Effects of pH and Cultivation Time on the Formation of Styrene and Volatile Compounds by Penicillium expansum

Styrene can be formed by the microbial metabolism of bacteria and fungi. In our previous study, styrene was determined as a spoilage marker of Fuji apples decayed by Penicillium expansum, which is responsible for postharvest diseases. In the present study, P. expansum was cultivated in potato dextrose broth added with phenylalanine—which is a precursor of styrene—using different initial pH values and cultivation times. Volatile compounds were extracted and analyzed using gas chromatography-mass spectrometry (GC-MS) combined with stir-bar sorptive extraction. The 76 detected volatile compounds included 3-methylbutan-1-ol, 3-methyl butanal, oct-1-en-3-ol, geosmin, nonanal, hexanal, and γ-decalactone. In particular, the formation of 10 volatile compounds derived from phenylalanine (including styrene and 2-phenylethanol) showed different patterns according to pH and the cultivation time. Partial least square-discriminant analysis (PLS-DA) plots indicated that the volatile compounds were affected more by pH than by the cultivation time. These results indicated that an acidic pH enhances the formation of styrene and that pH could be a critical factor in the production of styrene by P. expansum. This is the first study to analyze volatile compounds produced by P. expansum according to pH and cultivation time and to determine their effects on the formation of styrene.


Introduction
Styrene has been found in diverse foods such as beef meat, cereals, coffee beans, fruits, apple-based alcoholic beverages, and wheat beer [1][2][3]. The presence of styrene in foods can adversely affect their aroma due to its strong pungent and unpleasant odor [4]. Styrene can originate from food packaging materials [5] as well as the natural metabolism of raw agricultural materials [1]. It can also be formed by the microbial metabolism of bacteria [6] and fungi [7]. Several studies have investigated the production of styrene by fungi such as Pichia carsonii [8], Fusarium oxysporum [7], Penicillium citrinum [9], and Penicillium expansum [10]. Among them, P. expansum is a filamentous fungus that is widely found in certain types of spoiled fruits such as apples and plums [11] and is well known to produce styrene [12]. This fungus is responsible for the blue mold that is a major postharvest disease of apples [11,13]. This disease, related to styrene formation, can also result in off-flavors of processed apple products [14].

49
The formation of volatile compounds by a fungus can be affected by various cultivation 50 conditions, including the culture medium composition, temperature, and pH [16]. In particular,

51
Miao et al. determined that the pH of the cultivation medium could significantly affect the formation 52 of secondary volatile compounds [16]. Furthermore, Lee et al. determined that the production of 53 volatile compounds by Saccharomycopsis fibuligera KJJ81 depends on the cultivation time [17].

54
However, there has been only one report on the effects of various culture media on the formation of 55 volatile compounds by P. expansum [18]. That study found that cultivating P. expansum on various 56 media such as pine leaves, pine stems, pine wood, mature dark bark, and potato dextrose broth 57 (PDB) resulted in the production of different volatile compounds, including styrene.

58
While some previous studies have investigated the formation of styrene by P. expansum, the 59 critical effects of culturing conditions have not been elucidated. Therefore, the objectives of this

71
In particular, the well-known fungal volatile compound 3-methylbutan-1-ol [12,20] was detected at a 72 higher level than other alcohols throughout the cultivation period. Other 3-methyl branched-chain 73 volatiles such as 3-methyl butanal were also detected. Both 3-methylbutan-1-ol and 3-methyl butanal 74 are commonly generated from leucine [19,20]. Reduction by alcohol dehydrogenase can convert 75 3-methyl butanal into 3-methylbutan-1-ol [21]. The levels of these volatile compounds derived from 76 leucine were higher at pH 9 than at pH 5, which suggests that the leucine metabolism of P. expansum 77 was more strongly activated at an alkaline pH.  2) Retention indices were determined using n-alkanes C7-C30 as external standards on a Stabilwax column.

