feruloyl esterase using agricultural residues as substrates

Journal Pre-proofs Efficient ferulic acid and xylo-oligosaccharides production by a novel multimodular bifunctional xylanase/feruloyl esterase using agricultural residues as substrates Ruonan Wang, Jinshui Yang, Jin Myong Jang, Jiawen Liu, Yu Zhang, Liang Liu, Hongli Yuan PII: DOI: Reference:

S0960-8524(19)31717-1 https://doi.org/10.1016/j.biortech.2019.122487 BITE 122487

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

2 November 2019 19 November 2019 21 November 2019

Please cite this article as: Wang, R., Yang, J., Jang, J.M., Liu, J., Zhang, Y., Liu, L., Yuan, H., Efficient ferulic acid and xylo-oligosaccharides production by a novel multi-modular bifunctional xylanase/feruloyl esterase using agricultural residues as substrates, Bioresource Technology (2019), doi: https://doi.org/10.1016/j.biortech. 2019.122487

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Ltd.

1

Efficient ferulic acid and xylo-oligosaccharides production by a novel

2

multi-modular bifunctional xylanase/feruloyl esterase using

3

agricultural residues as substrates

4

Ruonan Wanga, Jinshui Yanga, Jin Myong Janga,b, Jiawen Liua, Yu Zhanga, Liang Liua,

5

Hongli Yuana,*

6

a

7

Ministry of Agriculture, College of Biological Sciences, China Agricultural University,

8

Beijing, China.

9

b

State Key Laboratory of Agrobiotechnology and Key Laboratory of Soil Microbiology,

School of Lifesciences, Kim Il Sung University, Pyongyang, Democratic People's

10

Republic of Korea

11

Ruonan Wang: [email protected]

12

Jinshui Yang: [email protected]

13

Jin Myong Jang: [email protected]

14

Jiawen Liu: [email protected]

15

Yu Zhang: [email protected]

16

Liang Liu: [email protected]

17

*Corresponding author: Hongli Yuan E-mail address: [email protected]

1

18 19

Abstract Liberating high value-added compounds ferulic acid (FA) and

20

xylo-oligosaccharides (XOSs) from agricultural residues is a promising strategy for the

21

utilization of lignocellulose. In this study, a bifunctional xylanase/feruloyl esterase from

22

bacterial consortium EMSD5 was heterogeneously expressed in Escherichia coli.

23

Depending on the inter-domain synergism of the recombinant enzyme rXyn10A/Fae1A,

24

high yields of FA (2.78, 1.82, 1.15 and 7.31 mg/g substrate, respectively) were obtained

25

from 20 mg in-soluble wheat arabinoxylan, de-starched wheat bran, ultrafine-grinding

26

corn stover and steam-exploded corncob. Meanwhile, 3.210, 1.235, 1.215 and 0.823 mg

27

xylose/XOSs were also released. For cost-saving enzyme production, we firstly

28

constructed a recombinant E. coli, which could secrete the bifunctional

29

xylanase/feruloyl esterase out of cells. When the recombinant E. coli was cultured in

30

medium containing 200 mg de-starched wheat bran, 474 μg FA and 18.2 mg

31

xylose/XOSs were also detected. Hence, rXyn10A/Fae1A and the recombinant strain

32

showed great applied potential for FA and XOSs production.

33

Keywords: Bifunctional xylanase/feruloyl esterase; Inter-domain synergism; Ferulic

34

acid; Agricultural residues; Extracellular secretion

2

35 36

1. Introduction As the main kinds of hydroxycinnamic acids, ferulic acid (FA) and p-coumaric acid

37

(p-CA) have been widely used in pharmaceutical, cosmetics and food industries due to

38

their anti-oxidant and anti-inflammatory biological activities (Oliveira et al., 2019;

39

Wang et al., 2016b). The content of FA in agricultural residues such as wheat bran and

40

maize bran ranged from 0.5% to 3.0% (w/w), which could be used as cheaper raw

41

materials for preparing FA (Long et al., 2018). Compared with alkali-extract method,

42

enzymatic hydrolysis was an environmental-friendly alternative to produce FA from

43

lignocellulose (Nieter et al., 2016). And extensive research revealed that FA is a key

44

recalcitrant component in grass lignocellulose and thus impedes biomass

45

saccharification (Oliveira et al., 2015). Therefore, depolymerizing biomass by enzymes

46

not only produces FA but also increases the saccharification of biomass.

47

Feruloyl esterases (FAEs) are key accessory enzymes for hemicellulose degradation,

48

which hydrolyze the ester-bonds between polysaccharides and FA (Wong, 2006). FAEs

49

from different microorganisms have been purified and applied to release FA from

50

agricultural residues, such as de-starched wheat bran (Cao et al., 2015; Uraji et al.,

51

2014), corn stover (Zhang et al., 2013), corn cob (Li et al., 2011), wheat straw (Cheng et

52

al., 2012) and sugarcane bagasse (Damasio et al., 2013), etc. However, due to the

53

complex structure of hemicellulose, the most reported researches showed that FAEs

54

could release FA from lignocellulose only in the presence of xylanase (Nieter et al.,

55

2016; Sang et al., 2011; Zhang et al., 2013). When xylanase was added in hydrolysis

3

56

mixtures, yields of FA were 4.2- to 47-fold of that obtained with feruloyl esterases from

57

de-starched wheat bran (Topakas et al., 2004; Wu et al., 2017). It is speculated that

58

xylanases may firstly cleave the xylan main-chain and produce feruloylated

59

xylo-oligosaccharides (FXOSs); then FAEs may remove FA from FXOSs and release

60

prebiotic xylo-oligosaccharides (XOSs) (Oliveira et al., 2019). Considering the cost of

61

multiple enzymes production, the efforts were made to construct chimeric bifunctional

62

xylanase/feruloyl esterase named FLX by fusing a xylanase (XYNB) to a feruloyl

63

esterase (FAEA). FLX released 1-fold more FA from de-starched wheat bran (DSWB)

64

than the mixture of individual XYNB and FAEA (Levasseur et al., 2005). These results

65

demonstrated that the physical proximity of xylanase and feruloyl esterase domain in

66

bifunctional enzyme generates an enhanced synergy on the degradation of complex

67

substrates. Furthermore, bifunctional xylanase/feruloyl esterase could alone liberate

68

high-value-added products FA and XOSs from lignocellulose, which makes the

69

bioconversion of biomass more economically feasible.

