FEMS Microbiology Letters Advance Access published April 28, 2015
1
Identification of riboflavin: revealing different metabolic
2
characteristics between Escherichia coli BL21(DE3) and
3
MG1655
4 5
Running title: Riboflavin accumulation in Escherichia coli BL21(DE3)
6
Xinran Wang, Qian Wang*, Qingsheng Qi
8
National Glycoengineering Research Center, State Key Laboratory of Microbial
9
Technology,
10
Shandong University, Jinan 250100, P. R. China
11 12
* Corresponding author: Tel & Fax: +86-531-88365628;
13
E-mail address: [email protected]
14 15
1
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
7
15
Abstract
17
There are many physiological differences between Escherichia coli B and K-12 strains,
18
owing to their different origins. Deeper insight into the metabolic and regulative
19
mechanisms of these strains will inform improved usage of these industrial workhorses.
20
In the present study, we observed that BL21 fermentation broth gradually turned yellow
21
during cultivation. By spectral analysis and liquid chromatography-mass spectroscopy
22
identification, we confirmed for the first time that the yellow substance accumulated in
23
the fermentation broth is riboflavin. Comparing the enzyme sequences involved in
24
riboflavin metabolism between BL21 and MG1655, we identified a site mutation on the
25
115 residue of bifunctional riboflavin kinase/FMN adenylyltransferase (RibF) in BL21.
26
This His115Leu mutation was found to reduce enzyme activity to 55% that of MG1655,
27
which is probably one reason for riboflavin accumulation in BL21. Quantitative PCR
28
analysis showed that genes of the entire branch of the riboflavin and FAD biosynthesis
29
pathways in BL21 were up-regulated. Several physiological and metabolic characteristics
30
of BL21 and MG1655 were found to be different, and may also be related to the
31
riboflavin accumulation.
32
2
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
16
32
Introduction
34
Escherichia coli has been a workhorse, not only in fundamental biological studies but
35
also in various biotechnological applications. Almost all laboratory strains of E. coli are
36
derivatives of nonpathogenic K-12 or B strains. Since two K-12 strains - MG1655 and
37
W3110 are sequenced in 1997 (Blattner, et al., 1997, Hayashi, et al., 2006), Increasing
38
omics analyses (Choi, et al., 2003, Franchini & Egli, 2006, Han & Lee, 2006,
39
Nandakumar, et al., 2006, Ishii, et al., 2007) and in silico metabolic modeling (Covert, et
40
al., 2004, Feist, et al., 2007) have been performed in K-12. In contrast, there have been
41
few studies for B strains, although they have been widely used for large-scale production
42
of recombinant proteins, ethanol, and other biomolecules (Choi, et al., 2006). Recently,
43
the genome sequences of two B strains, REL606 and BL21(DE3), have been determined.
44
Comparative omics analysis of B and K-12 show that they are very similar (Schneider, et
45
al., 2002) and closely related (Yoon, et al., 2012), although the strains show many
46
important behavioral differences in cell metabolism and physiology (Swartz, 1996). The
47
B strains have a lower acetate production than K-12 derivatives. A common explanation
48
for that is the activated glyoxylate shunt, which is the main pathway for acetate utilization.
49
Deletions of arcA and iclR in E. coli MG1655 endowed the host with central metabolic
50
fluxes similar to that of E. coli BL21(DE3) under glucose abundant conditions
51
(Waegeman, et al., 2011).
52
Riboflavin (vitamin B2), the precursor of coenzymes flavin adenine dinucleotide (FAD)
53
and flavin mononucleotide (FMN), is an essential component of basic metabolism
54
(Massey, 2000). Riboflavin is required for a wide variety of cellular processes and has a
55
key role in the metabolism of fats, ketone bodies, carbohydrates, and proteins (Depeint, et
56
al., 2006). Many microorganisms including fungi, and also plants, synthesize riboflavin,
57
while humans and animals need external riboflavin supply to avoid symptoms of
58
deficiency (Stahmann, et al., 2000). 3
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
33
Riboflavin biosynthesis has been studied in both gram-positive and gram-negative
60
bacteria, such as Bacillus subtilis and E. coli (Richter, et al., 1992, Richter, et al., 1997).
61
Riboflavin is formed from one GTP molecule and two molecules of ribulose 5-phosphate
62
(Ru-5-P), through a seven-step pathway (Bacher, et al., 2000). In B. subtilis, the encoding
63
genes of riboflavin biosynthesis reside within one rib operon (Vitreschak, et al., 2002). In
64
contrast, these genes in E. coli do not converge to a single operon but are scattered on the
65
chromosome (Bacher, et al., 2001). It has been reported that reducing the enzymatic
66
activity of bifunctional riboflavin kinase/FMN adenylyltransferase (encoded by the ribC
67
gene) could result in riboflavin overproduction in B. subtilis (Mack, et al., 1998). In E.
68
coli, this enzyme is encoded by ribF, which is an essential gene for survival. By
69
modulating the expression of ribF, riboflavin production was improved in E. coli (Lin, et
70
al., 2014). However, the exact regulation mechanism of riboflavin in E. coli has not been
71
revealed.
