Metabolic engineering of carbon overflow metabolism of Bacillus subtilis for improved N-acetyl-glucosamine production

Accepted Manuscript Metabolic engineering of carbon overflow metabolism of Bacillus subtilis for improved N-acetyl-glucosamine production Wenlong Ma, Yanfeng Liu, Hyun-dong Shin, Jianghua Li, Jian Chen, Guocheng Du, Long Liu PII: DOI: Reference:

S0960-8524(17)31824-2 https://doi.org/10.1016/j.biortech.2017.10.007 BITE 19048

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

12 September 2017 2 October 2017 4 October 2017

Please cite this article as: Ma, W., Liu, Y., Shin, H-d., Li, J., Chen, J., Du, G., Liu, L., Metabolic engineering of carbon overflow metabolism of Bacillus subtilis for improved N-acetyl-glucosamine production, Bioresource Technology (2017), doi: https://doi.org/10.1016/j.biortech.2017.10.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

1

Metabolic engineering of carbon overflow metabolism of Bacillus subtilis for improved

2

N-acetyl-glucosamine production

3

Wenlong Ma , Yanfeng Liu , Hyun-dong Shin , Jianghua Li , Jian Chen , Guocheng Du , Long Liu

1,2

1,2

3

1,2

2

2*

1,2

4 5

1. Key Laboratory of Carbohydrate Chemistry and Biotechnology, Ministry of Education, Jiangnan

6

University, Wuxi 214122, China

7

2. Key Laboratory of Industrial Biotechnology, Ministry of Education, Jiangnan University, Wuxi 214122,

8

China

9

3. School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta 30332,

10

USA

11

*Corresponding author:

12

Guocheng Du, Tel.: +86-510-85918309, Fax: +86-510-85918309, E-mail: [email protected].

13

1

14

Abstract

15

Bacillus subtilis is widely used as cell factories for the production of important industrial biochemicals.

16

Although many studies have demonstrated the effects of organic acidic byproducts, such as acetate,

17

on microbial fermentation, little is known about the effects of blocking the neutral byproduct overflow,

18

such as acetoin, on bioproduction. In this study, we focused on the influences of modulating overflow

19

metabolism on the production of N-acetyl-D-glucosamine (GlcNAc) in engineered B. subtilis. We found

20

that acetoin overflow competes with GlcNAc production, and blocking acetoin overflow increased

21

GlcNAc titer and yield by 1.38- and 1.39-fold, reaching 48.9 g/L and 0.32 g GlcNAc/g glucose,

22

respectively. Further blocking acetate overflow inhibited cell growth and GlcNAc production may be

23

induced by inhibiting glucose uptake. Taken together, our results show that blocking acetoin overflow

24

is a promising strategy for enhancing GlcNAc production. The strategies developed in this work may be

25

useful for engineering strains of B. subtilis for producing other important biochemicals.

26

Keywords

27

Bacillus subtilis, N-acetyl-D-glucosamine, Overflow metabolism, Acetoin, Acetate

28

2

29

30

1. Introduction

N-acetyl-D-glucosamine (GlcNAc), also known as 2-acetamido-2-deoxy-D-glucose, is a bioactive

31

amino monosaccharide. GlcNAc and their derivatives, alone or in combination, have mainly been used

32

in pharmaceutical and health food products due to its anti-inflammatory and anti-sarcoma effects in

33

osteoarthritis (du Souich, 2014; Hochberg et al., 2016; Liu et al., 2013a). Moreover, recent studies

34

have shown that GlcNAc can also be used as a new functional material with high potential in various

35

fields, for example as a potential candidate for the diagnosis of tumors via chemical exchange

36

saturation transfer magnetic resonance imaging (Longo et al., 2017; Rivlin & Navon, 2016) as well as a

37

marker to target nano-sized carriers for cancer diagnosis (Kumar et al., 2017). Given the widespread

38

application of GlcNAc and their derivatives, production of GlcNAc has been of considerable interest,

39

especially by environmentally friendly processes and microbial production methods (Liu et al., 2013a).

40

Among them, B. subtilis, which is generally recognized as safe, is a favorable industrial candidate as

41

cell factories owing to its genetically well-known and metabolically robust properties (Liu et al., 2013b;

42

Ozturk et al., 2016; Yang et al., 2017).

43

Previously, we constructed a GlcNAc-producing recombinant B. subtilis strain, BSGN5, by blocking

44

the importation of extracellular GlcNAc and catabolism of intracellular GlcNAc as well as by

45

eliminating acidic byproduct lactate formation (Liu et al., 2013c; Liu et al., 2014a). However, overflow

46

of the neutral byproduct, such as acetoin, has not been engineered to further enhance GlcNAc

47

production. Acetoin, accumulated in rich medium up to 20–30 g/L, is one of the major overflowed

48

byproducts produced by B. subtilis during fast growth (Fradrich et al., 2012; Fujita, 2009; Ramos et al.,

49

2000). Synthesis of these overflowed acetoin required 40–60 g/L glucose, which diverted valuable

50

carbon from biomass formation and markedly decreased GlcNAc yield. This overflow pathway leading

3

51

to low product yield is recognized as one of the major disadvantages that limits the productivity of B.

52

subtilis in biotechnological processes. The reactions relevant to acetoin synthesis are illustrated as

53

follows: →

54 -

55 56 57

-

(1) (2) (3)

In this work, to reduce the competitive acetoin formation and improve GlcNAc yield, acetolactate

58

synthase (AlsS) and acetolactate decarboxylase (AlsD) (Renna et al., 1993) were inactivated. During

59

fed-batch fermentation in a 3-L bioreactor, blocking acetoin overflow pushed carbon flux from

60

fructose-6-phosphate to GlcNAc synthesis pathway and successfully increased the GlcNAc titer and

61

yield by 38 % and 39 % to 48.9 g/L and 0.32 g GlcNAc/g glucose, respectively. The strategy used here is

62

useful, and can be used to guide approaches for minimizing overflow metabolism of B. subtilis in other

63

applications.

