxylation of the saturated analog (4) of 3-n-butylidenephthalide (1) (Figure 6), we also performed biotransformations of 3-n-butylphthalide (4) using the three previously used microorganisms. However, our experiments showed that 3-n-butylphthalide (4) was not converted to product (2). We used 13 C NMR as a diagnostic tool to analyze the biotransformation mixture. Spectra of the biotransformation mixtures showed a lack of the characteristic signal for the C-3 carbon in 3-butyl-3-hydroxyphthalide (2) at δ = 107.9 ppm (Figure S13, Supplementary Materials). Additionally, the saturated analog (4) of 3-n-butylidenephthalide (1) However, our experiments showed that 3-n-butylphthalide (4) was not converted to product (2). We used 13 C NMR as a diagnostic tool to analyze the biotransformation mixture. Spectra of the biotransformation mixtures showed a lack of the characteristic signal for the C-3 carbon in 3-butyl-3-hydroxyphthalide (2) at δ = 107.9 ppm (Figure S13, Supplementary Materials). Additionally, the saturated analog (4) of 3-n-butylidenephthalide (1) was not observed during the biotransformations. Therefore, the possibility of the conversion of the compound (1) through the reduction of a double bond followed by hydroxylation was rejected.
Another proposed mechanism involves the microbial hydrolysis of 3-n-butylidenephthalide (1), keto-enol tautomerism and the closing of the lactone moiety ( Figure 6). Aldol-lactonizations may be selectively conducted by enzymes [36]; however, in the proposed mechanism, the molecule with a positive charge on C-3 carbon is formed (sp 2 hybridization). Such flat carbocation is attacked by all sides of the plane, and thus, obtainment of a racemate is probable. Therefore, as we obtained an optically active product, we excluded this mechanism as a possibility.
The most probable pathway included hydration, which is the addition of water to a double bond in the C-3 position ( Figure 6). The hydration route was also proposed by Yan et al. when they investigated the conversion pathway of ligustilide in rat plasma, as well as in later research concerning the metabolism of 3-n-butylidenephthalide (1). The researchers proposed the aromatization of ligustilide to 3-n-butylidenephthalide (1), with a further addition of water to obtain 3-butyl-3-hydroxyphthalide (2) [28,37]. Hydratases reportedly have certain advantages over synthetic catalyst-mediated enantioselective hydration, as the latter often requires harsh environmental conditions [38].
Unfortunately, to the best of our knowledge, there is no available data on phthalide hydratases. In order to exclude acid-catalyzed hydration as a possible mechanism, compound (1) was additionally dissolved in a water and organic solvent mixture (35:65) in the presence of HCl (pH = 1). Compound (2) was not observed during HPLC analysis of the sample.
Interestingly, Li et al. observed the further methylation of product (2) in rat bile as a metabolite of a Chinese herbal formula containing phthalides, including 3-n-butylidenephthalide (1) [39]. Even though 3-butyl-3-methoxyphthalide (3) (Figure 7) was not observed during biotransformations in the cultures of fungal strains, we chemically synthesized this compound. We conducted hydrolysis and a reaction with boron trifluoride of the biotransformation mixture from A. candidus AM 386. We then examined whether the incorporation of a methoxy group affects the bioactivity. The substitution of the hydroxyl group in 3-butyl-3-methoxyphthalide (3)  was not observed during the biotransformations. Therefore, the possibility of the conversion of the compound (1) through the reduction of a double bond followed by hydroxylation was rejected. Another proposed mechanism involves the microbial hydrolysis of 3-n-butylidenephthalide (1), keto-enol tautomerism and the closing of the lactone moiety ( Figure 6). Aldol-lactonizations may be selectively conducted by enzymes [36]; however, in the proposed mechanism, the molecule with a positive charge on C-3 carbon is formed (sp 2 hybridization). Such flat carbocation is attacked by all sides of the plane, and thus, obtainment of a racemate is probable. Therefore, as we obtained an optically active product, we excluded this mechanism as a possibility.