82
3) Mean values of relative peak area to that of internal standard ± standard deviation from three replicates.   5) Identification of the compounds was based on the following: A, mass spectrum and retention index agreed with those of authentic compounds under the same conditions (positive identification); B, mass spectrum and retention index were consistent with those from NIST (National Institute of Standards and Technology) database (tentative identification) and literatures [22][23][24]; C, mass spectrum was consistent with that of W9N08 (Wiley and NIST) and manual interpretation (tentative identification). 6) Significant differences (p < 0.05) between samples according to cultivation time using Duncan's multiple comparison test. 7) N.D. = not detected.
Aldehydes can be produced from various precursors such as amino acids, carbohydrates, and fatty acids [39]. Some aldehydes such as decanal, nonanal, and hexanal were detected at higher levels than the other aldehydes. Korpi et al. found nonanal to be one of the main microbial volatile aldehydes in laboratory culture experiments, although it was not reported in field samples [40]. Hexanal, which is a straight long-chain aldehyde, can be formed from long-chain fatty acids such as palmitic acid and stearic acid via enzymatic oxidation [41]. In addition, hexan-1-ol can be converted reversibly into hexanal by alcohol dehydrogenase [42]. The level of hexanal was higher than that of hexan-1-ol in all of the present cultivation samples.
Most ketones are generated by lipid oxidation via β-oxidation of free fatty acids during microbial metabolism. Some ketones such as octan-3-one, 6-methylhept-5-en-2-one, and 5-hexyloxolan-2-one (γ-decalactone) were detected in this study, with octan-3-one only being identified at pH 5. This ketone has a musty and mushroom odor note and is reportedly a microbial volatile organic compound [43] that can be formed via the aerobic oxidation of linolenic acid and linoleic acid [41]. The precursors of γ-decalactone included oleic acid, linoleic acid, and other unsaturated fatty acids. In the first of three steps, ricinoleic acid is formed through the hydroxylation of oleic acid. Then, 4-hydroxy decanoic acid is formed via the reduction of ricinoleic acid from acetyl CoA (acetyl coenzyme A). The last step is lactonization, in which 4-hydroxy decanoic acid is converted into γ-decalactone [44,45].
A particularly interesting finding of this study was that the level of styrene was significantly elevated throughout the cultivation period at pH 5, whereas this tendency was not observed at pH 9. Other volatile compounds derived from phenylalanine also showed characteristic patterns of formation according to pH and cultivation time. Therefore, this study compared the contents of volatile compounds derived from phenylalanine in P. expansum according to pH and cultivation time.  degradation by phenylalanine ammonia-lyase (PAL). Fungi can then participate in the conversion of cinnamic acid into styrene by cinnamic acid decarboxylation [46]. The amount of styrene produced 145 in the present study was significantly higher after 24 hours of cultivation at pH 5 than after 16 hours.