70

To date, only three natural bifunctional xylanase/feruloyl esterase enzymes have

71

been reported, including cellulosomal XynY and XynZ from Clostridium thermocellum

72

(Blum et al., 2000) and Xyn10D-Fae1A from anaerobic rumen bacterium Prevotella

73

ruminicola 23 (Dodd et al., 2009). However, no application tests of XynY and

74

Xyn10D-Fae1A on natural substrates were investigated. Although XynZ has been used

75

in co-production of XOSs and phenolic compounds from sugarcane bagasse (Mandelli

76

et al., 2014), the efficiency of hydrolysis was limited. Thus, exploring more efficient

4

77

bifunctional xylanase/feruloyl esterase enzymes still remains a considerable concern.

78

In our previous study, a stable hemicellulase-producing bacterial consortium

79

EMSD5 was isolated from compost soil (Lv et al., 2008). Its corn stover-induced

80

extracellular metaproteome contained a protein (48211) with the highest abundance at

81

the beginning of cultivation, indicating 48211 plays a pioneering role in lignocellulose

82

degradation by EMSD5 (Zhu et al., 2016). Moreover, 48211 contained xylanase and

83

esterase domains according to the annotation in dbCAN database. In this study, protein

84

48211 was heterogeneously expressed in Escherichia coli and the recombinant

85

bifunctional enzyme was applied to produce FA and XOSs from agricultural residues

86

such as DSWB, ultrafine-grinding corn stover (UGCS) and steam-exploded corncob

87

(SECC). The synergistic effect of its inter-domain in the release of FA and XOSs from

88

complex substrate was also investigated. To simplify the enzyme purification process,

89

an E. coli extracellular secretion expression system of the bifunctional xylanase/feruloyl

90

esterase was successfully constructed for the first time. Finally, the recombinant E. coli

91

was directly used in the production of FA and XOSs from DSWB.

92

2. Materials and methods

93

2.1 Strains, plasmid, chemicals and culture conditions

94

Escherichia coli DH5α was used for gene cloning and plasmid maintenance. E. coli

95

BL21 (DE3) was used for gene expression with plasmid pET30a. These strains were

96

cultured using Luria-Bertani (LB) broth at 37°C. Kanamycin was added to the medium

5

97

at a final concentration of 50 μg/mL when it was required. Cultivation of the microbial

98

consortium EMSD5 followed the method described previously (Lv et al., 2008). Methyl

99

ferulate (MFA) was purchased from Alfa Aesar (China) Development Co., Ltd. Methyl

100

p-coumarate (MpCA) and methyl caffeate (MFA) were purchased from TCI (Shanghai)

101

Development Co., Ltd. Methyl sinapinate (MSA) was purchased from Carbosynth

102

Beijing Laboratories. Ethyl ferulate (EFA) was purchased from Yuanye (Shanghai,

103

China). p-nitrophenyl acetate (pNPAc) was purchased from Sigma-Aldrich (China).

104

Beechwood xylan (BWX), wheat arabinoxylan (WAX) and insoluble wheat

105

arabinoxylan (I-WAX) were purchased from Megazyme (Wicklow, Ireland). Bagasse

106

xylan (BX) was purchased from GENERAY (Shanghai, China). Q5® High-Fidelity

107

DNA Polymerase and Q5 Site-Directed Mutagenesis Kit were purchased from New

108

England BioLabs (Massachusetts, USA). Restriction enzymes were purchased from

109

Takara Biomedical Technology (Beijing) Co., Ltd. T4 DNA ligase was purchased from

110

Promega (Madison, USA).

111

2.2 Agricultural residues

112

Corncob, collected from Henan Province, was milled through a 50-mesh sieve. The

113

steam-exploded corncob (SECC) was prepared as the method of Liu et al. (2019).

114

Wheat bran was purchased from the local market (Beijing, China). De-starched wheat

115

bran (DSWB) was prepared according to the method of Xu et al. (2019). Corn stover

116

was collected from the Shangzhuang experimental station of China Agricultural

117

University. The ultrafine-grinding corn stover (UGCS) was prepared as the method of

6

118

Li et al. (2019).

119

2.3 Gene cloning and expression

120

According to the method published earlier (Zhu et al., 2016), the metagenomic

121

DNA of EMSD5 were extracted using a TIANamp DNA Extraction Kit. The extracted

122

metagenomic DNA was used as template to amplify the gene sequences encoding

123

rXyn10A/Fae1A and its truncated mutants using the specific primers. Then the PCR

124

products were restricted with endonucleases BamHI and SalI and ligated with

125

BamHI/SalI-digested vector pET30a, followed by transformation into E. coli DH5a. The

126

recombinant plasmids were isolated from transformants. After verifying by DNA

127

sequencing, the recombinant plasmids were transformed into E. coli BL21(DE3) and the

128

gene expression was induced by 0.5 mM of isopropyl-1-thio-β-D-galactopyranoside

129

(IPTG) at 37°C and 200 rpm for 3 h. The site-directed mutants were constructed using

130

Q5 site-directed mutagenesis kit according to the manufacturer’s protocol.

131

2.4 Purification of recombinant protein

132

After induction, the cells were harvested by centrifugation at 4°C with 8000  g for

133

5 min. The cell pellets were resuspended in a buffer containing 20 mM Tris-HCl, 500

134

mM NaCl (pH 8.0) and lysed by ultrasound (10 min, 2 s off, 2 s on). Supernatants were

135

collected after cell-lysates centrifugated at 4°C with 8000  g for 15 min. The

136

recombinant proteins in the supernatants were purified using Ni2+ His-tag column with

137

nonlinear imidazole gradient from 20 to 300 mM (in resuspension buffer). The

138

imidazole was removed by dialyzing 24 h in 20 mM Tris-HCl (pH 8.0). The purity of

7

139

recombinant proteins was examined by electrophoresis in 10% (w/v) sodium dodecyl

140

sulfate-polyacrylamide gel (SDS-PAGE). The protein concentration was estimated

141

using Quick Start™ Bradford Kit (Bio-Rad, USA) with Bovine serum albumin as

142

standard.

143

2.5 Enzyme assays

144

For xylanase activity assay, beechwood xylan was used as the substrate. The

145

reaction mixture consisted of 100 μL properly diluted enzyme and 100 μL of 10 mg/mL

146

BWX in 0.05 M citric-Na2HPO4 (pH 6.0). After 10 min incubation at 50°C, the

147

enzymatic activity was terminated by adding 150 μL 3,5-dinitrosalicylic acid reagent

148

(DNS) followed by boiling for 5 min. The concentrations of reducing sugar were

149

measured at 540 nm after the mixture was cooled down to room temperature. Activities

150

were calculated with xylose as the standard. One unit of enzyme activity was defined as

151

the amount of enzyme catalyzing the release of 1 μmol of reducing sugars in 1 min.

