1、Introduction
2-Phenylethanol (2-PE) is an aromatic alcohol, which can be naturally found in essential oils of many plants, such as hyacinths, jasmine and lilies. 2-PE has been widely used in cosmetics, perfume and food industries owning to its delicate rose scent (Scognamiglio, Jones, Letizia, & Api, 2012). Furthermore, 2-PE is also an important raw material for the derivatives synthesis, among which phenylethyl acetate is a valuable fragrance compound, and p-hydroxyphenylethanol is widely used in pharmaceutical and fine chemicals industries (Masuo, Osada, Zhou, Fujita, & Takaya, 2015; Yin et al., 2015). Although 2-PE can be extracted from flowers and plants, however, the extremely low concentration hinders it application (Feng et al., 2015). Alternatively, 2-PE can be chemically synthesized, however, its quality is greatly affected because of the harsh condition and toxic reagents used (Etschmann, Bluemke, Sell, & Schrader, 2002). Currently, the price of 2-PE produced through natural routes, such as extraction from rose petals or bio-converted from renewable resources is approximately USD 1,000/kg. However, it is only USD 5.0/kg for chemically synthesized 2-PE from benzene and styrene (Hua & Xu, 2011). Therefore, bio-synthesis of 2-PE has become an appealing option owning to its environmentally friendly property and mild conditions.
In nature, many wild-type microorganisms have been identified and characterized to be capable of producing 2-PE, most of which are from eukaryotes, including Saccharomyces sp., Kluyveromyces marxianus, Yarrowia lipolytica, Aspergillus oryzae, and Pichia sp., etc (Huang, Lee, & Chou, 2001; Masuo et al., 2015). In addition, some prokaryotes have also been reported to produce 2-PE, such as Microbacterium foliorum, Proteus vulgaris, and Psychrobacter sp., 2-PE can be either converted from L-phenylalanine (L-phe) through three steps catalysis of Ehrlich pathway or produced from glucose through Shikimate pathway. As multiple steps are needed for 2-PE production through Shikimate pathway, lower concentration usually occurred when 2-PE was directly synthesized from glucose. Accordingly, Ehrlich pathway is thought to be more promising for 2-PE production. It should be noticed that the lipophilic 2-PE could make the lipid membrane structure a preferential binding target, resulting in the collapse of transmembrane gradients and consequently the loss of cell viability (Sikkema, de Bont, & Poolman, 1995), hence, strain development including mutation, selection, or genetic modification has been comprehensively adopted to improve the final 2-PE production. For instance, the newly isolated K. marxianus CCT 7735 could generate 3.44 g/L of 2-PE through Ehrlich pathway under optimized conditions (Azevedo, Santos, Vieira, Gomes, & Batista, 2018). S. cerevisiae BY4741, which overexpressed ARO10 and contained an aro8Δ deletion could produce 96 mg/L of 2-PE from glucose (Shen, Nishimura, Matsuda, Ishii, & Kondo, 2016). S. cerevisiae SPO810, in which ARO8 and ARO10 were co-expressed could finally produce 2.61 g/L of 2-PE with fed batch fermentation  (Yin et al., 2015).
In the present study, novel 2-PE-generating microbes were first isolated and characterized. To further improve 2-PE production, statistical design of experimental strategy and in situ extraction strategy were used to improve final 2-PE titer. Transcriptome analysis on genes involved in 2-PE synthesis were also identified and characterized.
 
2、Material and methods
2.1. Isolation and molecular identification of strain YLG18
Soil samples from Xuanwu Lake, Nanjing, China were used as inocula in order to screen bacteria capable of producing 2-PE. The soil samples were added into flasks containing 50 mL of defined mineral salts medium (MMT) spiked with 40 g/L of glucose as the carbon source. After 96 h incubation at 30 °C and a shaking speed of 200 rpm, 0.5 mL culture was spread on a Petri dish containing MMT medium, which was conducted in aseptic environment. Then different colonies were picked into mineral salts medium using L-phe as the sole nitrogen source for detection of 2-PE production capabilities. Eventually, a 2-PE-producing bacterium named YLG18 was obtained. Unless stated otherwise, the strain was grown in mineral salts medium with L-phe as the sole nitrogen source at 30 °C.
The defined mineral salts medium contained (per liter of distilled water): glucose 40 g, L-phe 5 g, NaCl 1.0 g, MgCl2·6H2O 0.5 g, KH2PO4 0.2 g, KCl 0.3 g, CaCl2·2H2O 0.015 g, TES 2.292 g, uracil 0.1 g, thiamine 3 mg. In addition, 1 mL of trace element solution and 10 mL of salts solution were added to 1 L medium. Then the medium was dispended into triangle shaker, which were wrapped with gauze, autoclaved for 20 minutes and cooled down to room temperature.