64

2. Materials and methods

65

2.1. Strains, plasmids, and culture conditions

66

The B. subtilis and E. coli strains and plasmids used in this study are listed in Table 1. All primers

67

(Table 2) were generated based on B. subtilis strain 168 (NC_000964.3). E. coli JM109 was used for

68

gene cloning studies. BSGN5 (B. subtilis 168 d

69

whose importation pathway of extracellular GlcNAc and catabolism pathway of intracellular GlcNAc

70

were blocked, was used as a starting strain to create GlcNAc overproduction variants (Liu et al., 2014a).

71

During the plasmid and strain construction, strains were cultured at 37°C in Luria-Bertani (LB) medium

, ΔnagPΔgamPΔgamAΔnagAΔnagBΔldh::lox72),

4

72 73

74

75

(tryptone, 10 g/L; yeast extract, 5 g/L; NaCl, 10 g/L) or LB agar. When necessary, the following b

w

df

(μg/mL): mp

, 00; k

m

, 5;

dz

, 5.

2.2. Blocking of acetoin synthesis

To block the formation of acetoin, alsR, alsS, and alsD, which are involved in acetoin synthesis

76

(Fradrich et al., 2012; Renna et al., 1993), were knocked out from the BSGN5 chromosome. The

77

marker-free knockout approach used here has been described previously (Yan et al., 2008). Briefly, the

78

front and back homology arms (800 bp) flanking the deletion target were respectively amplified

79

through PCR using the genomic DNA from the BSGN5 strain as a template. The Lox71-zeo-lox66

80

cassette was amplified from the plasmid p7Z6. Then, these three fragments were fused via PCR and

81

used for electro-transformation of BSGN5 competent cells (Xue et al., 1999). Afterward, zeo

82

transformants were selected and verified by colony PCR. The resistance marker cassette on the

83

genome of zeo transformants was evicted with the help of plasmid pTSC, which expresses Cre

84

recombinase. Finally, pTSC was removed by incubating the strains at 50°C for 24 h before the next

85

round of genome editing.

86

2.3. Identification of acetate

87

r

r

To characterize the organic acidic byproduct that accumulated in the fermented broth, the

88

culture supernatant was first analyzed by high-performance liquid chromatography (HPLC, Agilent

89

1260 series, Santa Clara, CA, USA), and then compared with 17 organic acidic metabolites synthesized

90

in the Embden–Meyerhof–Parnas (EMP), tricarboxylic acid (TCA), and pentose phosphate pathway.

91

The HPX-87H column (Bio-Rad Hercules, CA), refractive index detector, and UV detector (wavelength

92

at 210 nm) were used. HPLC analysis was carried out with 5 mM H2SO4 as the mobile phase at a flow

5

93

rate of 0.5 mL/min, and columns were maintained at 40°C (Liu et al., 2014b). Next, the organic acidic

94

byproduct was reconfirmed by mass spectrometer (MS) and nuclear magnetic resonance (NMR). Five

95

microliters of the culture supernatant were introduced into the electrospray ionization source of the

96

MS directly. Mass spectrometry was performed in negative ion mode with scanning over the m/z

97

range from 50–2,000 at 10 s/scan. NMR studies were carried out using a Bruker AVANCE III 400 MHz

98

NMR spectrometer.

99

2.4. Blocking of acetate synthesis

100

To block acetate synthesis, pta and ackA, which code for phosphotransacetylase and acetate

101

kinase, respectively, were knocked-out individually and combinatorically in BSGN10 using the

102

marker-free knockout approach as described above.

103

2.5. Shake flask fermentation

104

During shake flask fermentation, rich medium (RM1) consisting of (g/L): yeast extract, 12; trptone,

105

6; (NH4)2SO4, 6; K2HPO4·3H2O, 18.75; KH2PO4, 2.5; MgSO4, 3; FeSO4·7H2O, 0.06; CaCl2, 0.06; and

106

glucose, 50 were used. Strains kept at -80°C was transferred to 25 mL of LB seed medium at pH 7.2 in

107

250 mL shake flasks and incubated at 37°C for 6–8 h on a rotary shaker at 200 rpm. Then, 5% of the

108

seed culture was inoculated into the RM1 fermentation medium. The shake flask fermentation was

109

conducted in a 500 mL Erlenmeyer flask containing around 75 mL fermentation medium and

110

incubated at 37°C and 200 rpm for 32 h. The experiments were conducted in triplicate. To buffer the

111

pH in the BSGN10 fermentation processes, powdered calcium carbonate (CaCO 3) (1.5, 3.0, and 4.5%)

112

was added into RM1 as a neutralizing agent.

6

113

114

2.6. Fed-batch fermentation in a 3-L bioreactor

The glucose feedback-controlled fermentation was carried out in a 3 L stirred fermenter (LiFlus

115

GM BioTRON, Bucheon, Korea) using a feed-back control glucose feeding strategy integrated with

116

step-wise regulation of the dissolved oxygen level (Zhu et al., 2015) with a working volume of 1.5 L.

117

The fermentation medium (RM2) was prepared with the following (g/L): yeast extract, 12; corn syrup

118

powder, 18; (NH4)2SO4, 6; K2HPO4·3H2O, 18.75; KH2PO4, 2.5; MgSO4, 3; FeSO4·7H2O, 0.06; CaCl2, 0.06;

119

CaCO3, 30. In briefly, fed-batch fermentation was initiated with an initial glucose concentration of 20

120

g/L. Glucose concentration was maintained at 5 g/L by feeding high concentration of glucose (500 g/L)

121

using an automatic glucose analyzer (SBFC-2010, Zhonghui Science and Technology Ltd., China). The

122

total glucose concentration used was 150 g/L. The aeration rate was 1.5 vvm. The agitation speed was

123

changed between 500 and 800 rpm to regulate the DO level: 0–6 h, 30%; 6–27 h, 50%; 27–60 h, 40%.