152

For feruloyl esterase activity assay, methyl ferulate was used as substrate. The

153

reaction mixture consisted of 10 μL properly diluted enzyme and 190 μL of 1 mM MFA

154

in 0.05 M citric-Na2HPO4 (pH 7.0). After 10 min incubation at 50°C, the enzymatic

155

activity was terminated by adding 100 μL acetonitrile. Concentration of ferulic acid in

156

the mixture was determined by Essentia LC-15C high performance liquid

157

chromatography (HPLC) (Shimadzu, Kyoto, Japan) using a C18 column (Advanced

158

Chromatography Technologies Ltd, 1 Berry Street Aberdeen, Scotland) and a SPD-15C

159

Essentia UV/VIS detector (Shimadzu, Kyoto, Japan) at 40°C, UV wavelength was set at

8

160

322 nm. Separation was performed within 6 min using a mobile phase consisting of 0.1%

161

formic acid and 60 % acetonitrile at a rate of 1 mL per min. One unit of enzyme activity

162

was defined as the amount of enzyme that released 1 μmol of ferulic acid per min.

163

2.6 Effects of temperature and pH on enzyme activity and stability

164

The effect of pH on enzyme activity was determined by measuring xylanase and

165

feruloyl esterase activity in 0.05 M citric-Na2HPO4 buffer (pH 4.0-8.0) and

166

glycine-NaOH (pH 9.0-10.0) at 50°C for 10 min. Enzyme stability against pH was

167

determined by pre-incubating 0.7 mg/mL of enzyme in 0.05 M different buffers at 4°C

168

for 1 h and then measuring the residual activity under the standard assay condition.

169

Glycine–HCl (pH 2.2–3.0), sodium acetate (pH 4.0–5.0), sodium phosphate (pH

170

6.0–8.0), glycine–NaOH (pH 9.0–10.0) and NaH2PO4–NaOH (pH 11.0–12.0) buffers

171

were used.

172

The effect of temperature on enzyme activity was investigated by measuring the

173

xylanase in 0.05 M citric-Na2HPO4 buffer (pH 6.0) and feruloyl esterase activity in 0.05

174

M citric-Na2HPO4 buffer (pH 7.0) at 20–60°C for 10 min. Thermostability of enzyme

175

was determined by measuring the residual activity under the standard assay condition

176

after incubating 0.7 mg/mL of enzyme in 20 mM Tris-HCl (pH 8.0) at different

177

temperature (4, 30, 40, 50, 60 and 70°C) for 1 h.

178

2.7 Kinetic parameters and substrate specificity

179 180

Xylanase enzyme kinetic assays were carried out in 0.05 M citric-Na2HPO4 buffer (pH 6.0) at 50°C for 10 min, using BWX and WAX at concentrations of 0-20.0 mg/mL.

9

181

Feruloyl esterase enzyme kinetic assays were carried out in 0.05 M citric-Na2HPO4

182

buffer (pH 7.0) at 50°C for 10 min, using MFA at concentrations of 0-2.0 mM. Kinetic

183

parameters were calculated by non-linear regression fit directly to the Michaelis-Menten

184

equation using Graphpad prism 7 software.

185

The substrate preferences of the purified recombinant enzyme were investigated

186

with different xylans (including BWX, WAX and BX) and hydroxycinnamic acid esters

187

(including MFA, EFA, MpCA, MCA and MSA) according to the method described

188

previously. The acetyl xylan esterase activity of the purified recombinant enzyme was

189

investigated with 1 mM p-nitrophenyl acetate (pNPAc) in 0.05 M citric-Na2HPO4 buffer

190

(pH 7.0) at 50°C for 10 min. The produced pNP was monitored spectrophotometrically

191

at 410 nm. One unit of enzyme activity was defined as the amount of enzyme required

192

to release 1 μmol of p-nitrophenol per min.

193

2.8 Hydrolysis of esterified biopolymers

194

To test the performance of rXyn10A/Fae1A versus the combined mixture of

195

equimolar concentrations of individual truncated mutants, hydrolysis experiments of

196

in-soluble wheat arabinoxylan were performed. Samples of 20 mg in-soluble wheat

197

arabinoxylan were incubated with 0.4 μM enzyme individually or co-incubation under 1

198

mL 0.05 M citric-Na2HPO4 buffer, pH 6.0 at 40°C and 200 rpm for 15 min. The

199

reaction was terminated by boiling for 10 min and centrifugated (8000  g, 10 min). The

200

supernatant was collected and then filtered through a 0.22 μm filter. The reducing sugar

201

(xylose/XOSs) was estimated using the DNS method with xylose as a standard. The

10

202

concentrations of hydroxycinnamic acid in the filtrates were determined by HPLC as

203

described before. All hydrolysis experiments were conducted in triplicate. Statistical

204

significance between the groups was determined through one-way analysis of variance

205

(ANOVA) followed by Duncan’s multiple range test (p＜0.05) using SPSS, version 23.

206

To evaluate of the applied potentials, 20 mg dry mass of pretreated agricultural

207

residue was hydrolyzed by 0.4 μM rXyn10A/Fae1A under 1 mL 0.05 M citric-Na2HPO4

208

buffer, pH 6.0 at 40°C and 200 rpm for 24 h. The reducing sugar (xylose/XOSs) and

209

hydroxycinnamic acids in the hydrolysate were quantified by DNS and HPLC.

210

The total amount of ester-linked hydroxycinnamic acids in the substrates were

211

determined by alkaline hydrolysis (2.0 M sodium hydroxide for 4 h at 50°C and 200

212

rpm). After acidification with HCl, the samples were filtered and analyzed with HPLC

213

as described above.

214

For saving the process of enzyme purification, extracellular secretory expression of

215

bifunctional xylanase/feruloyl esterase in E. coli was attempted. Protein 48211 without

216

signal peptide coding sequences were amplified using the specific primers. The PCR

217

products were restricted with endonucleases SalI and XhoI and were ligated with

218

SalI/XhoI-digested vector pET22b+.The transformation was performed as described in

219

section 2.3. The recombinant E. coli strain was inoculated in the autoclaved 10 mL LB

220

with 200 mg de-starched wheat bran medium and cultured at 37 °C. After induced by

221

0.5 mM IPTG, the samples were taken out and analyzed at time intervals to detect the

222

production of ferulic acid and reducing sugars (xylose/XOSs).