The most probable pathway included hydration, which is the addition of water to a double bond in the C-3 position ( Figure 6). The hydration route was also proposed by Yan et al. when they investigated the conversion pathway of ligustilide in rat plasma, as well as in later research concerning the metabolism of 3-n-butylidenephthalide (1). The researchers proposed the aromatization of ligustilide to 3-n-butylidenephthalide (1), with a further addition of water to obtain 3-butyl-3-hydroxyphthalide (2) [28,37]. Hydratases reportedly have certain advantages over synthetic catalyst-mediated enantioselective hydration, as the latter often requires harsh environmental conditions [38].
Unfortunately, to the best of our knowledge, there is no available data on phthalide hydratases. In order to exclude acid-catalyzed hydration as a possible mechanism, compound (1) was additionally dissolved in a water and organic solvent mixture (35:65) in the presence of HCl (pH = 1). Compound (2) was not observed during HPLC analysis of the sample.
Interestingly, Li et al. observed the further methylation of product (2) in rat bile as a metabolite of a Chinese herbal formula containing phthalides, including 3-n-butylidenephthalide (1) [39]. Even though 3-butyl-3-methoxyphthalide (3) (Figure 7) was not observed during biotransformations in the cultures of fungal strains, we chemically synthesized this compound. We conducted hydrolysis and a reaction with boron trifluoride of the biotransformation mixture from A. candidus AM 386. We then examined whether the incorporation of a methoxy group affects the bioactivity. The substitution of the hydroxyl group in 3-butyl-3-methoxyphthalide (3) was evidenced by the presence of a threeproton singlet at 3.04 ppm (H-3a) ( Figure S8

Influence of Lipophilicity and Fungistatic Activity
Xenobiotics are usually converted to more polar metabolites so that they are easily excreted from the organism. The converted compounds may exhibit pharmacological activity or become inactive [21]. We planned to determine whether there is a dependent relationship between the polarity of the metabolite (2) of 3-n-butylidenepthalide (1) and

Influence of Lipophilicity and Fungistatic Activity
Xenobiotics are usually converted to more polar metabolites so that they are easily excreted from the organism. The converted compounds may exhibit pharmacological activity or become inactive [21]. We planned to determine whether there is a dependent relationship between the polarity of the metabolite (2) of 3-n-butylidenepthalide (1) and its bioactivity. In order to achieve this, we evaluated the pharmacodynamic parameters and fungistatic properties of these compounds.
To assess the lipophilicity of phthalides and their metabolites, the chromatographic partition coefficient (log kw) and hydrophobicity index (ϕ 0 ) values were determined using RP-HPLC based on retention parameters. The partition coefficient (log P) values were calculated using theoretical methods proposed at ALOGPS 2.1.
RP-HPLC is a useful method for preliminary research into drug lipophilicity due to the small amount of compound required [40]. The method is based on the evaluation of the retention factor (k). It has been previously stated that there is a linear correlation between the log k value and the volume fraction of organic solvent [41]. This dependence, involving the log k values of five compounds, is presented in Figure 8.
its bioactivity. In order to achieve this, we evaluated the pharmacodynamic parameters and fungistatic properties of these compounds.
To assess the lipophilicity of phthalides and their metabolites, the chromatographic partition coefficient (log kw) and hydrophobicity index (φ0) values were determined using RP-HPLC based on retention parameters. The partition coefficient (log P) values were calculated using theoretical methods proposed at ALOGPS 2.1.
RP-HPLC is a useful method for preliminary research into drug lipophilicity due to the small amount of compound required [40]. The method is based on the evaluation of the retention factor (k). It has been previously stated that there is a linear correlation between the log k value and the volume fraction of organic solvent [41]. This dependence, involving the log k values of five compounds, is presented in Figure 8. The chromatographic lipophilicity parameters are presented in Table 2. Overall, high coefficient values of determination (≥0.9633) were obtained for all the examined compounds. The highest log kw value was achieved for 3-n-butylidenephthalide (1) at −3.094 for isomer (E). The metabolite of the abovementioned compound, 3-butyl-3-hydroxyphthalide (2), had a notably higher polarity (log kw = 2.397). This result confirmed the introduction of a new polar group into the compound, which is typical for phase 1 reactions [21]. Introduction of a methyl group into the structure of 3-butyl-3-methoxyphthalide (3) resulted in increased hydrophobicity (φ0 = 81.200) in comparison to (2) (φ0 = 73.079). The lack of a double bond in the side chain of 3-n-butylphthalide (4) resulted in a higher polarity (log kw = 2.7623) compared with that of compound (1). The log P values according to ALOGPS 2.1 did not exceed 3.38 for the abovementioned compounds.