124

During fermentation, the temperature was maintained at 37 °C using the temperature control system

125

on fermenter and the pH was maintained at 7.4 by automatic addition of ammonium hydroxide (50 %

126

[vol/vol]) using a computer-coupled peristaltic pump. The total fermentation period was 56 hours.

127

2.7. Analytical methods

128

The concentration of GlcNAc, acetate, and acetoin in the fermentation broth was analyzed by

129

HPLC as described above. The glucose concentration in the supernatant was measured using a

130

glucose-glutamate analyzer (SBA-40C, Biology Institute of Shandong Academy of Sciences, Jinan,

131

China). Cell growth was monitored by measuring the absorbance at 600 nm (OD 600). The correlation

132

between OD600 and dry cell weight (DCW) was 1 OD600 = 0.35 DCW (g/L) (Liu et al., 2013c).

7

133

2.8. Statistical analysis

All the experiments were performed independently at least three times. The statistical analysis

134 135

w

p f m dw h

d

’ t-test. P values of <0.05 were considered statistically significant.

136

3. Results and discussion

137

3.1. Blocking acetoin overflow inhibited cell growth, GlcNAc production, and glucose uptake

138

Acetoin overflow can be significantly blocked in the three following approaches: 1) knocking out

139

alsS and alsD; 2) knocking out alsR; 3) or knocking out alsR, alsS, and alsD. The acetoin concentration

140

was reduced to 0.7 g/L, which represented a decrease by 95% of that in the parent strain (16.7 g/L)

141

with any of the three approaches for blocking acetoin overflow (Fig. 2A). The conversion rate of

142

glucose to acetoin decreased from 0.33 to 0.04 g acetoin/g glucose via blocking acetoin formation.

143

One phenomenon that should be noted here was that knockout of alsD alone did not block acetoin

144

synthesis during the growth phase. The acetoin concentration of the alsD mutant (10.9 g/L) was

145

reduced to 65% of that of BSGN5 (16.7 g/L) (Fig. 2. A). The spontaneous conversion of acetolactate to

146

acetoin presents a possible route for the generation of acetoin in the alsD mutant. The acetoin

147

produced by other mutants was due to the expression of ilvBH in the branched-chain amino acid

148

synthesis pathway, which is also responsible for acetolactate synthesis (Renna et al., 1993).

149

However, an unexpected substantial inhibition of cell growth and reduction of GlcNAc titer were

150

observed after blocking acetoin overflow. The DCW and GlcNAc titer, 2.55 g/L and 1.7 g/L, decreased

151

to 35% and 25% of that in the parent strain, respectively (Fig. 2. B, C). Meanwhile, the rate of glucose

152

consumption was lower, with decreased glucose uptake from 50 g/L to 18 g/L in comparison with

8

153

BSGN5 (Fig. 2. D). These physiological changes reminded us the role of acetoin in maintaining

154

intracellular pH by converting pyruvic acid into a neutral species (Ramos et al., 2000).

155

As expected, we observed a substantial decline in pH in the fermentation broth of these mutants,

156

dropping from 6.45 to 4.95 (Fig. 2. E), which indicated that some unknown acidic metabolites were

157

being secreted into the broth. Based on pathway analysis and literature mining, the most likely acidic

158

metabolite was acetate (Shirk et al., 2002; Toya et al., 2015).

159

3.2. Blocking acetoin production led to overflow of acetate

160

To clarify whether or not the acid metabolite was acetate, the mutant BSGN10 was selected for 1

161

further studies via HPLC, MS, and H-NMR. HPLC analysis revealed that the retention time of the

162

unknown acid metabolite was identical with that of acetate (Fig. 3. A). Additionally, MS analysis of the

163

supernatant depicted one distinct major molecular ion peak at 59 m/z, which corresponds to H3CCOO

164

(Fig. 3. B). Furthermore, H-NMR analysis showed a major peak corresponding to the methyl protons

165

of H3CCOO (Fig. 3. C). Thus, these results demonstrated clearly that blocking acetoin production leads

166

to overflow of acetate in the BSGN5 strain, which accumulated in low concentrations without blocking

167

acetoin formation (1.4 g/L of acetate). The culture supernatant of these mutants were then analyzed

168

by HPLC to quantify acetate, which demonstrated that acetate accumulated in the broth reaching 6

169

g/L, corresponding to 0.34 g acetate/g glucose (Fig. 3. D).

170

-

1

-

Acetate is a lipophilic agent that is harmful to cell growth because it induces an uncoupling

171

mechanism and dissipation of the proton motive force (Russell & DiezGonzalez, 1998). Moreover,

172

acetate accumulation leads to greatly reduced heterologous gene expression (De Anda et al., 2006;

173

Goel et al., 1999). From an evolutionary perspective, the harmful effects of acetate were avoided by

174

overflow of acetoin, the neutral fermentation end-product, which played a crucial role in bacterial

9

175

fitness. This is because synthesis of acetoin reduced intracellular carbon flux from pyruvate to acetate,

176

allowing the mutants to metabolize glucose for fast growth without a drop in pH (Ali et al., 2001;

177

Fradrich et al., 2012; Schilling et al., 2007). To minimize the toxicity of acetate, adding powdered

178

calcium carbonate to neutralize the toxic acid to prevent the drop in pH was implemented.