11

223 224

2.9 Bioinformatics analysis The signal peptide was predicted at SignalP 4.1 server

225

(http://www.cbs.dtu.dk/services/SignalP/). The domain architecture was annotated using

226

dbCAN (http://bcb.unl.edu/dbCAN2/). Sequence similarity was assessed using BLAST

227

at the NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

228

3. Results and discussion

229

3.1 Expression and purification of recombinant bifunctional xylanase/feruloyl

230

esterase

231

SignalP analysis of the amino acid sequence of 48211 confirmed the presence of an

232

N-terminal signal peptide of 28 amino acid residues, suggesting that 48211 was a

233

secreted protein. In dbCAN database, 48211 was annotated as a putative multi-modular

234

bifunctional xylanase/feruloyl esterase composed of four domains (from N-terminal to

235

C-terminal): xylanase domain from GH10, carbohydrate binding module (CBM) from

236

family 13, carbohydrate esterase domain from CE1 and CBM from family 2 (Fig. 1A).

237

Blast analysis of the amino acid sequence of 48211 showed that it had the highest (78%)

238

homology with a hypothetical protein (Genebank: WP_092476704.1) from Clostridium

239

polysaccharolyticum. Furthermore, a homology search of the Protein Data Bank (PDB)

240

found the highest homology (32.4%) of GH10 domain with xylanase domain 2W5F_A

241

from Clostridium thermocellum XynY (Najmudin et al., 2010). And CE1 domain

242

showed the highest homology of 63.2% with feruloyl esterase (5CXU_A) (Gruninger et

12

243

al., 2016) from ruminant anaerobic fungi Anaeromyces mucronatus and 42.5% with

244

CE1 domain (1JJF_A) from XynZ (Prates et al., 2001). 5CXU_A was more closely

245

related to anaerobic bacterial FAEs than fungal enzymes (Qi et al., 2011). Multiple

246

amino acid sequence alignment of the 48211 catalytic domains with other xylanases and

247

feruloyl esterases indicated that it had conservative Glu196 and Glu306 in xylanase and

248

Ser765-Asp833-His864 in feruloyl esterase for catalysis. Compared with previously

249

reported bifunctional xylanase/feruloyl esterase enzymes (cellulosomal XynY and

250

XynZ, Xyn10D-Fae1A and fused FLX), 48211 has a different domain structure and low

251

sequence similarity, which suggested it is a novel multi-modular bifunctional

252

xylanase/feruloyl esterase.

253

To investigate the function of 48211, its gene without signal peptide coding

254

sequences was successfully cloned into pET30a vector and expressed in E. coli

255

BL21(DE3). The predicted molecular weight of recombinant protein was 114.3 kDa and

256

it displayed consistent molecular weight according to SDS-PAGE (Fig. 1B).

257

3.2 Biochemical characteristics of rXyn10A/Fae1A

258

As shown in Fig. 2, recombinant protein showed activities towards all the tested

259

substrates, including xylans [beechwood xylan (BWX), soluble wheat arabinoxylan

260

(WAX) and bagasse xylan (BX)], hydroxycinnamic acid esters [methyl ferulate (MFA),

261

ethyl ferulate (EFA), methyl p-coumarate (MpCA), methyl caffeate (MCA) and methyl

262

sinapinate (MSA)] and p-nitrophenyl acetate (pNPAc). The specific activity against

263

pNPAc was the lowest (0.9 U/mg) with no acetic acid release from lignocellulose by

13

264

recombinant protein. These results indicated that, in agreement with annotation, the

265

recombinant protein was a bifunctional enzyme with both xylanase and feruloyl esterase

266

activities and termed as rXyn10A/Fae1A. rXyn10A/Fae1A displayed the highest

267

activity on WAX (49.1 U/mg). As for hydroxycinnamic acid esters, rXyn10A/Fae1A

268

showed the highest activity with MSA (12.1 U/mg). And rXyn10A/Fae1A followed the

269

activity pattern of MSA>MFA>MpCA>MCA, which was in agreement with its

270

homology 5CXU_A (Qi et al., 2011). The xylanase and feruloyl esterase activities of

271

XynZ were 21 U/mg and 0.4 U/mg, respectively (Mandelli et al., 2014) and the feruloyl

272

esterase activity of Xyn10D-Fae1A was even only 0.001 U/mg (Dodd et al., 2009).

273

Therefore, the specific activity of rXyn10A/Fae1A was higher than previously reported

274

natural bifunctional xylanase/feruloyl esterase enzymes.

275

Enzymatic properties of rXyn10A/Fae1A were investigated using BWX and MFA

276

as substrates. As shown in Fig. 3A and 3C, the optimal temperatures of both activities

277

were 50°C. rXyn10A/Fae1A was stable at 4-40°C and drastically reduced its activities

278

at temperature higher than 50°C. The optimal reaction pH for rXyn10A/Fae1A xylanase

279

activity was 6.0, and over 97% of the maximal activities were kept between pH 5.0 and

280

7.0, indicating it is a weak acidic-neutral xylanase. When incubated at the pH range of

281

4.0-10.0 for 1 h at 4°C, more than 95% of the xylanase activity was retained. The

282

optimal reaction pH for rXyn10A/Fae1A feruloyl esterase activity was 7.0 and it

283

retained more than 85% of the activity at the pH range of 2.2-12.0 (Fig. 3B and 3D).

284

rXyn10A/Fae1A has nearly similar optimal conditions for both enzymes activities

14

285 286

which would be beneficial to its application. The kinetic parameters Km and Kcat/Km of rXyn10A/Fae1A against xylan and

287

hydroxycinnamic acid ester were shown in Table 1. And in order to decipher the

288

contribution of the different modules to the function of rXyn10A/Fae1A, several

289

truncated mutants were constructed (Fig. 1B). We found that truncated mutants have the

290

same optimal conditions with rXyn10A/Fae1A. However, the catalytic efficiencies of

291

truncated mutants with removal of another catalytic domain were 1.4- to 5.4-fold of

292

rXyn10A/Fae1A against those model soluble substrates (Table 1). In a previous study

293

by Su et al. (2012), it was reported that deleting Man5A domain of bifunctional

294

cellulase/mannanase CbCel9B/Man5A led to two-fold increase in the Kcat/Km of

295

CbCel9B. These observations suggested that the activity of each catalytic domain was

296

modulated by the other catalytic module in a single polypeptide.