It has been previously stated that 90% of compounds that reach phase 2 clinical trials do not exceed a log P value of 5 [33]. Studies conducted by Tamaian et al. (2015) on thiazolyl-carbonyl-thiosemicarbazides and thiazolyl-azoles emphasized the importance of optimum hydrophilic-lipophilic balance. These studies showed that the highest bioactivity was observed for the compounds that had medium values for the lipophilicity parameters [42]. Log P values that are too high may result in poorer drug solubility. For instance, the widely used fungistatic ketoconazole with a log P value of 3.73 reportedly belongs to class II of the Biopharmaceutics Classification System, suggesting that the compound has high permeability and low solubility [43,44]. Conversely, another antimycotic fluconazole has a log P value of 0.5 [45]. In theory, such low lipophilicity should be connected with lower membrane permeability [46]. The chromatographic lipophilicity parameters are presented in Table 2. Overall, high coefficient values of determination (≥0.9633) were obtained for all the examined compounds. The highest log k w value was achieved for 3-n-butylidenephthalide (1) at −3.094 for isomer (E). The metabolite of the abovementioned compound, 3-butyl-3hydroxyphthalide (2), had a notably higher polarity (log k w = 2.397). This result confirmed the introduction of a new polar group into the compound, which is typical for phase 1 reactions [21]. Introduction of a methyl group into the structure of 3-butyl-3-methoxyphthalide (3) resulted in increased hydrophobicity (ϕ 0 = 81.200) in comparison to (2) (ϕ 0 = 73.079). The lack of a double bond in the side chain of 3-n-butylphthalide (4) resulted in a higher polarity (log k w = 2.7623) compared with that of compound (1). The log P values according to ALOGPS 2.1 did not exceed 3.38 for the abovementioned compounds. Table 2. Chromatographic partition coefficient (log kw) and hydrophobicity index (ϕ 0 ) for substrate (1) and products of its conversion (3-butyl-3-hydroxyphthalide (2), 3-butyl-3-methoxyphthalide (3) and 3-n-butylphthalide (4)) along with the calculated partition coefficient.  It has been previously stated that 90% of compounds that reach phase 2 clinical trials do not exceed a log P value of 5 [33]. Studies conducted by Tamaian et al. (2015) on thiazolylcarbonyl-thiosemicarbazides and thiazolyl-azoles emphasized the importance of optimum hydrophilic-lipophilic balance. These studies showed that the highest bioactivity was observed for the compounds that had medium values for the lipophilicity parameters [42].
Log P values that are too high may result in poorer drug solubility. For instance, the widely used fungistatic ketoconazole with a log P value of 3.73 reportedly belongs to class II of the Biopharmaceutics Classification System, suggesting that the compound has high permeability and low solubility [43,44]. Conversely, another antimycotic fluconazole has a log P value of 0.5 [45]. In theory, such low lipophilicity should be connected with lower membrane permeability [46].

Compound
3-n-Butylidenephthalide (1) efficiently inhibited the growth of the C. albicans clinical isolates 595/20 and 38, with an MIC 50 below 50 µg/mL. It also inhibited strain 636/20, showing an MIC 50 value of 88 µg/mL. By contrast, strain 636/20 was not inhibited by the fluconazole at the highest concentration used (250 µg/mL). This phthalide was slightly less active against C. albicans ATTC 90028 and had an MIC 50 of 110 µg/mL. 3-n-Butylphthalide (4) was also significantly active against the clinical isolate C. albicans 595/20, showing an MIC 50 below 50 µg/mL. These inhibitory activities of phthalides (1) and (4) might correlate with their high lipophilicity, and thereby with their sufficient membrane permeability.