72
In the present study, we showed that E. coli BL21 was able to generate riboflavin under
73
normal cultivation conditions. Comparative analysis of E. coli strains BL21 and MG1655
74
indicated that riboflavin accumulation may be a reason for their metabolic difference. By
75
transcription and enzymatic analysis, we found that the riboflavin accumulation in BL21
76
is caused by not only a single mutation of ribF but also a certain regulation mechanism
77
which up-regulates the whole FAD synthesis pathway.
78 79
4
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
59
79
Material and methods
81
Bacterial strains and plasmids
82
The strains of E. coli and plasmids used in the study are summarized in Table 1. E. coli
83
DH5α was used as the host for molecular cloning. E. coli BL21(DE3) was used as the
84
host for protein expression. Plasmid pET28a was used as the expression vector for the
85
enzyme activity assay of different RibFs. Genomic DNA was isolated from wild-type E.
86
coli MG 1655 and BL21(DE3) with a TIANamp Bacterial DNA kit (Tiangen Biotech).
87
The two ribF genes were respectively amplified from the genome DNA of E. coli MG
88
1655
primer
pair
PribF
FP
89
(GCTTCCATGGCTATGAAGCTGATACGCGGCATACA)
and
PribF
RP
90
(ACTCCTCGAGAGCCGGTTTTGTTAGCCC). To construct the expression plasmids,
91
PCR fragments were digested with the restriction enzymes NcoI and XhoI and then
92
ligated into the corresponding sites of pET28a, which was cut with the same restriction
93
enzymes to generate plasmids pRibmhis (ribF from MG1655) and pRibbhis (ribF from
94
BL21(DE3)).
95
Culture conditions
96
LB medium (10 g/L tryptone, 5 g/L yeast extract and 10 g/L NaCl) was used for all DNA
97
manipulations and protein expression. During cultivation and fermentation, AM1 mineral
98
salts medium (Martinez, et al., 2007) supplemented with 10 g/L glucose was used.
99
Kanamycin (25 mg/ml) was added to provide selective pressure during cultivation when
and
BL21(DE3)
To
induce
using
the
100
necessary.
expression
101
isopropyl-β-D-thiogalactopyranoside (IPTG) was added to cultures, with a final
102
concentration of 0.2 mM.
103
Analysis methods during fermentation
104
Cell growth was detected by measuring optical density at 600 nm with a
105
spectrophotometer (Thermo Scientific Instruments LLC, USA). To detect the yellow 5
of
plasmid-borne
genes,
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
80
fluorescent substance in the fermentation broth, 2 μl of culture supernatant was spotted
107
onto filter papers and exposed to ultraviolet light, to check for the fluorescence. The
108
excitation-emission spectra were measured with a fluorescence spectrophotometer
109
(RF-5301PC, Shimadzu Corp., Japan).
110
Glucose and organic acids were determined by high-performance liquid chromatography
111
(HPLC) (Shimadzu) coupled with a refractive index detector (RID-10A), as reported
112
previously (Chen, et al., 2011). Samples were filtered through a 0.22 μm syringe filter
113
and analyzed on an HPX-87H column (Bio-Rad, USA) at 65 °C with 5 mM H2SO4 as
114
mobile phase and flow rate 0.6 ml/min. Physiological parameters of specific growth rate,
115
specific glucose uptake rate, and specific product secretion rates were determined for the
116
exponentially growing cells, according to methods described by (Sauer, et al., 1999).
117
Intracellular FAD concentration was measured using an FAD Colorimetric/Fluorometric
118
Assay Kit (Cat# K357-100; BioVision Inc., USA).
119
HPLC and LC-MS analysis of riboflavin
120
Supernatants were subsequently filtered through a 0.22-μm syringe filter, and a 20 μl
121
aliquot of the filtered solution was injected into an HPLC system (Shimadzu) equipped
122
with a UV-VIS detector (SPD-10A; excitation 470 nm; emission 530 nm). Riboflavin
123
was analyzed on an Agilent ZORBAX SB-C18 column (4.6 mm ID × 250 mm) with 30%
124
(v/v) methanol-0.1% (v/v) acetic acid as a mobile phase and flow rate 0.5 ml/min.
125
Riboflavin was qualitatively analyzed by liquid chromatography-mass spectroscopy
126
(LC-MS) using a TurboMass Gold mass spectrometer (PerkinElmer Inc., USA).
127
Riboflavin was identified by LC-MS based on retention time and fragmentation patterns
128
of known standards.
129
Protein expression of RibF
130
E. coli BL21(DE3) was transformed with plasmids pRibbhis and pRibmhis, and the
131
resulting strains were aerobically cultivated overnight at 37 °C in LB medium. For shake
132
flask culture, 2 ml of the overnight culture was added to a 300 ml Erlenmeyer flask 6
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
106
containing 50 ml LB medium at 37 °C. Gene expression was stimulated by adding 0.2
134
mM IPTG after the culture had reached an optical density at 600 nm of 0.8–1.0. Flasks
135
were transferred to 16 °C for soluble protein expression. After 20 h aerobic cultivation,
136
cells were harvested by centrifugation at 14,000 ×g at 4 °C for 10 min.