146
Additionally, the level of cinnamic acid peaked after 16 hours of cultivation at pH 5 and thereafter 147 tended to decrease. It seems that cinnamic acid-which is a highly efficient precursor of styrene-is 148 rapidly converted into styrene as soon as cinnamic acid is synthesized. In particular, Penicillium 149 strains are well known to have the ability to form styrene from cinnamic acid [8,47]. This means that high at an alkaline pH, and, accordingly, the production of phenylalanine peaks [49]. As a result, an 160 alkaline pH could result in the decreased production of styrene.
another major volatile compound derived from phenylalanine. First, 2-phenylacetaldehyde is 163 produced via the decarboxylation and deamination of phenylalanine, and then 2-phenylethanol is 164 biosynthesized from 2-phenylacetaldehyde by phenyl acetaldehyde reductase [47]. The amount of 165 2-phenylethanol formed was considerably greater at pH 9 than at pH 5 in the present study. Many Phenylalanine can be converted into cinnamic acid as the primary product of phenylalanine degradation by phenylalanine ammonia-lyase (PAL). Fungi can then participate in the conversion of cinnamic acid into styrene by cinnamic acid decarboxylation [46]. The amount of styrene produced in the present study was significantly higher after 24 hours of cultivation at pH 5 than after 16 hours. Additionally, the level of cinnamic acid peaked after 16 hours of cultivation at pH 5 and thereafter tended to decrease. It seems that cinnamic acid-which is a highly efficient precursor of styrene-is rapidly converted into styrene as soon as cinnamic acid is synthesized. In particular, Penicillium strains are well known to have the ability to form styrene from cinnamic acid [8,47]. This means that PAL, which can convert phenylalanine into cinnamic acid, might be a critical enzyme for the formation of styrene. In addition, Pagot et al. reported that the synthesis of styrene by PAL was strongly activated during the exponential phase in Penicillium strains [48]. The peaking of cinnamic acid after 16 hours (in the exponential phase) at pH 5 could therefore be related to the synthesis of a considerable amount of PAL. Accordingly, the formation of styrene by P. expansum was elevated at pH 5. On the other hand, both styrene and cinnamic acid were detected at much lower levels at pH 9 than at pH 5, and their levels did not increase significantly with the cultivation time. This also could be related to the activity of PAL, which is a reversible enzyme. The activity of PAL can be markedly affected by pH. The ability of PAL to convert cinnamic acid into phenylalanine (reverse reaction) is high at an alkaline pH, and, accordingly, the production of phenylalanine peaks [49]. As a result, an alkaline pH could result in the decreased production of styrene.
Moreover, 2-phenylethanol, which has fruity, floral, and rose-like odor notes [50][51][52][53], was another major volatile compound derived from phenylalanine. First, 2-phenylacetaldehyde is produced via the decarboxylation and deamination of phenylalanine, and then 2-phenylethanol is biosynthesized from 2-phenylacetaldehyde by phenyl acetaldehyde reductase [47]. The amount of 2-phenylethanol formed was considerably greater at pH 9 than at pH 5 in the present study. Many bacteria and fungi respond to a high extracellular pH by synthesizing deaminase that hydrolyzes amino acids [53,54]. Furthermore, Ghosh et al. identified that an alkaline pH enhances the production of aromatic alcohols [54]. Those authors found that the formation of three aromatic alcohols (tryptophol, 2-phenylethanol, and tyrosol) by Candida albicans was threefold higher under an alkaline condition. Accordingly, the production of a large amount of 2-phenylethanol at pH 9 in P. expansum in the present study could have been induced by the alkaline pH. In addition, other volatile compounds derived from phenylalanine such as phenyl acetaldehyde, 2-phenylacetonitrile, benzaldehyde, acetophenone, benzoic acid, and cinnamaldehyde were also detected. Among them, benzaldehyde was detected at a higher level at pH 5 than at pH 9, and its level peaked after 16 hours of cultivation. On the other hand, 2-phenylacetonitrile derived from phenylacetaldehyde was only detected at pH 5, which might have been due to all phenylacetaldehydes being converted into 2-phenylethanol at pH 9. In summary, volatile compounds derived from phenylalanine produced by P. expansum could be considerably affected by the extracellular pH and cultivation time.
Partial least square-discriminant analysis (PLS-DA) was conducted to determine the differences in volatile compounds produced by P. expansum and the significant effects of pH and cultivation time on the formation of volatile compounds. Figure 2  bacteria and fungi respond to a high extracellular pH by synthesizing deaminase that hydrolyzes 167 amino acids [53,54]. Furthermore, Ghosh et al. identified that an alkaline pH enhances the 168 production of aromatic alcohols [54]. Those authors found that the formation of three aromatic 169 alcohols (tryptophol, 2-phenylethanol, and tyrosol) by Candida albicans was threefold higher under 170 an alkaline condition. Accordingly, the production of a large amount of 2-phenylethanol at pH 9 in benzaldehyde was detected at a higher level at pH 5 than at pH 9, and its level peaked after 16 hours detected at pH 5, which might have been due to all phenylacetaldehydes being converted into Partial least square-discriminant analysis (PLS-DA) was conducted to determine the differences      All of the samples at pH 5 and 9 were located on the positive and negative PLS 1 axes, respectively, while all of the samples cultivated for 16 and 32 hours were located on the positive and negative PLS 2 axes, respectively. As the cultivation time increased, the samples moved along PLS 2. Tables 2 and 3 list the major volatile compounds (with a criterion of the variable importance plot (VIP) > 0.8) identified in P. expansum. The negative PLS 1 axis was related to most of the aldehydes and alcohols, while the positive PLS 1 axis was related to some benzenes such as styrene, benzaldehyde, and 1,3,5-trimethylbenzene, while styrene was also associated with the negative PLS 2 axis. These results demonstrated that the formation of styrene could be considerably influenced by an acidic pH and a longer cultivation time in P. expansum. In addition, 2-phenylethanol was positioned on the negative PLS 1 axis, which indicates that it could be affected by an alkaline pH in P. expansum. In addition, Figure 2 shows that the formation of volatile compounds by P. expansum, including styrene, could be affected more by the pH than by the cultivation time.

Strain and Cultivation of Penicillium expansum
P. expansum HR5-4 was isolated from naturally decayed apples. P. expansum was identified as previously reported [55] and cultivated in 40 mL of PDB media contained 0.1% phenylalanine. One milliliter of spore suspension (10 7 spores/mL) of P. expansum was inoculated in a 250 mL Erlenmeyer flask with screw cap and placed in a shaking incubator (Vision Scientific Co., Ltd., Bucheon-si, Gyeonggi-do, Korea) at 25 • C and 180 rpm. P. expansum was cultivated at different cultivation times (16,24, and 32 hours) and initial pH (pH 5 and pH 9). Each cultivation time was determined by growth phase of P. expansum (16 hours: Exponential phase, 24 and 32 hours: Stationary phase). Initial pH of media was adjusted by using 0.1 M HCl and NaOH.

