Biodiesel synthesis by a one-step method in a genetically engineered Escherichia coli using rice straw hydrolysate and restaurant oil wastes as raw materials



Yuhua Zhao, Institute of Microbiology, College of Life Sciences, Zhejiang University, Hangzhou 310058, China.




To reduce the cost of biodiesel, the utilization of rice straw hydrolysate and restaurant oil wastes as the substrates for biodiesel production in a recombinant strain by a one-step method was investigated.

Methods and Results

A recombinant Escherichia coli pET28a (+)-PAW, which encodes enzymes for the ethanol pathway and acyltransferase, was constructed. Fermentation results in Luria-Bertani medium revealed that xylose was favourable for fatty acid ethyl esters (FAEEs) production (2·79 g l−1) by E. coli pET28a (+)-PAW. When rice straw hydrolysate and restaurant oil wastes were utilized as the monosaccharide substituent and the fatty acids substituent, 0·27 g l−1 FAEEs and 0·249 g l−1 FAEEs were obtained, respectively. 1·27 g l−1 FAEEs was obtained in the restaurant oil wastes medium supplemented by 2% sodium oleate.


The fermentation results indicated that the strain was effective in biodiesel production. The rice straw hydrolysate and restaurant oil wastes could be utilized by E. coli as the substrates for FAEE production.

Significance and Impact of the Study

This novel exploration might pave the way for biodiesel production from waste materials by genetically engineered microorganism.


Developing alternative resources of energy as a substitute for traditional fossil fuels has attracted much interest in recent years (El Diwani et al. 2009). Among alternative fuels, biodiesel is considered to be a renewable, nontoxic and biodegradable diesel fuel. Biodiesel consists of the alkyl monoesters of fatty acids with short-chain alcohols, such as fatty acid methyl esters (FAMEs) and fatty acid ethyl esters (FAEEs; Steinbuchel et al. 2006; Ghanei et al. 2011). Today, more than 95% of the world's biodiesel is produced from edible vegetable oils (Gui et al. 2008), which results in the production of biodiesel costing approximately 1·5 times that of traditional diesel (Olutoye and Hameed 2011). New sources of cheaply available fatty acids and sugar substitutes are necessary to reduce the cost of biodiesel production (Elbahloul and Steinbuchel 2010).

More than 15 million tons of waste oil are generated annually throughout the world (Gui et al. 2008). Restaurant oil waste disposal is a world-wide environmental problem, and the problem is of particular concern in China because more than 4·5 million tons of waste oil are generated each year from restaurants, food processing industries and fast food shops (Zhang et al. 2012). The large amounts of restaurant oil wastes are usually dumped into rivers and landfills in China (Meng et al. 2008). Generally, restaurant oil wastes that are rich in free fatty acids (FFAs) can be utilized as the substrates of biodiesel because the physical and chemical properties of restaurant oil wastes are similar to those of fresh edible oil. The exploration of strategies to utilize restaurant oil wastes for biodiesel production offers significant advantages, such as enhancing the economic viability of biodiesel, not competing with the food market and maintaining the health of the environment (Chhetri et al. 2008; do Nascimento et al. 2011). Several published studies have investigated biodiesel produced from restaurant oil wastes by means of base- or acid-catalysed one-step transesterification reactions and base-catalysed two-step transesterification reactions (Felizardo et al. 2006). No previous studies are available to date that provide information on biodiesel production by genetically engineered microorganisms from restaurant oil wastes. In this study, restaurant oil wastes were explored as fatty acid sources for biodiesel production by the fermentation of engineered Escherichia coli. E. coli has been proven to be a biodiesel producer after genetic engineering. A previous report by Steen et al. (2010) found that genetically engineered E. coli produced fatty acid-derived molecules from hemicelluloses or glucose. Another engineered E. coli TOP10 pMicrodiesel constructed by Steinbuchel et al. (2006) yielded 0·26 g l−1 FAEEs under submerged conditions using glucose as the carbon source. Genetic engineering can give E. coli the ability to produce biodiesel by introducing two enzymes that are necessary for ethanol production and one enzyme that encodes acyltransferase. Because ethanol was produced indirectly from sugar glycolysis, the exploration of cheap carbohydrates as the sugar source is essentially important for further development (Tsigie et al. 2011). Lignocellulosic materials as inexpensive materials for microbial oil production have proven attractive in recent years (Angerbauer et al. 2008). Using these residues as substitutes for the sugar source in E. coli-mediated biodiesel production is promising.