It appeared that (-)-3-butyl-3-hydroxyphthalide (2) was slightly active only in the case of strains 636/20 and 38, with MIC 50 values of 203 and 250 µg/mL, respectively. The compound did not influence the growth of other strains. The lack of activity of 3butyl-3-hydroxyphthalide (2) suggests that 3-n-butylidenephthalide (1) is subjected to an inactivation pathway during its conversion. The lack of activity of (2) corresponds with significantly lower logk w values, which may indicate that the permeability of the compound is too low. Introduction of the methoxy group in the C-3 position resulted in enhanced activity toward strain 595/20 (MIC 50 = 115 µg/mL) compared with that of the metabolite, but it did not improve its fungistatic activity toward strain 636/20 (MIC 50 = 244 µg/mL). However, the lipophilicity of (2) had a value close to that displayed by 3-n-butylidenephthalide (1). These results suggest that lipophilicity is not the main determinant of the antifungal activity of phthalides. In the case of 3-butyl-3-methoxyphthalide (3), its molecular structure and the occurrence of spatial hindrance may also influence its decreased activity.
Structure-activity dependences for phthalides have been previously proposed. Considering the dependence between the structure and fungistatic activity, available data focuses on preliminary structure-activity tests of more complex 3-substituted phthalides against selected phytopathogens. It has been stated that -OH and -NH 2 groups incorporated in the benzene ring may increase the antifungal activity of phthalides due to the formation of a hydrogen bond with the active site of the pathogen enzyme [49]. In fact, the lack of the activity of the tested compounds (2) and (3) may be explained by the -OH and -OCH 3 groups at the C-3 position. Xiao et al. showed the linkage between the structure of marine fungal phthalide derivatives and peroxisome proliferator-activated receptor gamma (PPAR-γ) binding and activation properties. This study also showed that the occurrence of an -OH group in the benzene ring has a positive effect on higher bioactivity; conversely, the presence of a double side chain at C-3 in phthalides resulted in lower binding and activation properties compared to compounds with a single chain [50].

Compounds
The substrate for the biotransformations-3-n-butylidenephthalide (1) was purchased from Sigma-Aldrich (St. Louis, MO, USA) as the mixture of (E) and (Z) isomers, at the ratio of 9:1 according to the 1 H NMR and HPLC (wavelength 274 nm).
3-butyl-3-methoxyphthalide (3) was obtained through 24 h alkaline hydrolysis of a mixture of products (80 mg) obtained in the biotransformation catalyzed by A. candidus AM 386. The biotransformation was conducted using 3 mL 0.5M KOH in MeOH, and esterification was carried out using 4 mL 20% BF 3 in methanol. After 1 h, the mixture was briefly filtered using silica gel to remove BF 3 . The mixture was then eluted by diethyl ether . Crude product was concentrated in a rotary evaporator and purified using silica gel and a hexane:ethyl acetate 19:1 (v/v) eluent. The yield of 3-butyl-3-methoxyphthalide (3) was 41.52 mg (51.9%).
3-n-Butylphthalide (4) was synthesized as follows: n-butyllithium in hexane (n-BuLi; Sigma-Aldrich, St. Louis, MO, USA) was added dropwise to the phthalic anhydride dissolved in tetrahydrofuran (7.4 g, 0.05 mol) at −78 • C (isopropanol/CO 2 (s)). The molar ratio of the anhydride to n-BuLi was 3:1. The reaction was run for 20 min and then quenched with 10% hydrochloric acid, extracted with diethyl ether, washed with water and dried with MgSO 4 . The solvent was then evaporated. The crude extract was dissolved in 30 mL THF and NaBH 4 (1.13 g, 0.03 mol; Sigma-Aldrich, St. Louis, MO, USA) was added. The reaction was stopped by adding 10% hydrochloric acid. THF was evaporated and the mixture was extracted with diethyl ether and washed with water. The organic layer was dried with MgSO 4 and collected. After concentration, the crude product was purified by column chromatography using petroleum ether:acetone 3:1 (v/v) as eluent. The yield of 3-n-butylphthalide (4) was 2.12 g (67%). The progress of the reactions was monitored by analytical TLC and GC. The structures of the compounds (3,4) were confirmed by NMR spectroscopy.