297

3.3 Inter-domain synergism of rXyn10A/Fae1A on insoluble wheat arabinoxylan

298

Model substrate insoluble wheat arabinoxylan (I-WAX), which maintains the

299

ferulic acid crosslinks in the native arabinoxylan after purification, was also used to test

300

the activity of rXyn10A/Fae1A. rXyn10A/Fae1A (46 μg/20 mg substrate) showed the

301

yield of 161 mg/g substrate of reducing sugars (xylose/XOSs) and 2.78 mg/g substrate

302

of FA from I-WAX at 40°C for 15 min (Fig. 4A and 4B). Until now, only one

303

bifunctional xylanase/feruloyl esterase (XynZ) has been used in hydrolysis of I-WAX

304

(Mandelli et al., 2014). Although the production of phenolic compounds was similar,

305

rXyn10A/Fae1A released nearly five-fold reducing sugars of XynZ (33.9 mg/g

15

306

substrate). Recently reported feruloyl esterases (FaeC and FaeD) exhibited the highest

307

yield of FA (3.0 mg/g substrate and 2.3 mg/g substrate, respectively) from I-WAX

308

(Dilokpimol et al., 2017; Makela et al., 2018). However, we should consider that longer

309

hydrolysis time (24 h) were needed for FaeC and FaeD and the I-WAX was pretreated

310

by a high dosage of commercial xylanase (400 μg/20 mg substrate) for 72 h. Therefore,

311

rXyn10A/Fae1A was superior to other reported feruloyl esterases and xylanases for FA

312

and XOSs production.

313

We hypothesized that the inter-domain synergism of rXyn10A/Fae1A is a

314

necessary factor for efficiently releasing FA and XOSs from I-WAX. Experiments were

315

carried out to compare the hydrolysis of rXyn10A/Fae1A with its truncated mutants

316

individually or in combination on I-WAX. As shown in Fig. 4A, equimolar mix of

317

GH10-CBM13 and CE1-CBM2 or GH10 and CE1 could achieve same production of

318

reducing sugars on I-WAX to that of rXyn10A/Fae1A. And the synergy factors of

319

GH10-CBM13 plus CE1-CBM2 and GH10 plus CE1 were 1.1 and 1.2, respectively. To

320

further confirm the observed role of the feruloyl esterase in the release of reducing

321

sugars from I-WAX, we mutated the catalytic residue serine (Ser-765) of esterase

322

domain to alanine to inactivate the enzyme. The mutant (S765A) showed comparable

323

reducing sugars with rXyn10A/Fae1A. These results suggested that the facilitation of

324

the feruloyl esterase to the xylanase was very faint. Compared with the production of

325

FA by rXyn10A/Fae1A, that by CE1-CBM2 and CE1 were dramatically decreased by

326

92% and 93%. With the addition of GH10-CBM13 or GH10, the release of FA

16

327

increased by 856% or 967%, respectively. These results indicated that xylanase acted

328

synergistically with feruloyl esterase in the release of FA and the help of xylanase was a

329

necessary factor for high production of FA. These conclusions were further

330

demonstrated by the result of mutant E196A. The mutant E196A, which was the

331

xylanase-inactive form of rXyn10A/Fae1A, released only 16.9% FA production of

332

rXyn10A/Fae1A. However, it is very interesting that the mix enzymes GH10-CBM13

333

plus CE1-CBM2 and GH10 plus CE1 could only achieved 70.3% and 70.9% FA

334

production of rXyn10A/Fae1A. We speculated that the physical proximity of xylanase

335

and feruloyl esterase in a polypeptide may also aid the release of FA from complex

336

substrate. To the best of our knowledge, this study firstly reported that the inter-domain

337

synergism of bifunctional xylanase/feruloyl esterase is a necessary factor for efficiently

338

releasing FA but not XOSs from I-WAX.

339

3.4 FA and XOSs production from agricultural residues using rXyn10A/Fae1A

340

From an industrial point of view, the ability of rXyn10A/Fae1A to release FA and

341

XOSs from agricultural waste materials was tested using de-starched wheat bran

342

(DSWB), ultrafine-grinding corn stover (UGCS) and steam-exploded corncob (SECC).

343

When 46 μg rXyn10A/Fae1A was used, the yields of FA were 36.4, 23.0 and 146 μg on

344

the basis of 20 mg DSWB, UGCS and SECC, respectively, after 24 h incubation (Table

345

2). DSWB used in this study contained 2.5 mg FA per g DSWB. After 24 h enzymatic

346

hydrolysis, a conversion rate of 71.9% of alkaline extractive was obtained, with a yield

347

of 1.82 mg/g DSWB. Although EstF27 M6 released comparable amounts of FA (1.80

17

348

mg/g DSWB), the enzyme dosage was nearly five-fold of rXyn10A/Fae1A (Cao et al.,

349

2015). The yield of FA by rXyn10A/Fae1A was higher than those of 1.52 mg/g DSWB

350

by LhFae (Wang et al., 2016a), 1.36 mg/g DSWB by ScFaeD1 and 1.31 mg/g DSWB

351

by ScFaeD2 (Nieter et al., 2016). Moreover, the high production of FA mentioned

352

above all depended on the addition of commercial xylanase, which increased the cost of

353

bioconversion. While the fused bifunctional xylanase/feruloyl esterase FLX could

354

completely release FA from DSWB, the enzyme loading reached 2000 μg which was 43

355

times higher than rXyn10A/Fae1A (Levasseur et al., 2005). Except DSWB,

356

rXyn10A/Fae1A was also able to release 1.15 mg FA per g UGCS, with a conversion

357

rate of 41.1%. The yield was much higher than those of 0.5 mg/g sugarcane bagasse by

358

AcFAE (Damasio et al., 2013) and 0.2 mg/g steam-exploded corn stover by AfFaeA

359

(Zhang et al., 2013). Corncob is an important byproduct of corn with a global yield

360

exceeding 800,000,000 tons, and most of corncobs are burned and cause environmental

361

pollution (Xian et al., 2019). After pretreated by steam explosion, rXyn10A/Fae1A

362

could release 7.31 mg FA per g SECC with a conversion rate of 67.7%. To the best of

363

our knowledge, the FA yield from SECC by rXyn10A/Fae1A was the highest among

364

previously reported enzymes using biomass as substrates. And amount of 1.24, 1.22 and

365

0.823 mg reducing sugars (xylose/XOSs) were obtained from 20 mg DSWB, UGCS and

366

SECC simultaneously. High efficiency of rXyn10A/Fae1A in conversing different kinds

367

of agricultural residues to FA and XOSs demonstrated the great biotechnological

368

potential.