Genomic DNA of the cultures was extracted and purified with DNeasy tissue kit (Qiagen, Germany) according to manufacturer’s instructions. The genomic DNA was used as a template for PCR amplification of the 18S rRNA gene with a pair of universal fungi primer ITS1 (5′-CCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′). The obtained PCR products were purified with a PCR purification kit (Qiagen, Germany) and sequenced using an ABI DNA sequencer (Applied Biosystems, USA). The 18S rRNA gene sequence was aligned using the BLAST algorithm and was deposited in the GenBank database with an accession number of WHVZ00000000.
2.2. Optimization of fermentation conditions
All batch fermentation studies were carried out in 250 mL triangle shakers containing 50 mL of mineral salts medium with initial pH of 5.8. The medium-filled triangle shakers were wrapped with gauze and aluminum foil before being autoclaved. L-phe was first filter-sterilized and then added into the medium as the sole nitrogen source. Inoculation was carried out by adding 2% of seed culture (strain YLG18) to the medium. The bottles were incubated at 30 °C and 200 rpm in a shaking incubator for 96 h. Detailed optimization studies regarding the production of 2-PE were carried out by using the response surface methodology (RSM). Three factors (temperature, initial glucose concentration, and initial L-phe concentration) were chosen as independent variables, while 2-PE concentration was the dependent variable. An experimental strategy based on central composite design (CCD) was obtained using Design Expert version 7.0 (Stat-Ease, Minneapolis, USA). The values of the response variable obtained were fitted to a second-order polynomial equation:
Yi=β0+∑βi xi+∑βiixi2+∑βijxixj
where Yi is the predicted response, xi, xj are independent variables, which influence the dependent variable Y; β0 is the offset term; βi is the ith linear coefficient; βii is the ith quadratic coefficient and βij is the ijth interaction coefficient. Statistical analysis of the model was performed by using analysis of variance (ANOVA) in a statistical software package.
Subsequently, the optimized condition obtained from RSM was applied to batch fermentation in a 3.0-L bioreactor (BIOSTAT® B plus, Sartorius, Germany). The bioreactor was filled with 2.0 L of mineral salts medium and operated at 30 °C with an agitation rate of 200 rpm.
2.3 Analytic method
Fermentation broth samples were analyzed for biomass growth, glucose and L-phe utilization, and 2-PE production. Biomass was determined by measuring optical density at 600 nm with appropriate dilution using a UV–visible spectrophotometer (Lambda-25, Perkin-Elmer, USA). Glucose was analyzed by a 1260 Series HPLC system (Agilent Technologies Inc.) equipped with an Aminex HPX-87H column (BioRad, Richmond, CA, USA) and a Refractive Index Detector (RID). Samples were run at 75 °C with 0.6 mL/min eluent of 5 mM sulfuric acid. Substrate (L-phe) and product (2-PE) were measured by a 1260 Series HPLC system (Agilent Technologies Inc.) equipped with an AcclaimTM 120-C18 column (Thermo Scientific, China) and UV-detector at 210 nm. 50% sterile water and 50% (v/v) methanol were pumped isocratically through Agilent 1260 quat pump at 0.6 mL/min. The column was kept at temperature of 30℃. Five-point standard curves were obtained by running standard solutions containing L-phe and 2-PE.
2.4. In situ product removal (ISPR) for 2-PE production
Extractive bioconversion was carried out with the addition of different organic solvents as extractants. These extractants were added directly with an aqueous/organic ration of 1:1 (v/v) when the bioconversion started. Controls were conducted without addition of any solvent. The bioconversion was carried out at 180 rpm, 30 ºC in 500-mL flask. At the end of the bioconversion, samples were collected and 2-PE concentration in both organic and aqueous phases was measured (Stark, 2003).
2.5. qRT-PCR-based validation and analysis
qRT-PCR was conducted to verify the expression of genes potentially involved in the biosynthesis pathway. Total RNA was extracted from strain YLG18 cultured after 48 h using the FastPure Cell/Tissue Total RNA Isolation Kit (Vazyme, Nanjing, China) and reverse-transcribed into cDNA using the HiScript II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China). PCR primers were designed with SnapGene 1.1.3 based on the nucleotide sequence of reference and antioxidant-related genes from the RNA-Seq data (Table 1). qRT-PCR was performed using ChammQTM SYBR qPCR Master Mix (High ROX Premixed) (Vazyme, Nanjing, China) and an ABI StepOnePlus (Applied Biosystems). qRT-PCR was performed with the follow thermocycling parameters: 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s and 55 °C for 30 s. Triplicate analyses of each cDNA sample were performed, and the relative expression levels of genes in each group were normalized to 18SrRNA expression. The 2−ΔΔCt comparative threshold cycle (Ct) method was used to evaluate the relative expression levels of target genes (Livak & Schmittgen, 2000). The values reported represent the average of 3 biological replicates (Perdiguero et al., 2012).
3、Results and discussion