179

3.3. Addition of calcium carbonate as a neutralizing agent promoted GlcNAc production

180

Powdered calcium carbonate was added to RM1 medium to serve as an acid buffer pool to neutralize

181

the acetic acids during the BSGN10-P43-CeGNA1 fermentation process. As shown in Fig. 4, the

182

observed growth defect in BSGN10-P43-CeGNA1 was largely complemented by the addition of

183

calcium carbonate. When the calcium carbonate load reached 30 g/L, the maximal DCW of 6.1 g/L

184

was obtained (Fig. 4. A). Meanwhile, a higher GlcNAc titer was obtained, which increased from 6.2 to

185

7.4 g/L in comparison with that of the BSGN5 strain (Fig. 4. B). As noted, addition of calcium carbonate

186

promoted glucose consumption. Further increments in the calcium carbonate loading did not have

187

significant effects on the fermentation; therefore, 30 g/L calcium carbonate was used in the following

188

experiments. The calcium carbonate promoted GlcNAc production in BSGN10 due to its buffering

189

effects and its ability to influence key cellular processes such as the EMP pathway and TCA cycle (Han

190

et al., 2013; Salek et al., 2015).

191

3.4. GlcNAc production in a 3-L bioreactor

192

Next, we investigate the ability of BSGN10 to produce GlcNAc as well as its growth characteristics

193

in a 3-L fermenter by automatically maintaining the pH of the broth at 7.4. During the GlcNAc

194

fermentation in pH stats with calcium carbonate, the maximal cell density was 23.5 g/L, which was

195

15.2% higher than the value obtained for the parent strain BSGN5-P43-CeGNA1 (20.4 g/L). Meanwhile,

10

196

the highest GlcNAc titer of 48.9 g/L was almost 1.35-fold higher than that of BSGN5 under the same

197

conditions (Fig. 5. A, B), with the GlcNAc yield increased from 0.23 to 0.32 g GlcNAc/g glucose. Though

198

overflowed acetate reached 12.1 g/L in the fermentation broth, which was 5-fold higher compared

199

with that of BSGN5, it was reused subsequently and its deleterious effect on cell growth was alleviated

200

by the addition of ammonium hydroxide (Fig. 5. D). Together, these results indicated that pH-stat

201

control of fermentation relieved acetate toxicity and restored cell growth, and blocking acetoin

202

overflow redirected more carbon flux for GlcNAc production.

203

Consistent with our deduction, the toxicity of acetate led to the observed defects of BSGN10

204

during shake flask fermentation. This result also provides an additional metabolic engineering

205

modification target, the acetate overflow pathway, to promote GlcNAc production. As acetoin and

206

acetate are both derived from pyruvate, blocking the synthesis of acetate further may lead to

207

pyruvate overflow (Fig. 1). Thus, it was uncertain whether blocking acetate synthesis to remove its

208

deleterious effect on the background of BSGN10 can further push carbon flux to the GlcNAc synthesis

209

module or lead to pyruvate overflow.

210

3.5. Blocking acetate production led to pyruvate overflow

211

To investigate the above speculation, mutants deficient in genes for acetate synthesis were

212

studied. In B. subtilis, the Pta-AckA pathway is the major route for acetate synthesis during aerobic

213

growth (Grundy et al., 1993). The deletion of pta and ackA, each of which encodes enzyme to produce

214

acetate from acetyl-CoA, markedly decreased acetate concentration (Fig. 6. A). Unfortunately, blocking

215

acetate overflow did not result in a recovery of cell growth and GlcNAc production, but rather it

216

further inhibited these processes when compared to that of BSGN10. Furthermore, the DCW

217

decreased by 14.5%, from 6.1 to 5.2 g/L, and the GlcNAc titer decreased by 12.1%, from 7.4 g/L to 6.5

11

218

g/L (Fig. 6. B, C). Additionally, glucose uptake was also suppressed, with glucose intake decreasing

219

from 34.4 g/L to 28.1 g/L (Fig. 6. D).

220

We noted that pyruvate overflowed amounted to 6 g/L as analyzed by HPLC (Fig. 6. E). Pyruvate

221

accumulation may inhibit glucose uptake through the phosphoenolpyruvate (PEP): carbohydrate

222

phosphotransferase system (PTS), because the ratio of intracellular PEP/PYR affects the

223

phosphorylation of EIIA . A decrease in this ratio leads to a decrease in EIIA phosphorylation, thus

224

lowering glucose uptake (Deutscher et al., 2014; Himmel et al., 2012). Furthermore, a previous

225

studied demonstrated that blocking the synthesis of acetoin resulted in at least a 2.5-fold increase in

226

intracellular pyruvate concentration, whereas PEP concentration remained the same (Toya et al.,

227

2015). This result also supported our deduction that the low PEP/PYR ratio caused an inhibition of

228

glucose uptake in BSGN11 as well as the observed inhibition of cell growth and GlcNAc production.

229

Glc

Glc

Other than the inhibition of glucose uptake, energy metabolism may also be affected by blocking

230

acetoin and acetate overflow because the overflow of acetoin participates in the regulation of the

231

NAD /NADH ratio (Xiao & Xu, 2007). When the synthesis of acetoin was blocked, NADH formed in the

232

synthesis of biomass and secondary fermentation products, such as pyruvic and succinic acid, may not

233

be efficiently re-oxidized to NAD , thus affecting the NAD /NADH balance as well as energy supply. As

234

described elsewhere, a limited respiratory capacity or inefficient energy supply may be other factors

235

that can lead to cell growth inhibition (Marshall et al., 2016; Vemuri et al., 2007). This inhibitory effect

236

was especially noticeable when acetate overflow was further blocked, of which its biosynthesis is

237

accompanied by equimolar substrate-level ATP generation and provides additional energy when the

238

respiration capacity is saturated (Fig. 1) (Chang et al., 1999). Together, these results suggested that

+

+

+

12

239

maybe a limited energy supply constrains fast cell growth, which then constrains GlcNAc production

240

after blocking the overflow metabolism.