18

369

3.5 Hydrolysis of de-starched wheat bran by the bifunctional xylanase/feruloyl

370

esterase extracellular secretory recombinant E. coli

371

Extracellular secretion of recombinant enzyme in E. coli could simplify the process

372

of protein purification and the recombinant strain could be directly used in the

373

conversion of agricultural residues to value-added products (Xu et al., 2019). Based

374

upon its high hydrolysis efficiency, the extracellular secretion of the bifunctional

375

xylanase/feruloyl esterase in E. coli was attempted. When the recombinant E. coli was

376

inoculated in medium containing 200 mg DSWB, products of FA and reducing sugars

377

(xylose/XOSs) rapidly increased along with the fermentation. After 36 h cultivation,

378

474 μg FA and 18.2 mg reducing sugars (xylose/XOSs) were detected (Fig. 5). These

379

results confirmed that the active form of the bifunctional xylanase/feruloyl esterase was

380

indeed secreted into extracellular environment. This study was believed to be the first

381

report of extracellular secretion of bifunctional xylanase/feruloyl esterase in E. coli and

382

production of FA and XOSs from biomass by recombinant E. coli, which providing a

383

more convenient method for FA and XOSs production.

384

4. Conclusions

385

This study reported a novel bifunctional xylanase/feruloyl esterase rXyn10A/Fae1A.

386

High amounts of FA from I-WAX, DSWB, UGCS and SECC (2.78, 1.82, 1.15 and 7.31

387

mg/g substrate, respectively) were obtained by rXyn10A/Fae1A depending on the

388

synergism of its inter-domain. Furthermore, extracellular secretion expression system of

19

389

the bifunctional xylanase/feruloyl esterase in E. coli was constructed. 474 μg FA and

390

18.2 mg xylose/XOSs were also detected in the medium during the cultivation of

391

recombinant E. coli in LB medium containing DSWB. Hence, rXyn10A/Fae1A and the

392

recombinant E. coli were excellent candidates in FA and XOSs production.

393

Author Contribution Statement

394

Ruonan Wang and Hongli Yuan conceived and designed the experiments. Ruonan

395

Wang performed the majority of the laboratory work, analyzed the results and wrote the

396

manuscript. Jinshui Yang contributed to the interpretation of the results and revision of

397

the manuscript. Jin Myong Jang assisted in data analysis. Jiawen Liu, Yu Zhang and

398

Liang Liu carried out the material pretreatment. Hongli Yuan supervised the overall

399

work, discussed the results, and revised the manuscript. All authors read and approved

400

the final manuscript.

401

Appendix A. Supplementary data

402

E-supplementary data for this work can be found in e-version of this paper online.

403

Conflict of interest

404

None.

405

Acknowledgements

20

406

This work was supported by the project for extramural scientists of state key

407

laboratory of agrobiotechnology: 2018SKLAB6-28.

408

References

409

1. Blum, D.L., Kataeva, I.A., Li, X.L., Ljungdahl, L.G., 2000. Feruloyl esterase activity

410

of the Clostridium thermocellum cellulosome can be attributed to previously

411

unknown domains of XynY and XynZ. J. Bacteriol. 182 (5), 1346-1351.

412

2. Cao, L.C., Chen, R., Xie, W., Liu, Y.H., 2015. Enhancing the thermostability of

413

feruloyl esterase EstF27 by directed evolution and the underlying structural

414

basis. J. Agric. Food Chem. 63 (37), 8225-8233.

415

3. Cheng, F.S., Sheng, J.P., Cai, T., Jin, J., Liu, W.Z., Lin, Y.M., Du, Y.X., Zhang,

416

M.Q., Shen, L., 2012. A protease-insensitive feruloyl esterase from China

417

Holstein cow rumen metagenomic library: expression, characterization, and

418

utilization in ferulic acid release from wheat straw. J. Agric. Food Chem. 60 (10),

419

2546-2553.

420

4. Damasio, A.R.L., Braga, C.M.P., Brenelli, L.B., Citadini, A.P., Mandelli, F., Cota, J.,

421

de Almeida, R.F., Salvador, V.H., Paixao, D.A.A., Segato, F., Mercadante, A.Z.,

422

de Oliveira Neto, M., do Santos, W.D., Squina, F.M., 2013.

423

Biomass-to-bio-products application of feruloyl esterase from Aspergillus

21

424 425

clavatus. Appl. Microbiol. Biotechnol. 97 (15), 6759-6767. 5. Dilokpimol, A., Makela, M.R., Mansouri, S., Belova, O., Waterstraat, M., Bunzel, M.,

426

de Vries, R.P., Hilden, K.S., 2017. Expanding the feruloyl esterase gene family

427

of Aspergillus niger by characterization of a feruloyl esterase, FaeC. N.

428

Biotechnol. 37 (Pt B), 200-209.

429

6. Dodd, D., Kocherginskaya, S.A., Spies, M.A., Beery, K.E., Abbas, C.A., Mackie,

430

R.I., Cann, I.K.O., 2009. Biochemical analysis of a beta-D-xylosidase and a

431

bifunctional xylanase-ferulic acid esterase from a xylanolytic gene cluster in

432

Prevotella ruminicola 23. J. Bacteriol. 191 (10), 3328-3338.

433

7. Gruninger, R.J., Cote, C., McAllister, T.A., Abbott, D.W., 2016. Contributions of a

434

unique beta-clamp to substrate recognition illuminates the molecular basis of

435

exolysis in ferulic acid esterases. Biochem. J. 473 (7), 839-849.

436

8. Levasseur, A., Navarro, D., Punt, P.J., Belaich, J.P., Asther, M., Record, E., 2005.

437

Construction of engineered bifunctional enzymes and their overproduction in

438

Aspergillus niger for improved enzymatic tools to degrade agricultural

439

by-products. Appl. Environ. Microbiol. 71 (12), 8132-8140.

440

9. Li, J.B., Cai, S.C., Luo, Y.M., Dong, X.Z., 2011. Three feruloyl esterases in

441

cellulosilyticum ruminicola H1 act synergistically to hydrolyze esterified

442

polysaccharides. Appl. Environ. Microbiol. 77 (17), 6141-6147.