The purpose of this study was to explore the feasibility of rice straw hydrolysate and restaurant oil wastes as raw materials for biodiesel production by a recombinant E. coli. Particularly, a lower-cost one-step method using microbial fermentation was constructed for FAEE synthesis from wastes.

Materials and methods

Strains, plasmids and culture condition

Zymomonas mobilis ZM4 (CICC10273) and Acinetobacter baylyi strain ADP1 (BD413; ATCC33305) were obtained from the China Center of Industrial Culture Collection and the American Type Culture Collection, respectively. E. coli DH5α and E. coli BL21 (DE3) were used as the hosts for plasmid amplification and expression, respectively. Plasmids pMD19-T (Takara, Dalian, China) and pET28a (+) (Invitrogen, Shanghai, China) were used as vectors. Z. mobilis ZM4 was cultured in medium that consisted of 10% glucose, 0·5% yeast extract, 0·1% KH2PO4, 0·1% (NH4)2SO4, 0·05% MgSO4 and an additional 2% agar when needed. The E. coli and the recombinant strains were grown in Luria-Bertani (LB) medium at 37°C overnight with vigorous shaking. Ampicillin (100 μg ml−1) or kanamycin (50 μg ml−1) was added to the medium when needed. Where indicated, sodium oleate was added from a 10% (w/v) stock solution in H2O to a final concentration of 0·2% (w/v). If not stated otherwise, the pH was controlled at 7·0 with adjustment using 1 mol l−1 HCl or 1 mol l−1 NaOH.


Restriction endonucleases, Taq DNA polymerase and PrimeSTAR HS DNA polymerase were purchased from Takara, Dalian, China. T4 DNA ligase was purchased from New England Biolabs Ltd (Beijing, China). The gel extraction kit was obtained from Axygen, China. The oligonucleotide primers were synthesized by Invitrogen. Ethyl oleate and other ethyl esters utilized as reference substances for FAEEs were purchased from Meitech Technology Ltd (Wuhan, China). Nonesterified fatty acid (NEFA) detection kit was purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). All the other reagents were of analytical grade.

DNA manipulation and transformation

Extraction of chromosomal DNA and plasmid DNA, enzyme reactions, and transformation of E. coli cells were carried out as described by Sambrook and Russell (2001).

Construction of the recombinant strain

The oligonucleotides primers used for PCR amplification are list in Table 1. The genes encoding pyruvate decarboxylase (pdc) and alcohol dehydrogenase (adh) were amplified from the total genomic DNA of Z. mobilis ZM4 with the primers P1–P4. Gene atfA, which encodes wax ester synthesis/acyl coenzyme A: diacylglycerol acyltransferase (WS/DGAT), was amplified from the total genomic DNA of Acinetobacter sp with the primers P5-P6. PCR was performed under the following conditions: 98°C, 2 min; 30 cycles of 98°C, 10 s, 55°C, 15 s, 72°C, 2 min, with a final extension at 72°C for 5 min.

Table 1. Primers used for DNA amplification and cloning. Restriction sites are underlined
PrimersSequence (5′-3′)











CT (SalI)

Overlap extension PCR was used to obtain the fusion gene containing pdc and adh with the primers P7–P10. The target segment, which contained the fusion gene and the atfA gene, was amplified by the overlap extension PCR with the primers P11–P14. The PCR products were separated by electrophoresis and cloned into the pMD 19-T vector for sequencing according to standard methods (Sambrook and Russell 2001). The purified PCR product of the target segment was digested with EcoRI and SalI and ligated into pET28a (+), which had been digested with the same restriction enzymes. The expression vector for the recombinant protein was used to transform E. coli BL21 (DE3). To express the recombinant protein, the recombinant strain was cultivated in a 100 ml LB medium containing 50 μg ml−1 kanamycin at 37°C and grown to an OD600 of 0·5, followed by induction with 0·1 mmol l−1 isopropyl-ß-D-thiogalactopyranoside (IPTG) at 23°C overnight in a shaking incubator at 130 rev min−1.