241

This is in agreement with recent publications, which suggested that the proteome cost of energy

242

biogenesis by respiration exceeds that by fermentation and overflow metabolism is a programmed

243

global response used by cells to balance the conflicting proteomic demands of energy biogenesis and

244

biomass synthesis for rapid growth (Basan et al., 2015; Polz & Cordero, 2016). Based on this theory,

245

blocking overflow metabolism forces cells to mainly use the slow high-yield respiration pathway for

246

energy biogenesis. However, the proteome cost of energy biogenesis by respiration exceeds that by

247

fermentation, thus leading to suboptimal growth.

248

In this study, the toxicity of acetate was alleviated by the addition of calcium carbonate, a

249

neutralizing agent. However, in the BSGN5 parent strain, the toxicity was efficiently avoided by guiding

250

the carbon flux to neutral acetoin. It is interesting to note that the accumulation of acetate or low pH

251

induced synthesis of acetoin, so as to direct carbon flux to the neutral species, thereby avoiding

252

acetate toxicity (Biswas et al., 2012). The trade-off between the synthesis of acetoin and acetate

253

suggests that it might be a good way to further improve GlcNAc titer and yield by expressing the key

254

enzymes glutamine-fructose-6-phosphate transaminase (GlmS) and glucosamine 6-phosphate

255

N-acetyltransferase (GNA1) under the control of the promoter PalsS, which is induced through

256

interactions between acetate and the transcription factor AlsR (Fradrich et al., 2013; Fradrich et al.,

257

2012). Thus, the expression of GlmS will pull the fructose-6-phosphate (F6P) flux from EMP to

258

D-glucosamine-6-phosphate (GlcN-6P) synthesis, reducing carbon flux to pyruvate. Expression of

259

GNA1 will pull AcCoA flux to N-acetyl-D-glucosamine-6-phosphate (GlcNAc-6P), promoting pyruvate

260

utilization. Finally, both high-yield GlcNAc production and overflow metabolism alleviation will be

13

261

achieved. We are now attempting to engineer the transcription factor AlsR and promoter P alsS to

262

control the expression of GlmS and GNA1 to enhance GlcNAc production.

263

4. Conclusions

In summary, the findings showed that blocking acetoin overflow pushed more carbon flux from

264 265

fructose-6-phosphate to GlcNAc synthesis pathway and greatly improved GlcNAc production, with the

266

GlcNAc titer and yield increasing by 38 % and 39 %, respectively. It is a promising strategy for

267

minimizing overflow metabolism for biochemical production in B. subtilis. For instance, strategies for

268

blocking acetoin overflow may be applied to metabolic engineering of B. subtilis for producing other

269

value-added commodity chemicals derived from pyruvate and acetyl-CoA. To further improve GlcNAc

270

production, expression of GlmS and GNA1 should be enhanced and optimized to pull carbon flux to

271

GlcNAc synthesis pathway.

272

Appendix A. Supplementary data

273

Supplementary data of this work can be found in online of this paper.

274

Acknowledgements

275

This work is financially supported by the National Natural Science Foundation (31622001, 31671845,

276

21676119, 31600068), the Natural Science Foundation of Jiangsu

277

F

278

111-2-06) and the China Postdoctoral Science Foundation (2016M600363, 2017T100327).

d m

R

hF

d f

h

U

” (JU R

279

14

(BK 0 60 76), “ h 7 5), h

j

(

.

280

References

281

1. Ali, N.O., Bignon, J., Rapoport, G., Debarbouille, M. 2001. Regulation of the acetoin catabolic

282 283

pathway is controlled by sigma L in Bacillus subtilis. J Bacteriol, 183(8), 2497-2504. 2. Basan, M., Hui, S., Okano, H., Zhang, Z.G., Shen, Y., Williamson, J.R., Hwa, T. 2015. Overflow

284

metabolism in Escherichia coli results from efficient proteome allocation. Nature, 528(7580),

285

99-104.

286

3. Biswas, R., Yamaoka, M., Nakayama, H., Kondo, T., Yoshida, K.-i., Bisaria, V.S., Kondo, A. 2012.

287

Enhanced production of 2,3-butanediol by engineered Bacillus subtilis. Appl Microbiol

288

Biotechnol, 94(3), 651-658.

289

4. Chang, D.E., Shin, S., Rhee, J.S., Pan, J.G. 1999. Acetate metabolism in a pta mutant of Escherichia

290

coli W3110: Importance of maintaining acetyl coenzyme a flux for growth and survival. J

291

Bacteriol, 181(21), 6656-6663.

292

5. De Anda, R., Lara, A.R., Hernandez, V., Hernandez-Montalvo, V., Gosset, G., Bolivar, F., Ramirez, O.T.

293

2006. Replacement of the glucose phosphotransferase transport system by galactose

294

permease reduces acetate accumulation and improves process performance of Escherichia

295

coli for recombinant protein production without impairment of growth rate. Metab Eng, 8(3),

296

281-290.

297

6. Deutscher, J., Ake, F.M.D., Derkaoui, M., Zebre, A.C., Cao, T.N., Bouraoui, H., Kentache, T., Mokhtari,

298

A., Milohanic, E., Joyet, P. 2014. The bacterial phosphoenolpyruvate:carbohydrate

299

phosphotransferase system: regulation by protein phosphorylation and

300

phosphorylation-dependent protein-protein interactions. Microbiol Mol Biol Rev, 78(2),

301

231-256.

15

302 303

7. du Souich, P. 2014. Absorption, distribution and mechanism of action of SYSADOAS. Pharmacology & Therapeutics, 142(3), 362-374.

304

8. Fradrich, C., Krausze, J., Quade, N., Heinz, D., Jahn, D., Hartig, E. 2013. Purification, crystallization

305

and preliminary X-ray analysis of the effector domain of AlsR, an LysR-type transcriptional

306

regulator from Bacillus subtilis. Acta Crystallogr Sect F Struct Biol Cryst Commun, 69(5),

307

581-584.