443

10. Li, J.B., Zhang, H.Y., Lu, M.S., Han, L.J., 2019. Comparison and intrinsic

444

correlation analysis based on composition, microstructure and enzymatic

22

445

hydrolysis of corn stover after different types of pretreatments. Bioresour.

446

Technol. 293.

447

11. Liu, L., Yang, J.S., Yang, Y., Luo, L.J., Wang, R.N., Zhang, Y., Yuan, H.L., 2019.

448

Consolidated bioprocessing performance of bacterial consortium EMSD5 on

449

hemicellulose for isopropanol production. Bioresour. Technol. 292, 121965.

450

12. Long, L., Zhao, H., Ding, D., Xu, M., Ding, S., 2018. Heterologous expression of

451

two Aspergillus niger feruloyl esterases in Trichoderma reesei for the

452

production of ferulic acid from wheat bran. Bioprocess Biosyst. Eng. 41 (5),

453

593-601.

454

13. Lv, Z.W., Yang, J.S., Yuan, H.L., 2008. Production, purification and

455

characterization of an alkaliphilic endo-β-1,4-xylanase from a microbial

456

community EMSD5. Enzyme Microb. Technol. 43 (4-5), 343-348.

457

14. Makela, M.R., Dilokpimol, A., Koskela, S.M., Kuuskeri, J., de Vries, R.P., Hilden,

458

K., 2018. Characterization of a feruloyl esterase from Aspergillus terreus

459

facilitates the division of fungal enzymes from Carbohydrate Esterase family 1

460

of the carbohydrate-active enzymes (CAZy) database. Microbial Biotechnol. 11

461

(5), 869-880.

462

15. Mandelli, F., Brenelli, L.B., Almeida, R.F., Goldbeck, R., Wolf, L.D., Hoffmam,

463

Z.B., Ruller, R., Rocha, G.J.M., Mercadante, A.Z., Squina, F.M., 2014.

464

Simultaneous production of xylooligosaccharides and antioxidant compounds

465

from sugarcane bagasse via enzymatic hydrolysis. Ind. Crops Prod. 52, 770-775.

23

466

16. Najmudin, S., Pinheiro, B.A., Prates, J.A.M., Gilbert, H.J., Romao, M.J., Fontes,

467

C.M.G.A., 2010. Putting an N-terminal end to the Clostridium thermocellum

468

xylanase Xyn10B story: crystal structure of the CBM22-1-GH10 modules

469

complexed with xylohexaose. J. Struct. Biol. 172 (3), 353-362.

470

17. Nieter, A., Kelle, S., Linke, D., Berger, R.G., 2016. Feruloyl esterases from

471

Schizophyllum commune to treat food industry side-streams. Bioresour. Technol.

472

220, 38-46.

473

18. Oliveira, D.M., Finger-Teixeira, A., Mota, T.R., Salvador, V.H., Moreira-Vilar, F.C.,

474

Molinari, H.B., Mitchell, R.A., Marchiosi, R., Ferrarese-Filho, O., dos Santos,

475

W.D., 2015. Ferulic acid: a key component in grass lignocellulose recalcitrance

476

to hydrolysis. Plant Biotechnol. J. 13 (9), 1224-1232.

477

19. Oliveira, D.M., Mota, T.R., Oliva, B., Segato, F., Marchiosi, R., Ferrarese, O.,

478

Faulds, C.B., dos Santos, W.D., 2019. Feruloyl esterases: Biocatalysts to

479

overcome biomass recalcitrance and for the production of bioactive compounds.

480

Bioresour. Technol. 278, 408-423.

481

20. Prates, J.A.M., Tarbouriech, N., Charnock, S.J., Fontes, C.M.G.A., Ferreira, L.M.A.,

482

Davies, G.J., 2001. The structure of the feruloyl esterase module of xylanase

483

10B from Clostridium thermocellum provides insights into substrate recognition.

484

Structure 9 (12), 1183-1190.

485

21. Qi, M., Wang, P., Selinger, L.B., Yanke, L.J., Forster, R.J., McAllister, T.A., 2011.

486

Isolation and characterization of a ferulic acid esterase (Fae1A) from the rumen

24

487 488

fungus Anaeromyces mucronatus. J. Appl. Microbiol. 110 (5), 1341-1350. 22. Sang, S.L., Li, G., Hu, X.P., Liu, Y.H., 2011. Molecular cloning, overexpression

489

and characterization of a novel feruloyl esterase from a soil metagenomic library.

490

J. Mol. Microbiol. Biotechnol. 20 (4), 196-203.

491

23. Su, X.Y, Mackie, R.I., Cann, I.K.O., 2012. Biochemical and mutational analyses of

492

a multidomain cellulase/mannanase from Caldicellulosiruptor bescii.

493

Appl. Environ. Microbiol. 78 (7), 2230-2240.

494

24. Topakas, E., Stamatis, H., Biely, P., Christakopoulos, P., 2004. Purification and

495

characterization of a type B feruloyl esterase (StFAE-A) from the thermophilic

496

fungus Sporotrichum thermophile. Appl. Microbiol. Biotechnol. 63 (6),

497

686-690.

498

25. Uraji, M., Arima, J., Inoue, Y., Harazono, K., Hatanaka, T., 2014. Application of

499

two newly identified and characterized feruloyl esterases from Streptomyces sp.

500

in the enzymatic production of ferulic acid from agricultural biomass. PLoS One

501

9 (8), e104584.

502

26. Wang, L., Li, Z., Zhu, M.J., Meng, L., Wang, H.X., Ng, T.B., 2016a. An acidic

503

feruloyl esterase from the mushroom Lactarius hatsudake: A potential animal

504

feed supplement. Int. J. Biol. Macromol. 93 (Pt A), 290-295.

505

27. Wang, Y.K., Zhang, X., Chen, G.L., Yu, J., Yang, L.Q., Gao, Y.Q., 2016b.

506

Antioxidant property and their free, soluble conjugate and insoluble-bound

507

phenolic contents in selected beans. J. Funct. Foods 24, 359-372.

25

508 509

28. Wong, D.W.S., 2006. Feruloyl esterase - A key enzyme in biomass degradation. Appl. Biochem. Biotechnol. 133 (2), 87-112.