Fermentation in the LB medium with glucose or xylose

In a 250-ml flask, 1% size of inoculum was added to the 100 ml LB medium containing 0·2% (w/v) sodium oleate as FFAs, 2% (w/v) glucose or 2% (w/v) xylose as the sugar source, 0·1 mmol l−1 IPTG and appropriate antibiotics for plasmid selection. The media were sampled to determine the pH, sugar concentration, cell dry mass, OD600, FAEEs yield and ethanol yield. For the quantification of the FAEEs produced by the recombinant strain, 100 ml of culture broth was ultrasonically extracted with 100 ml of a chloroform/methanol mixture (2 : 1, v/v) for 30 min. The mixture was centrifuged at 15 000 g for 10 min at room temperature, and the organic phase was withdrawn carefully. The ultrasonic extraction was repeated two times, and the organic phases from the same sample were combined together. The solvent was removed by evaporation from the combined extracts, and the residual material was redissolved in 2 ml of the chloroform/methanol mixture (2 : 1, v/v). FAEEs and ethanol were analysed by gas chromatography, external standard method (SP-6800; Lunan Ruihong, Shandong, China), equipped with an OV-225 capillary column (30 m × Φ 0·32 mm × 0·5 μm). Fermentation experiments were repeated three times. All samples were analysed in triplicate, and the mean values were calculated. The intracellular lipid droplets accumulating in the recombinant strain, which was cultivated in LB medium containing 2% (w/v) xylose, were analysed using transmission electron microscopy (TEM; JEM-1230, JEOL).

Fermentation with rice straw hydrolysate

The rice straw was obtained from Yongan, Fujian Province, China, and the preparation of the hydrolysate was according to a modified method of Huang et al. (2009). The materials were smashed to <0·5 mm and then mixed with dilute sulphuric acid (1·5%, v/v) to give a mixture with a solid loading of 10% (w/v). The mixture was treated in an autoclave at 121°C for 90 min, and the liquid fraction was separated by vacuum filtration after cooling. The pH of the hydrolysate was controlled at 10·0 by the addition of Ca(OH)2. After 1 h, the hydrolysate was filtrated under vacuum, acidified to pH 5·5 with 5 mol l−1 H2SO4 and filtrated again 1 h after acidification. Then, 0·1 g l−1 Na2SO3 was added to the filtrate, followed by boiling at 100°C for 15 min. After that, the hydrolysate was recovered by vacuum filtration. Furthermore, activated charcoal was used as the adsorbent of the hydrolysate by the modified method of Huang et al. (2011). The total sugar concentration of rice straw hydrolysate was measured using the 3, 5-dinitrosalicylic acid (DNS) method. Fermentation experiments were repeated three times. All samples were analysed in triplicate, and the mean values were calculated. The recombinant strain was cultivated in the 100 ml LB medium, with 0·2% sodium oleate as the FFAs composition and the hydrolysate and in which the sugar concentration was diluted to 20 g l−1.

Sugar analysis by high-performance liquid chromatography (HPLC)

A total of 200 μl of fermentation liquor was mixed with 200 μl of 0·3 mol l−1 sodium hydroxide and 200 μl of 0·3 mol l−1 methanolic solution of 1-phenyl-3-methyl-5-pyrazolone (PMP). The whole mixture was heated to 70°C and incubated for 30 min. After the reaction, the mixture was cooled to room temperature and neutralized with 200 μl of 0·3 mol l−1 hydrochloric acid. The mixture was shaken vigorously for 1 min after chloroform (800 μl) was added to the solution and then centrifuged at 15 000 g for 1 min. The chloroform layer after centrifugation was discarded, and the extraction process was repeated three times. The supernatants were filtered through a 0·22-μm pore membrane filter for HPLC (Agilent 1200 series, Santa Clara, CA, USA) analysis using a C18 reversed phase column (5 μm, 4·6 × 250 mm; Agilent). The mobile phase was a mixture of 0·1 mol l−1 ammonium acetate buffer (pH 5·5) and acetonitrile at the ratio of 77 : 23 (v/v), and the flow rate was 1 ml min−1. The wavelength of UV-detection was set to 245 nm, and the column temperature was fixed at 35°C. All sugar samples were analysed in triplicate, and the mean values were calculated.

Fermentation with restaurant oil wastes

Restaurant oil wastes used in this study were obtained from the cafeteria at Zijingang Campus, Zhejiang University. The raw restaurant oil wastes were filtered through a 200-mesh sieve to remove all the insoluble impurities and stored at −20°C. The fatty acid composition of the raw restaurant oil wastes and the FAEEs extracted from the recombinant strain were determined using a gas chromatography-mass spectrometry (Focus-GC-DSQ II) equipped with a DB-5MS column (30 × 0·25 mm, 0·25 nm; Agilent). The concentration of FFA in the restaurant oil wastes was analysed using the NEFA detection kit. The recombinant strain was cultivated in the 100 ml of restaurant oil waste medium with different concentrations of glucose or xylose added (Table 2). The media were sampled to determine pH, sugar concentration, cell dry mass, OD600, FAEE yield and ethanol yield. Fermentation experiments were repeated three times. All samples were analysed in triplicate, and the mean values were calculated.