308

9. Fradrich, C., March, A., Fiege, K., Hartmann, A., Jahn, D., Hartig, E. 2012. The transcription factor

309

AlsR binds and regulates the promoter of the alsSD operon responsible for acetoin formation

310

in Bacillus subtilis. J Bacteriol, 194(5), 1100-1112.

311 312 313

10. Fujita, Y. 2009. Carbon catabolite control of the metabolic network in Bacillus subtilis. Biosci Biotechnol Biochem, 73(2), 245-259. 11. Goel, A., Lee, J., Domach, M.M., Ataai, M.M. 1999. Metabolic fluxes, pools, and enzyme

314

measurements suggest a tighter coupling of energetics and biosynthetic reactions associated

315

with reduced pyruvate kinase flux. Biotechnol Bioeng, 64(2), 129-134.

316

12. Grundy, F.J., Waters, D.A., Takova, T.Y., Henkin, T.M. 1993. Identification of genes involved in

317

utilization of acetate and acetoin in Bacillus subtilis. Mol Microbiol, 10(2), 259-271.

318

13. Han, B., Ujor, V., Lai, L.B., Gopalan, V., Ezeji, T.C. 2013. Use of proteomic analysis to elucidate the

319

role of calcium in acetone-butanol-ethanol fermentation by Clostridium beijerinckii NCIMB

320

8052. Appl Environ Microbiol, 79(1), 282-293.

321

14. Himmel, S., Zschiedrich, C.P., Becker, S., Hsiao, H.H., Wolff, S., Diethmaier, C., Urlaub, H., Lee, D.,

322

Griesinger, C., Stulke, J. 2012. Determinants of interaction specificity of the Bacillus subtilis

16

323

GlcT antitermination protein-functionality and phosphorylation specificity depend on the

324

arrangement of the regulatory domains. J Biol Chem, 287(33), 27731-27742.

325

15. Hochberg, M.C., Martel-Pelletier, J., Monfort, J., Moller, I., Castillo, J.R., Arden, N., Berenbaum, F.,

326

Blanco, F.J., Conaghan, P.G., Domenech, G., Henrotin, Y., Pap, T., Richette, P., Sawitzke, A., du

327

Souich, P., Pelletier, J.P., Grp, M.I. 2016. Combined chondroitin sulfate and glucosamine for

328

painful knee osteoarthritis: a multicentre, randomised, double-blind, non-inferiority trial

329

versus celecoxib. Annals Rheum Diseases, 75(1), 37-44.

330

16. Kumar, P., Tambe, P., Paknikar, K.M., Gajbhiye, V. 2017. Folate/N-acetyl glucosamine conjugated

331

mesoporous silica nanoparticles for targeting breast cancer cells: A comparative study.

332

Colloids Surfaces B-Biointerfaces, 156, 203-212.

333

17. Liu, L., Liu, Y.F., Shin, H.D., Chen, R., Li, J.H., Du, G.C., Chen, J. 2013a. Microbial production of

334

glucosamine and N-acetylglucosamine: advances and perspectives. Appl Microbiol Biotechnol,

335

97(14), 6149-6158.

336

18. Liu, L., Liu, Y.F., Shin, H.D., Chen, R.R., Wang, N.S., Li, J.H., Du, G.C., Chen, J. 2013b. Developing

337

Bacillus spp. as a cell factory for production of microbial enzymes and industrially important

338

biochemicals in the context of systems and synthetic biology. Appl Microbiol Biotechnol,

339

97(14), 6113-6127.

340 341 342

19. Liu, Y.F., Liu, L., Shin, H.D., Chen, R.R., Li, J.H., Du, G.C., Chen, J. 2013c. Pathway engineering of Bacillus subtilis for microbial production of N-acetylglucosamine. Metab Eng, 19, 107-115. 20. Liu, Y.F., Zhu, Y.Q., Li, J.H., Shin, H.D., Chen, R.R., Du, G.C., Liu, L., Chen, J. 2014a. Modular pathway

343

engineering of Bacillus subtilis for improved N-acetylglucosamine production. Metab Eng, 23,

344

42-52.

17

345

21. Liu, Y.F., Zhu, Y.Q., Ma, W.L., Shin, H.D., Li, J.H., Liu, L., Du, G.C., Chen, J. 2014b. Spatial modulation

346

of key pathway enzymes by DNA-guided scaffold system and respiration chain engineering for

347

improved N-acetylglucosamine production by Bacillus subtilis. Metab Eng, 24, 61-69.

348

22. Longo, D.L., Moustaghfir, F.Z., Zerbo, A., Consolino, L., Anemone, A., Bracesco, M., Aime, S. 2017.

349

EXCI-CEST: Exploiting pharmaceutical excipients as MRI-CEST contrast agents for tumor

350

imaging. Inter J Pharm, 525(1), 275-281.

351

23. Marshall, D.D., Sadykov, M.R., Thomas, V.C., Bayles, K.W., Powers, R. 2016. Redox imbalance

352

underlies the fitness defect associated with inactivation of the Pta-AckA pathway in

353

Staphylococcus aureus. J Proteome Res, 15(4), 1205-1212.

354 355

24. Ozturk, S., Calik, P., Ozdamar, T.H. 2016. Fed-batch biomolecule production by Bacillus subtilis: A state of the art review. Trends Biotechnol, 34(4), 329-345.

356

25. Polz, M.F., Cordero, O.X. 2016. Genomics of metabolic trade-offs. Nature Microbiol, 1(11).

357

26. Ramos, H.C., Hoffmann, T., Marino, M., Nedjari, H., Presecan-Siedel, E., Dreesen, O., Glaser, P.,

358

Jahn, D. 2000. Fermentative metabolism of Bacillus subtilis: Physiology and regulation of gene

359

expression. J Bacteriol, 182(11), 3072-3080.