Microorganisms
The following strains were used in the biotransformations: Mucor spinosus AM 398, Absidia cylindrospora AM 336, Pycnidiella resinae AM 50, Penicillium chrysogeum AM 112, Chaetomium indicum AM 158, Fusarium culmorum AM 9, Aspergillus candidus AM 386, Laetisporus sulphurens AM 515, Fusicoccum amygdali AM 258 and Ascosphaera apis AM 496. These were all obtained from the Department of Chemistry collection at the Wrocław Uni-versity of Environmental and Life Sciences. These strains were maintained on Sabouraud or Czapek agar slants at 4 • C.
Fungistatic activity was detected against Candida albicans ATCC 90028 from American Type Culture Collection, and its clinical isolates Candida albicans 636/20, Candida albicans 595/20 and Candida albicans 38 were obtained from Wrocław Medical University, Poland.

Biotransformations
The biotransformations were conducted in 300 mL Erlenmeyer flasks. Each strain was inoculated using the sterile Sabouraud medium (75 mL), which consisted of 30 g glucose (Chempur, PiekaryŚląskie, Poland), 10 g bactopeptone (Biocorp, Warszawa, Poland) and 1 L distilled water (pH 6.88). After 5 days of incubation at 25 • C on a rotary shaker, 3-nbutylidenephthalide (1) was added (20 mg dissolved in 0.5 mL of acetone) and samples were collected every 2 days to measure pH values. Product yield was determined on days 2, 4, 6, 8 and 14. The samples were acidified to pH 5 using hydrochloric acid (if necessary) and extracted using ethyl acetate. The samples were then dried using MgSO 4 and subjected to evaporation at 25 • C. Next, the samples were dissolved in methanol, filtered using a 0.45 µm PTFE filter and analyzed by HPLC.
Three microorganisms were chosen for the scale-up process: A. cylindrospora AM 336, C. indicum AM 158 and A. candidus AM 386. The preparative biotransformations were conducted in 2000 mL Erlenmeyer flasks using 500 mL of a sterile Sabouraud medium and precultivated inoculum, which constituted 10% of the medium volume. After 5 days of incubation at 25 • C on a rotary shaker, 3-n-butylidenephthalide (1) was added (150 mg dissolved in 1 mL acetone) to the culture. The samples were collected every 2 days and the pH values were measured. The samples were prepared as previously described (except for the evaporation step) and subjected to GC analysis in ethyl acetate. Product mixtures were collected on day 8 of the process for A. candidus AM 386 and on day 14 for C. indicum AM 158 and A. cylindrospora AM 336. The collected samples were acidified with 10% hydrochloric acid, extracted twice using ethyl acetate and analyzed by RP-TLC.
The product was purified using 250 µm thick TLC Silica gel 60 RP-18 F254s plates (Merck, Darmstadt, Germany) with 70% MeCN and 30% H 2 O acidified by 1% HCOOH eluent. After separation, the plates were visualized using a UV lamp at 254 and 365 nm. The product was scraped off and extracted with ethyl acetate, after which the organic layer was collected, dried with MgSO 4 and evaporated by the rotary evaporator. The structures of the compounds were determined by NMR.

Analysis
TLC analysis was performed on glass plates covered with silica gel 60 F254 (Merck) and developed using hexane:acetone 2:1 (v/v) eluent. After elucidation, the plates were visualized by a solution of 1% Ce(SO 4 ) 2 and 2% phosphoromolybdic acid in 10% H 2 SO 4 .