Table 2. Fatty acid ethyl ester (FAEE) production from fermentation by the recombinant strain with restaurant oil wastes as the fatty acid substitute
ParametersFermentation medium
  1. 1: 0% sugar; 2: 2% glucose; 3: 2% xylose;4: 1% glucose and 1% xylose; 5: 4% xylose; 6: 2% glucose and 4% xylose; 7: 4% glucose and 2% xylose.

Cell dry mass (CDM g l−1)3·958·33·34·046·65·15
Sugar conversion (%)39·345·220·224·930·224·424·9
Ethanol concentration (g l−1)1·051·911·811·594·975·113·38
FAEEs concentration (g l−1)0·0470·0720·2490·0800·1230·1100·236
FAEEs content (% of CDM)1·21·43·02·43·11·74·6

Statistical analysis

Correlation analysis of the fermentation data in the restaurant oil wastes medium was performed using MATLAB (ver. R2009a; The Mathworks Inc., Natick, MA, USA) software.


Construction of the recombinant strain

Three relevant genes, pdc, adh and atfA, were cloned into a high-copy-number plasmid pET28a (+) for biodiesel production. A 2·8-kb DNA fragment consisting of pdc and adh of Z. mobilis was amplified by overlap extension PCR, yielding a fusion gene that could produce ethanol. The expressed fusion protein PDC/ADH was analysed by SDS-PAGE, and a clear single band with molecular weight of approximately 100 kDa appeared when induced by IPTG. Both the PDC activity and the ADH activity of the fusion protein were detected, indicating that the fusion protein could be used for further research (data not shown). Furthermore, a 4·5-kb DNA fragment was amplified by combining the fusion gene with an additional T7 promoter and atfA using overlap extension PCR. Subsequently, this fragment was digested with EcoRI and SalI and ligated into pET28a (+) that had been digested with the same restriction enzymes, yielding pET28a (+)-PAW (Fig. 1). The plasmid pET28a (+)-PAW was transformed into E. coli BL21 (DE3), yielding the recombinant E. coli pET28a (+)-PAW. The recombinant plasmid carried three genes relevant for FAEE synthesis in a collinear orientation, with pdc and adh driven by a T7 promoter and with atfA controlled by a second T7 promoter, thereby ensuring effective transcription of all the three genes.

Figure 1.

Map of plasmid pET28a (+)-PAW. Relevant characteristics: Kan, kanamycin-resistance gene; T7, T7 promoter; fusion gene, pyruvate decarboxylase gene and alcohol dehydrogenase gene from Zymomonas mobilis ZM4; atfA, wax ester synthase/acyl-CoA/diacylglycerol acyltransferase gene from Acinetobacter baylyi strain ADP1.

Biodiesel production in the LB medium with sugar and fatty acids

The recombinant strain was cultivated in LB medium with 2% glucose and 0·2% sodium oleate and resulted in significant FAEE and ethanol formation, which was confirmed by gas chromatography. Of the glucose, 32·8% was consumed in the growth phase before induction, while 66·5% of the glucose was consumed during the phase of biodiesel biosynthesis. A cell dry mass of 6·4 g l−1 was reached at the end of the cultivation (Fig. 2a). The maximum yield of ethanol reached 8·12 g l−1 (0·1 g l−1 h−1) in the medium at 84 h, and the highest yield of FAEEs, 1·81 g l−1 (0·08 g l−1 h−1), was obtained at 24 h (Fig. 2b). The cellular FAEE content, described as milligram of biodiesel amount/cell dry mass in percentage (Steinbuchel et al. 2006), reached up to 62% during cultivation (Fig. 2b). The conversion of the substrate, described as cell dry mass per milligram of glucose assumption in percentage, was 72%.

Figure 2.

Fermentation of the recombinant Escherichia coli pET28a (+)-PAW with glucose (a,b) and xylose (c,d), respectively. △, pH; ▲, OD600; ■, cell dry mass (g l−1); ●, fatty acid ethyl esters (FAEEs) (g l−1); ○, ethanol (g l−1); ▼, sugar concentration (g l−1).