Extraction of Volatile Compounds
After vacuum filtration, aliquots (8 mL) of cultivation media were transferred into a 10 mL glass vial (Agilent Technologies, Santa Clara, CA, USA). Volatile compounds were extracted using a polydimethylsiloxane-coated stir bar (PDMS twister 10 mm length, 1.0 mm film thickness) (GERSTEL GmbH and Co. KG, Mülheim an der Ruhr, Germany). The PDMS twister was placed in samples and stirred at 1000 rpm and ambient temperature for 60 min. After extraction, PDMS twister was washed with distilled water and dehydrated with lint-free tissue paper. Then, PDMS twister was placed in desorption liner tube (GERSTEL GmbH and Co.) and inserted into a thermal desorption unit (TDU). Volatile compounds were desorbed by increasing the temperature of the TDU. The initial temperature of 40 • C was maintained for 0.5 min. After that, the temperature increased at a rate of 120 • C/min to 220 • C and held for 5 min. With cooled injection system (CIS), cryo-focusing temperature was maintained at −60 • C using liquid N 2 supply and the temperature of the CIS-4 PTV (Programmed Temperature Vaporizer) was increased thermally to 250 • C at a rate of 10 • C/sec and held for 3 min. The temperatures of transfer line and ion source were 280 and 250 • C, respectively. The splitless mode was used as injection mode.

GC-MS Analysis
Volatile compounds were analyzed using an Agilent 7890B gas chromatograph (GC) connected to a 5977A mass detector (Agilent Technologies, Santa Clara, CA, USA). A Stabilwax column (30 m length × 0.25 mm internal diameter × 0.25 µm film thickness, Restek Corporation, Bellefonte, PA, USA) was used. The oven temperature was initiated at 40 • C (5 min), increased to 220 • C at a rate of 4 • C/min and then held at 220 • C (10 min). Helium was used as carrier gas at a constant flow rate of 0.8 mL/min. The mass spectrum was obtained in EI (electron ionization) mode at 70 eV, mass scan rate of 4.5 scans/sec, and mass scan range of 35-350 m/z. In addition, the analysis of some volatile compounds, which could be derived from phenylalanine, were conducted using selective ion monitoring (SIM) mode. The list of volatile compounds measured and the SIM qualifying ions are presented in Table 4. Identification of the compounds was based on the following: A, mass spectrum and retention index agreed with those of authentic compounds under the same conditions (positive identification); B, mass spectrum and retention index were consistent with those from NIST (National Institute of Standards and Technology) database (tentative identification) and literatures [22][23][24]; C, mass spectrum was consistent with that of W9N08 (Wiley and NIST) and manual interpretation (tentative identification).

Identification and Semiquantification of Volatile Compounds
Volatile compounds were positively identified by comparing their mass spectral data and retention index (RI) values with those of authentic standard compounds. Otherwise, the other volatile compounds, whose authentic standard compounds were not available, were tentatively identified by comparing with the mass spectral libraries (Wiley 9th edition and NIST, version 08) and retention index (RI) values in published literatures [52][53][54]. The retention index (RI) values were calculated using n-alkanes from C 7 to C 30 as external standards. The semiquantification of volatile compounds was calculated by their peak areas to that of internal standard compound. Five microliters of 2,3,5-trimethyl pyrazine (100 mg/L in methanol) were used as an internal standard.

Statistical Analysis
Partial least square-discriminant analysis (PLS-DA) was performed to determine the differences of volatile compounds of P. expansum according to pH and cultivation time and the significant effect on the formation of volatile compounds using SIMCA-P (version 11.0, Umetrics, Umea, Sweden). Analysis of variance (ANOVA) was also conducted with general linear model procedure in SPSS (version 12.0, Chicago, IL, USA) to evaluate significant differences of volatile compounds in samples. In order to evaluate significant differences (p < 0.05), Duncan's multiple range test was conducted.

Conclusions
This study investigated volatile compounds produced by P. expansum according to pH and cultivation time. A total of 76 volatile compounds such as 3-methylbutan-1-ol, 3-methyl butanal, oct-1-en-3-ol, geosmin, nonanal, hexanal, and γ -decalactone were detected. In particular, the formation of volatile compounds derived from phenylalanine such as styrene showed characteristic patterns according to pH and cultivation time. In particular, the level of styrene was considerably higher at pH 5 than at pH 9. Moreover, as cultivation time increased, the production of styrene significantly increased at pH 5. On the other hand, styrene was detected at much lower level at pH 9 than at pH 5, and also its level was not significantly increased according to cultivation time. On PLS-DA plots, the formation of volatile compounds of P. expansum was more highly affected by pH condition than by cultivation time. As a result, the cultivation pH could be a critical factor in the production of styrene in P. expansum. This study is a first report on the analysis of volatile compounds according to pH and cultivation time and determines their effects on the formation of styrene in P. expansum.