Lipophilicity
The concentration of tested compounds (1-4) was 0.1 mg/mL. Dead time (t 0 ) was measured by injecting 1% aqueous NaNO 3 solution . The injection volume was 10 µL, the temperature was 35 • C and the flow rate was 1.5 mL/min. The retention time (t R ) of the samples was also assessed by RP-HPLC in triplicate using the abovementioned column and isocratic elution. The eluents consisted of methanol and water, both of which were acidified by 1% formic acid v/v at 60-80% of the organic phase. The retention factor (k) was calculated according to the following equation: k = (t R − t 0 )/t 0. A graph of log k as a function of the volume fraction of organic solvent was plotted. The value of the chromatographic lipophilicity index, log k w (log k value with 0% methanol in the mobile phase), was determined by extrapolating this correlation. The chromatographic hydrophobicity index ϕ 0 was also assessed as the volume fraction of methanol that yields a log k value of 0 [31]. Theoretical log P (partition coefficient) values were calculated with the use of ALOGPS 2.1. Pearson's correlation coefficients between the log P and log k w were determined using GraphPad Prism, and differences with p < 0.05 were considered statistically significant.

Fungistatic Activity
Compounds (1-4) were tested against four Candida albicans strains (C. albicans ATTC 90028 and its clinical isolates C. albicans 636/20, C. albicans 595/20 and C. albicans 38) using the broth microdilution technique. The medium used for the tests was YPD, consisting of 20 g glucose (Chempur, PiekaryŚląskie, Poland), 20 g bactopeptone (Biocorp, Warszawa, Poland), 10 g yeast extract (BTL, Łódź, Poland) and 1 L distilled water; the pH was adjusted to 6.5. The solutions of the compounds were prepared in dimethyl sulfoxide (DMSO) and diluted in YPD to obtain final concentrations in the range of 50-250 µg/mL. Fluconazole was tested at the range of 0.064-64 µg/mL with the exception of the resistant strain 636/20, which was additionally tested in the range of 50-250 µg/mL. Next, 100 µL of each solution was pipetted into wells of a 96-well microtiter plate. The inoculum was standardized to 0.5 McFarland standard and then diluted to obtain the final suspension with a cell density of 0.5 × 10 3 to 2.5 × 10 3 CFU/mL. The inoculum size used was 100 µL. The positive control comprised DMSO added in the same concentration as the tested compounds in the inoculum, while the negative control was DMSO diluted in the broth without the addition of inoculum. All samples were tested at least in triplicate. The microtiter plates were incubated at 35 • C for 24 h in a Biosan PST-60 HL ThermoShaker (Riga, Latvia) at 1000 rpm. The fungistatic activity of the compounds was assessed by measuring the absorbance at a wavelength of 595 nm (Epoch, BioTek, Winooski, VT, USA) to determine the MIC 50 value (i.e., the concentration of a compound required to inhibit the growth of 50% of microorganisms).

Conclusions
In this study, we confirmed the utility of microbially catalyzed biotransformation for obtaining the mammalian metabolite (2) of bioactive 3-n-butylidenephthalide (1). We show that microbiological transformations can efficiently produce the target compound and can be used to better understand phthalide lactone metabolism. Overall, 10 fungal strains were tested, five of which efficiently produced hydroxy-3-butylphthalide (2). We proposed three mechanisms by which 3-n-butylidenephthalide (1) conversion may occur. Even though hydroxylation remains popular in xenobiotic metabolism, we excluded this pathway based on 13 C NMR spectra and suggested the hydration pathway as the most probable mechanism. Valuable contributions of this research include the analysis of the lipophilicity and fungistatic activity of both phthalides (1) and (4) and the metabolite (2). Antifungal assays revealed the potential of both 3-n-butylidenephthalide (1) and 3-n-butylphthalide (4) against clinical isolates of C. albicans, with an MIC 50 value below 50 µg/mL. We did not observe the influence of a double bond-containing side chain on the activity of compounds (1) and (4). We noticed weak or even a lack of inhibitory properties for metabolite (2), which has lower lipophilicity and presumably lower permeability than compounds (1) and (4). We did not observe a straightforward correlation between lipophilicity and fungistatic activity. Apart from the lipophilicity, the spatial structure of compounds is the factor that influences their biological activity. The presence of -OH groups in compound (2) or -OCH 3 groups in compound (3) causes the decrease of antifungal activity compared to that of 3-n-butylphthalide (4). Compound (1), which has a double bond in its structure, shows similar polarity to 3-butyl-3-methoxyphthalide (3); however, this compound had high biological activity. Considering the possible applications for phthalides, there is a necessity for further research, particularly regarding the biological activity of their metabolites.