The effect of xylose on FAEE production by the recombinant strain was also investigated by adding 2% xylose into the LB medium with 0·2% sodium oleate. The xylose concentration declined from 20 to 6·9 g l−1 over culturing (Fig. 2d), and the glucose as the sugar source in the medium exhibited a similar trend. Of the xylose, 12·9% was consumed before being induced, while 60·3% of the xylose was consumed during the phase of biodiesel biosynthesis. The rate of xylose utilization was 19·9% lower than that of glucose in the strain growth phase. The conversion of the substrate as xylose to cell dry mass was 44%. The maximum ethanol production with glucose as the carbon source was 2·18 times higher than that with xylose. However, the 2·79 g l−1 yield of FAEEs (0·06 g l−1 h−1) at 48 h with xylose as the sugar source was 1·54 times higher than the FAEE yield with glucose, and 3·4 g l−1 of the cell dry mass was detected at this time (Fig. 2c). The cellular FAEE content in the xylose medium could reach 82·2% during cultivation (Fig. 2d). An evident intracellular accumulation of lipid droplets could be clearly observed by TEM in the recombinant E. coli pET28a (+)-PAW cells after induction by 0·1 mmol l−1 IPTG under the xylose condition (Fig. S1a), while no lipid droplets were found in the cells that were not induced (Fig. S1b).

FAEE production using the rice straw hydrolysate as s a substitute for sugar source

A maximum FAEE yield of 0·27 g l−1 (0·01 g l−1 h−1) was obtained, while a cell dry mass of 7·4 g l−1 was accumulated by 48 h in the rice straw hydrolysate medium (Fig. 3a). The yield of FAEEs was much lower than the 1·81 g l−1 yield of FAEEs in the LB medium with glucose as the sugar source. The cellular FAEE content in the rice straw hydrolysate medium reached up to 3·7% during the cultivation. It is worth noting that the utilization ratio of total sugar was 93%. The time courses of cell growth and biodiesel accumulation in the recombinant strain are shown in Fig. 3a. From 0 to 18 h, the synthesis of biodiesel was barely detectable. Then, the yield of biodiesel increased quickly because of the utilization of the sugar source and fatty acids, for example, xylose, from 18 to 48 h. After this increase, a clear decline in biodiesel content was observed, whereas the biomass increased continuously. To obtain a better understanding of the sugar utilization during the fermentation process, the sugar composition in the hydrolysate was analysed by HPLC (Fig. 3b). In addition to xylose and glucose, there were different kinds and varying amounts of other sugar sources in the rice straw hydrolysate, such as mannose, galactose, rhamnose and arabinose, which could explain why the total sugar concentration (10·93 g l−1) was higher than the concentration of xylose plus glucose (4·71 g l−1). The glucose residue was only 1·4%, compared to the 41·7% xylose residue during the growth phase. Furthermore, the utilization of xylose (98·1%) was significantly faster than that of glucose (77·5%) during the FAEE production phase.

Figure 3.

Fermentation of the recombinant Escherichia coli pET28a (+)-PAW in rice straw hydrolysate medium. △, pH; ●, cell dry mass (g l−1); ○, fatty acid ethyl esters (FAEEs) (g l−1); ◆, FAEEs content (%); ▲, xylose concentration (g l−1); ▼, glucose concentration (g l−1); ■, sugar concentration (g l−1).

FAEE production using the restaurant oil waste as a substitute for fatty acids

The concentration of FFA in the restaurant oil wastes was in the range of 5841·969–6212·435 μmol l−1. The fermentation results of the different media with restaurant oil wastes are presented in Table 2. Maximum sugar conversion (45·2%) was obtained with the 2% glucose treatment, but only 0·072 g l−1 of FAEEs was obtained in this treatment. The maximum ethanol yield reached 5·11 g l−1 (0·06 g l−1 h−1) in the 2% glucose and 4% xylose treatment, although the treatment only yielded 0·11 g l−1 of FAEEs. To investigate the effect of the initial amount of FFA on the production of FAEEs, an additional 2% sodium oleate was added to the same treatment, yielding 1·27 g l−1 of FAEEs, which was 11·5 times higher than the FAEEs yield (0·11 g l−1) obtained when no sodium oleate was added. This indicated the increase in the amount of FFA in the medium could promote the production of FAEEs. Fermentation of 2% xylose with the restaurant oil wastes led to the FAEEs yield of 0·249 g l−1 (0·007 g l−1 h−1) but resulted in the lowest sugar conversion (20·2%). The ethanol yield was positively and significantly correlated with the sugar content in the fermentation medium by Pearson correlation analysis (R = 0·844, P = 0·017). Three quadratic equations were obtained by regressing the sugar conversion, the ethanol concentration and the FAEE yield with the content of glucose and xylose in the fermentation medium as follows:

display math(1)
display math(2)
display math(3)

where Y1, Y2 and Y3 are sugar conversion, the ethanol concentration and the FAEE yield, respectively, and x1 and x2 are the glucose concentration and xylose concentration in the medium, respectively. All the determination coefficients in the three nonlinear regression models were above 0·98, indicating that the models could explain over 98% of the total variability in the data. The coefficients of Eqn (1) indicated a negative effect of xylose on the sugar conversion, but xylose was favourable for FAEE production, as shown in Eqn (3). Glucose was found to have positive influence on ethanol production by the analysis of Eqn (2).

Gas chromatography-mass spectrometry (GC-MS) analysis of fatty acids of the restaurant oil wastes and FAEEs

The FFA composition of the restaurant oil wastes and the components of FAEEs extracted from the recombinant strain in the restaurant oil wastes media were shown in Table 3. Hexadecanoic acid (26·25%), octadecadienoic acid (24·65%) and oleic acid (14·39%) dominated the crude fatty acid components of the restaurant oil wastes. In addition, the restaurant oil wastes consisted of a minor amount of dodecanoic acid (1·12%), tetradecanoic acid (3·28%) and palmitelaidic acid (1·56%). However, there were few FAEEs, such as hexadecanoic acid ethyl ester (0·22%) and 9,12-octadecadienoic acid ethyl ester (0·80%), in the restaurant oil wastes. The analysis of FAEEs extracted from E. coli BL21 pET28a (+)-PAW cultivated in the restaurant oil waste medium revealed a mixture of E-11-hexadecenoic acid ethyl ester (4·85%), hexadecanoic acid ethyl ester (10·84%), 9,12-octadecadienoic acid ethyl ester (13·38%) and (E)-9-octadecenoic acid ethyl ester (12·49%).

Table 3. Gas chromatography–mass spectrometry analysis of the free fatty acids (FFA) composition of the restaurant oil wastes and fatty acid ethyl esters (FAEEs) extracted from the recombinant strain
 ComponentsFormulaContent (%)
FFADodecanoic acidCH3(CH2)10COOH1·12
Tetradecanoic acidCH3(CH2)12COOH3·28
Palmitelaidic acidCH3(CH2)5CH=CH(CH2)7COOH1·56
Hexadecanoic acidCH3(CH2)14COOH26·25
9,12-Octadecadienoic acidCH3(CH2)4CH=CH(CH2)2(CH2)6COOH24·65
Oleic acidCH3(CH2)7CH=CH(CH2)7COOH14·39
Octadecanoic acidCH3(CH2)16COOH5·99

Dodecanoic acid,

ethyl ester


Tetradecanoic acid,

ethyl ester


E-11-Hexadecenoic acid,

ethyl ester


Hexadecanoic acid,

ethyl ester

9,12-Octadecadienoic acid,ethyl esterCH3(CH2)4CH=CH(CH2)2(CH2)6COOCH2CH313·38

(E)-9-Octadecenoic acid

ethyl ester



Efficient ethanol biosynthesis was achieved in E. coli by the heterologous expression of the recombinant proteins PDC and ADH because PDC was able to convert pyruvic acid to aldehyde and ADH could convert aldehyde to ethanol (Ingram et al. 1987; Alterthum and Ingram 1989). In addition, WS/DGAT (the atfA gene product) was confirmed to be capable of utilizing ethanol to some extent as an acyl acceptor substrate (Kalscheuer et al. 2004; Stoveken et al. 2005). Therefore, fermentation by E. coli pET28a (+)-PAW that expressed these three genes resulted in FAEE production with a maximum yield of 2·79 g l−1. Previously, an E. coli TOP10 that contained a similar plasmid pMicrodiesel produced 0·26 g l−1 of FAEEs after 48 h under aerobic conditions during batch fermentation using glucose and sodium oleate as substrates, and 1·28 g l−1 of FAEEs after 72 h when the fed-batch strategy was employed (Steinbuchel et al. 2006). The E. coli C41 (DE3), engineered by Steen et al. (2010), produced 0·674 g l−1 of FAEEs from hemicellulose. The FAEE yield in the present study with LB media was absolutely higher than the previously reported results, suggesting the promising value of the constructed strain.