360

27. Renna, M.C., Najimudin, N., Winik, L.R., Zahler, S.A. 1993. Regulation of the Bacillus subtilis alss,

361

alsd, and alsr genes involved in post-exponential-phase production of acetoin. J Bacteriol,

362

175(12), 3863-3875.

363 364

28. Rivlin, M., Navon, G. 2016. Glucosamine and N-acetyl glucosamine as new CEST MRI agents for molecular imaging of tumors. Sci Rep, 6, 32648.

18

365

29. Russell, J.B., DiezGonzalez, F. 1998. The effects of fermentation acids on bacterial growth, in:

366

Advances in Microbial Physiology, Vol 39, (Ed.) R.K. Poole. Academic Press Ltd-Elsevier

367

Science Ltd. London, pp. 205-234.

368 369 370

30. Salek, S.S., van Turnhout, A.G., Kleerebezem, R., van Loosdrecht, M.C.M. 2015. pH control in biological systems using calcium carbonate. Biotechnol Bioeng, 112(5), 905-913. 31. Schilling, O., Frick, O., Herzberg, C., Ehrenreich, A., Heinzle, E., Wittmann, C., Stulke, J. 2007.

371

Transcriptional and metabolic responses of Bacillus subtilis to the availability of organic acids:

372

Transcription regulation is important but not sufficient to account for metabolic adaptation.

373

Appl Environ Microbiol, 73(2), 499-507.

374

32. Shirk, M.C., Wagner, W.P., Fall, R. 2002. Isoprene formation in Bacillus subtilis: A barometer of

375

central carbon assimilation in a bioreactor? Biotechnol Prog, 18(5), 1109-1115.

376

33. Toya, Y., Hirasawa, T., Ishikawa, S., Chumsakul, O., Morimoto, T., Liu, S.H., Masuda, K., Kageyama, Y.,

377

Ozaki, K., Ogasawara, N., Shimizu, H. 2015. Enhanced dipicolinic acid production during the

378

stationary phase in Bacillus subtilis by blocking acetoin synthesis. Biosci Biotechnol Biochem,

379

79(12), 2073-2080.

380

34. Vemuri, G.N., Eiteman, M.A., McEwen, J.E., Olsson, L., Nielsen, J. 2007. Increasing NADH oxidation

381

reduces overflow metabolism in Saccharomyces cerevisiae. Proceedings of the National

382

Academy of Sciences of the United States of America, 104(7), 2402-2407.

383

35. Xiao, Z., Xu, P. 2007. Acetoin metabolism in bacteria. Crit Rev Microbiol, 33(2), 127-140.

384

36. Xu, X.F. 2016. Screening of glucosamine-6-phosphate N-acetyltransferase for N-acetylglucosamine

385

production and enzyme kinetics analysis. Thesis for Master's Degree, Jiangnan University.

19

386

37. Xue, G.P., Johnson, J.S., Dalrymple, B.P. 1999. High osmolarity improves the electro-transformation

387

efficiency of the gram-positive bacteria Bacillus subtilis and Bacillus licheniformis. J Microbiol

388

Met, 34(3), 183-191.

389 390 391

38. Yan, X., Yu, H.J., Hong, Q., Li, S.P. 2008. Cre/lox system and PCR-based genome engineering in Bacillus subtilis. Appl Environ Microbiol, 74(17), 5556-5562. 39. Yang, S., Du, G., Chen, J., Kang, Z. 2017. Characterization and application of endogenous

392

phase-dependent promoters in Bacillus subtilis. Appl Microbiol Biotechnol, 101(10),

393

4151-4161.

394

40. Zhu, Y., Liu, Y., Li, J., Shin, H.D., Du, G., Long, L., Jian, C. 2015. An optimal glucose feeding strategy

395

integrated with step-wise regulation of the dissolved oxygen level improves N

396

-acetylglucosamine production in recombinant Bacillus subtilis. Bioresour Technol, 177,

397

387-392.

398

20

399

Figure captions:

400

Fig. 1. GlcNAc production pathway and overflow metabolism in the recombinant B. subtilis strain,

401

BSGN5. Abbreviations: Glc, glucose; G6P, glucose-6-phosphate; F6P, fructose-6-phosphate; GlcN-6P,

402

glucosamine-6- phosphate; GlcNAc-6P, N-acetyl-D-glucosamine; Lac, lactate; 2-AL, 2-acetolactate;

403

AcCoA, acetyl coenzyme A; Ac-Pi, acetyl phosphate; Ace, acetate; OAA, oxaloacetic acid; Cit, citric acid;

404

Suc-CoA, succinyl coenzyme A; Suc, succinic acid; Fum, fumaric acid; Mal, malic acid. BsglmS, encoding

405

glutamine-fructose-6-phosphate transaminase; Cegna1, encoding glucosamine 6-phosphate

406

N-acetyltransferase from Caenorhabditis elegans; alsS and ilvBH, encoding acetolactate synthase; alsD,

407

encoding alpha-acetolactate decarboxylase; bdhA, encoding (R,R)-butanediol dehydrogenase; pta,

408

encoding phosphotransacetylase; ackA, encoding acetate kinase.

409 410

Fig. 2. Shake flask fermentation of BSGN5 and its derivatives with deletion of genes responsible for

411

acetoin synthesis. Effects of individually or combinatorially deletion of acetoin synthetic or regulatory

412

genes on acetoin synthesis (A), cell growth (B), GlcNAc production (C), glucose consumption (D), and

413

pH (E).

414 1

415

Fig. 3. Identification of acetate with HPLC, MS and H-NMR. (A) HPLC chromatogram of

416

BSGN5-P43-CeGNA1 culture supernatant, acetoin, BSGN10-P43-CeGNA1 culture supernatant and

417

acetate; (B) MS chromatogram of BSGN10-P43-CeGNA1 culture supernatant; (C) H-NMR

418

chromatogram of BSGN10-P43-CeGNA1 culture supernatant; (D) acetate concentration at 48h of the

419

shake flask fermentation of the BSGN5-P43-CeGNA1 and its derivatives.