510

29. Wu, H.L., Li, H.L., Xue, Y., Luo, G., Gan, L.H., Liu, J., Mao, L.H., Long, M.N.,

511

2017. High efficiency co-production of ferulic acid and xylooligosaccharides

512

from wheat bran by recombinant xylanase and feruloyl esterase. Biochem. Eng.

513

J. 120, 41-48.

514

30. Xian, L., Li, Z., Tang, A.X., Qin, Y.M., Li, Q.Y., Liu, H.B., Liu, Y.Y., 2019. A

515

novel neutral and thermophilic endoxylanase from Streptomyces ipomoeae

516

efficiently produced xylobiose from agricultural and forestry residues. Bioresour.

517

Technol. 285, 121293.

518

31. Xu, Z.S, Wang, T., Zhang, S.S, 2019. Extracellular secretion of feruloyl esterase

519

derived from Lactobacillus crispatus in Escherichia coli and its application for

520

ferulic acid production. Bioresour. Technol. 288, 121526.

521

32. Zhang, S.B., Zhai, H.C., Wang, L., Yu, G.H., 2013. Expression, purification and

522

characterization of a feruloyl esterase A from Aspergillus flavus. Protein Expr.

523

Purif. 92 (1), 36-40.

524

33. Zhu, N., Yang, J.S., Ji, L., Liu, J.W., Yang, Y., Yuan, H.L., 2016. Metagenomic and

525

metaproteomic analyses of a corn stover-adapted microbial consortium EMSD5

526

reveal its taxonomic and enzymatic basis for degrading lignocellulose.

527

Biotechnol. Biofuels 9, 243-265.

26

528

Table 1. Kinetic parameters of rXyn10A/Fae1A and its mutants of with various substrates BWX Protein

Kcat

WAX Km

Kcat/Km

MFA Kcat

Km

Kcat/Km

(s-1)

(mM)

(mM-1s-1)

15.60

27.98 ± 1.33

0.39 ± 0.05

71.74

10.84 ± 2.43

69.44

-

-

-

696.9 ± 97.1

14.11 ± 3.46

49.39

-

-

-

-

-

-

-

27.63 ± 1.09

0.22 ± 0.02

125.6

-

-

-

-

39.23 ± 1.92

0.39 ± 0.04

100.6

(s-1)

(mg/mL)

(mL

rXyn10A/Fae1A

96.53 ± 7.03

5.14 ± 0.94

GH10-CBM13

413.3 ± 21.8

GH10

mg-1s-1)

Kcat

Km

Kcat/Km

(s-1)

(mg/mL)

(mL

18.78

188.7 ± 16.4

12.10 ± 1.94

4.07 ± 0.59

101.5

752.8 ± 88.1

330.9 ± 19.5

3.84 ± 0.63

86.17

CE1-CBM2

-

-

CE1

-

-

Data reflect the mean ± SD (n =3).

27

mg-1s-1)

529

Table 2. Yield of ferulic acid, p-coumaric acid and reducing sugars from agricultural residues by rXyn10A/Fae1A rXyn10A/Fae1A released Substrate

Alkaline extracted ferulic acid (μg)

Ferulic acid (μg)

p-coumaric acid (μg)

Reducing sugars (mg)

de-starched wheat bran

50.6 ± 7.6

36.4 ± 1.88

not detected

1.24 ± 0.028

ultrafine-grinding corn stover

56.0 ± 2.0

23.0 ± 0.803

46.3 ± 1.80

1.22 ± 0.021

steam-exploded corncob

215.6 ± 6.0

146 ± 6.11

67.5 ± 5.17

0.823 ± 0.248

Release of hydroxycinnamic acids and reducing sugars from agricultural residues (20 mg) by rXyn10A/Fae1A (0.4 μM) was determined after incubation (40°C, pH 6.0, 24 h, 200 rpm). Data reflect the mean ± SD (n =3).

28

Figure captions: Figure 1. Sequence analysis and purification of rXyn10A/Fae1A and its mutants. (A) Domain organization of rXyn10A/Fae1A and other bifunctional xylanase/feruloyl esterase. The signal peptide is represented by the black rectangle. GH10/11: glycoside hydrolase family 10/11; CBM13/2/22/6: carbohydrate-binding module from different families; Doc: dockerin module; CE1: carbohydrate esterase family 1; Lipase_3: lipase family 3. (B) SDS-PAGE of rXyn10A/Fae1A and its site-directed or truncated mutants. Lanes: M, molecular mass markers; 1, rXyn10A/Fae1A; 2, E196A, xylanase-inactive mutant; 3, S765A, feruloyl esterase-inactive mutant; 4, GH10-CBM13; 5, CE1-CBM2; 6, GH10; 7, CE1. Figure 2. The substrate specificity of rXyn10A/Fae1A. Data reflect the mean ± SD (n =3). Figure 3. Effects of temperature and pH on the xylanase and feruloyl esterase activity of rXyn10A/Fae1A. Optimal temperature (A) and pH (B) for rXyn10A/Fae1A activity. Thermostability stability (C) and pH stability (D) of rXyn10A/Fae1A. Data reflect the mean ± SD (n =3). Figure 4. Hydrolysis of in-soluble wheat arabinoxylan (I-WAX) by rXyn10A/Fae1A and its truncated or site-directed mutants. (A) Production of reducing sugars. (B) Production of ferulic acid. E196A, the catalytic residue Glu196 for the xylanase domain of rXyn10A/Fae1A was mutated to Ala; S765A, the catalytic residue Ser765 for the feruloyl esterase domain of rXyn10A/Fae1A was mutated to Ala; Truncated mutants of rXyn10A/Fae1A: GH10-CBM13, CE1-CBM2, GH10 and CE1. Data reflect the mean ±

29

SD (n =3). Statistical significance is indicated by different letters on columns based on ANOVA (P < 0.05). Figure 5. The time course of produced ferulic acid or reducing sugars by xylanase/feruloyl esterase secretory recombinant E. coli cultured in 10 mL LB medium supplemented with 200 mg de-starched wheat bran. Data reflect the mean ± SD (n =3).

30

31

32

33

34

35

Highlights A novel bifunctional xylanase/feruloyl esterase (rXyn10A/Fae1A) was obtained The highest ferulic acid yield (7.31 mg/g substrate) was produced by rXyn10A/Fae1A Inter-domain synergism of rXyn10A/Fae1A is essential for the release of ferulic acid Recombinant E. coli was constructed and secreted the bifunctional enzyme out of cells Co-production of ferulic acid and XOSs during the cultivation of recombinant E. coli

36