Although xylose was found to be unfavourable for cell growth, more effective FAEE synthesis was found in the fermentation medium with xylose as sugar source than when glucose was the sugar source. The strain that utilized xylose as the carbon source consumed more energy than glucose for the sugar glycolysis pathway. Because glucose and xylose are two dominating sugars in lignocellulosic hydrolysates, the sufficient utilization of these sugars is required for economical biofuel production (Yu et al. 2011). The construction of recombinant E. coli capable of xylose utilization (Yomano et al. 2008) opens up new opportunities for developing a Zymomonas-based process for the conversion of corn stover to ethanol. In the present study, the recombinant E. coli pET28a (+)-PAW capable of FAEE production from both xylose and glucose paved the way for lignocellulose waste utilization.

Rice straw hydrolysate was selected as the substitute for sugar source for biodiesel production in this study. The rice straw hydrolysate has been used as feedstock in several biotechnological applications, that is, for bioethanol and lipid production using yeast (da Cunha-Pereira et al. 2011; Yu et al. 2011). However, the yield of biodiesel in the rice straw hydrolysate medium was significantly lower than in the LB medium containing glucose or xylose. There were different kinds and varying amounts of unfavourable compounds, such as acetic acid, 5-hydroxymethyl furfural (HMF) and furfural in hydrolysate (Roberto et al. 1996), which have been observed to inhibit the growth of microorganisms (Palmqvist et al. 1999) and to reduce biodiesel production. The results suggested that detoxification pretreatment to remove the inhibitors in the hydrolysate should be effective for increasing the biodiesel production with the hydrolysate. The utilization of glucose was faster than that of xylose as a result of carbon catabolite repression in microorganisms, which means that glucose was preferred over xylose for microbial growth and proliferation. However, the xylose was found to be more favourable for FAEE synthesis by the recombinant strain. The use of rice straw hydrolysate for biodiesel production provided an attractive alternative for the utilization of the renewable materials and avoided the use of highly toxic methanol in traditional industrial process of biodiesel.

Biochemical characterization of WS/DGAT has revealed that this acyltransferase exhibits a remarkably broad substrate range in vitro (Steinbuchel et al. 2006). Because the concentration of FFA in the restaurant oil wastes in the present study was in the range of 5841·969–6212·435 μmol l−1, which was similar to 0·2% sodium oleate in the LB medium. Thus, FAEE production by E. coli pET28a (+)-PAW was conducted in restaurant oil wastes, which are rich in fatty acids. In the restaurant oil wastes medium mixed with 2% xylose, 0·249 g l−1 of FAEE was the maximum yield that could be obtained when no sodium oleate was added. Regression analysis results further confirmed that xylose was favourable for biodiesel accumulation in this recombinant strain, although the utilization of xylose was lower than that of glucose in the growth phase, and the strain consumed more energy when utilizing xylose. The GC-MS analysis of fatty acids in the restaurant oil wastes and FAEEs indicated that some fatty acids had not been transesterified to ethyl esters in the restaurant oil wastes medium. This result was not surprising because many of the metabolic processes need specific fatty acids for lipid metabolism during growth. The significantly increased amounts of hexadecanoic acid ethyl ester and 9, 12-octadecadienoic acid ethyl ester likely resulted from the transesterification of hexadecanoic acid and octadecadienoic acid in the restaurant oil wastes to the final products as ethyl esters. Although the FAEE production was found to be far below what is needed for an industrial process, the recombinant strain produced FAEEs by one step without any soap and water inhibition compared to traditional catalysts. This new strategy has the advantages of no toxic wastewater, low equipment cost and no environmental pollution, in contrast to conventional FAEE production (Okuda et al. 2007; Nieves et al. 2011).

This work demonstrated that conversion of rice straw hydrolysate and restaurant oil wastes to FAEEs by E. coli pET28a (+)-PAW could be a promising alternative for biodiesel production.


This study was supported by the International Cooperation Project in Science and Technology of Zhejiang Province (2008C14038), the National Natural Science Foundation of China (31070079) and the Science and Technology Project of Zhejiang Province (2008C13014-3, 2011C13016)