1

420 421

Fig. 4. Shake flask fermentation of BSGN10 in RM1 medium with and without calcium carbonate.

422

Effects of calcium carbonate addition on cell growth (A), GlcNAc production (B), glucose consumption

423

(C).

424 425

Fig. 5. Time course of GlcNAc fed-batch fermentation by BSGN10 in a 3-L bioreactor. (A) cell growth

426

(squares), glucose concentration (diamonds), pH value (triangles); (B) GlcNAc titer; (C) acetoin and (D)

427

acetate concentration. BSGN5-P43-CeGNA1, filled symbols; BSGN10-P43-CeGNA1, open symbols.

428

Fed-batch fermentation was initiated with an initial glucose of 20 g/L. During the fermentation, the

429

glucose was maintained at 5 g/L. The total glucose concentration used was 150 g/L. The temperature,

430

pH and aeration rate were 37 C, 7.4 and 1.5vvm, respectively.

o

431

21

432

Fig. 6. Shake flask fermentation of BSGN10 and its derivatives with knockout of genes responsible for

433

acetate synthesis. Effects of individually or combinatorially deletion of acetate synthetic genes on

434

acetate production (A), cell growth (B), GlcNAc titer (C), glucose consumption (D), and pyruvate

435

production (E).

22

436

Tables

437

Table 1. Strains and plasmids used in this study. Characteristics

Reference

BSGN5

B. subtilis 168 derivate, ΔnagPΔgamPΔgamAΔnagAΔnagBΔldh::lox72

(Liu et al., 2014a)

BSGN6

BSGN5 derivate, BSGN5Δpta::lox72

(Liu et al., 2014a)

BSGN10

B

5d

,B

5ΔalsRSD::lox72

This work

B

5d

,B

5ΔalsR::lox72

This work

Strains

B

5Δ

R

B

5Δ

B

5d

,B

5ΔalsS::lox72

This work

B

5Δ

B

5d

,B

5ΔalsD::lox72

This work

B

5Δ

B

5d

,B

5ΔalsSD::lox72

This work

B

0Δ k

B

0d

,B

0ΔackA::lox72

This work

B

0Δp

B

0d

,B

0Δp ::lox72

This work

B

0d

,B

0Δp ΔackA::lox72

This work

BSGN11 Plasmids p7Z6

pMD18-T containing lox71-zeo-lox66 cassette r

r

Yan et al.(2008)

pTSC

Em Amp ; temperature sensitive in B. subtilis

Yan et al.(2008)

pP43-CeGNA1

pP43NMK derivate with CeGNA1 cloned

(Xu, 2016)

438

23

439 Primer

Table 2. Primers used in this study. Sequence

Gene Knockout alsSD-L-F

CCATGTATAGAGTAGGCCATGCTTCTTTAGC

alsSD-L-R

AGGATCCCCGGGTACCGAGCTCCACCCTCACTCCTTATTATGCATTTTAAACGTAAAA

alsSD-Z-F

TTTTACGTTTAAAATGCATAATAAGGAGTGAGGGTGGAGCTCGGTACCCGGGGATCCT

alsSD-Z-R

CCCTGCTAAAAGGGGCTTTCTTTTTTTCTTGCTTGCATGCCTGCAGGTCGAC

alsSD-R-F

GTCGACCTGCAGGCATGCAAGCAAGAAAAAAAGAAAGCCCCTTTTAGCAGGG

alsSD-R-R

CTACTGCGCTGTCAGAAGCAAAATCAG

pta-L-F

CGCATTTCGTACACTAGCTTCTTGAATTG

pta-L-R

AGGATCCCCGGGTACCGAGCTCAATAAAACCTCCTCAAAAAGTTACAAAAACGC

pta-Z-F

GCGTTTTTGTAACTTTTTGAGGAGGTTTTATTGAGCTCGGTACCCGGGGATCCT

pta-Z-R

CGAGAGCTGCCATTGTCTTCAATTTTAGCTTGCATGCCTGCAGGTCGAC

pta-R-F

GTCGACCTGCAGGCATGCAAGCTAAAATTGAAGACAATGGCAGCTCTCG

pta-R-R

CTGGATAGAACCGAAAGTCCCATGC

ackA-L-F

CTGGGAACTGGGAACCTTCTGTTTACC

ackA-L-R

CTGTTTCCTGTGTGAAATTGTTATCCGCTCGATTGACGCTCCTTTATACTCTGTATCAACA

ackA-Z-F

TGTTGATACAGAGTATAAAGGAGCGTCAATCGAGCGGATAACAATTTCACACAGGAAACAG

ackA-Z-R

ACATTCAAGAGAATGTGCTTTCATGCGATGCCAGGGTTTTCCCAGTCACGAC

ackA-R-F

GTCGTGACTGGGAAAACCCTGGCATCGCATGAAAGCACATTCTCTTGAATGT

ackA-R-R

GCTGAACAAATGCCAAATCTTTGCCC 440

Underlined letters represent homologous sequences for fusion PCR.

441 442 443 444 445 446 447 448

24

449

450 451

(Fig. 1)

452 453

25

454 455

(Fig.2)

456 457 458 459 460 461

26

462 463

(Fig.3)

464

27

465 466

(Fig.4)

467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483

28

484 485

(Fig.5)

486 487 488 489 490 491 492 493

29

494 495 496

(Fig.6)

30

497

●

Competitive overflow of acetoin was blocked to promote GlcNAc production;

498

●

Addition of calcium carbonate as a neutralizing agent minimized acetate toxicity;

499

●

The GlcNAc titer in 3-L bioreactor increased by 38% by metabolic engineering of overflow

500

metabolism.

501

31