137
SDS-PAGE analysis
138
Harvested cells were washed three times with PBS buffer (pH 7.5) and resuspended in the
139
same buffer, followed by ultrasonication at 4 °C (40% output, 10 min). After
140
centrifugation at 14,000 ×g at 4 °C for 10 min, the resulting supernatant was analyzed by
141
12% (v/v) SDS-PAGE. The gels were stained with Coomassie Brilliant Blue G-250.
142
Protein was estimated by the method of Bradford (Bradford, 1976) using a Bio-Rad
143
protein assay and bovine serum albumin as a standard.
144
Preparation of enzyme solutions
145
All procedures were carried out at 0–4 °C. Cells were washed with cold PBS buffer (pH
146
7.5). After centrifugation, cells were resuspended in 10–15 ml of the aforementioned
147
buffer solution. After ultrasonication, cells were centrifuged for 20 min to remove cell
148
debris and unbroken cells. The resulting supernatants were collected and used to measure
149
protein content and flavokinase activity. Total protein content was determined using an
150
Easy Protein Quantitative Kit (TransGen Biotech Inc., Beijing, China).
151
Enzyme activity assay of RibF
152
Flavokinase activity is expressed as nanomoles of FMN formed from riboflavin and ATP.
153
Flavokinase activity was measured in a final volume of 1 ml with 600 μl protein soluble
154
supernatants and 400 μl PBS buffer (pH 7.5) containing 50 mM riboflavin, 3 mM ATP,
155
15 mM MgCl2, and 10 mM Na2SO3. The mixture were pre-incubated at 37 °C for 5 min,
156
then the reaction was stopped by heating at 100 °C for 5 min (Nielsen, et al., 1983).
157
Proteins were removed from the substrates and product solution by centrifugation, and
158
supernatant compositions were analyzed by HPLC.
159
Quantitative real-time PCR (qRT-PCR) 7
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
133
160
The gapA gene (glyceraldehyde-3-phosphate dehydrogenase A), whose expression level
161
is relatively constant, was used as the control gene. After strains were cultivated in a
162
shake flask for 3 h, total cellular RNA was extracted using an RNeasy minikit (Tiangen).
163
cDNA was generated by reverse transcription PCR using random 6-mers and oligo dT
164
primers with a PrimeScript RT reagent kit (TaKaRa). qRT-PCR was carried out using the
165
LightCycler 480 Real-Time PCR system (Roche) with SYBR Premix ExTaq II (TaKaRa).
166
Measurement of each gene was repeated three times.
168
Results
169
Identification of riboflavin in fermentation broth of E. coli BL21(DE3)
170
During the comparative experiments of strains BL21(DE3) and MG1655, we noticed that
171
the color of the BL21 fermentation broth gradually turned yellow. This phenomenon
172
occurred from the late exponential phase of BL21, whereas it did not arise in MG1655
173
(see Fig. S1). The observed yellow substance emitted blue-violet fluorescence when
174
exposed to UV light (Fig. S2). Thus, we inferred that such obvious color change in the
175
BL21 culture medium was attributable to the accumulation of a fluorescent substance. To
176
identify the yellow substance in the fermentation broth, its excitation-emission spectra
177
were measured with a fluorescence spectrophotometer. The results showed that the BL21
178
fermentation broth had a strong absorption peak at 350–450 nm and strong emission peak
179
at 450–520 nm, which are proximate to the excitation and emission peaks of vitamin B2
180
or riboflavin. MG1655 did not have such obvious peaks in this wavelength range (Fig.
181
S3). Thus, we preliminarily inferred that the yellow substance in the BL21 fermentation
182
broth was riboflavin.
183
To verify this, the fermentation broths of both E. coli strains were collected and analyzed
184
with HPLC and LC-MS. A peak at 21.37 min was identified in BL21 but not MG1655.
185
This peak was subsequently proven to be riboflavin via LC-MS (Fig. 1). Upon diluting
186
the riboflavin to the same concentration as in the fermentation supernatant, it displayed a 8
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
167
similar light yellow color (Fig. S4). Therefore, we concluded that the fluorescent yellow
188
substance in the BL21 fermentation broth was riboflavin.
189
Analysis of relationship between riboflavin accumulation and cell physiology
190
changes
191
Growth behavior was compared for BL21 and MG1655. When cultivated in AM1
192
medium in the presence of 10 g/L glucose, both strains showed a similar glucose
193
consumption rate, but the growth rate of MG1655 was 1.5 times faster than that of BL21
194
(Table 2). MG1655 produced 1.2 g/L acetic acid after 14 h cultivation (Fig. 2a). Acetic
195
acid is the main byproduct of MG1655, which is consistent with the literature. However,
196
the acetic acid secretion rate of BL21 was 50% lower than that of MG1655 (Table 2).
197
Additionally, lactic acid was detected in the BL21 culture, which reached its maximum
198
yield of 1.7 g/L at 14 h (Fig. 2b). In BL21, riboflavin accumulated together with cell
199
growth (Fig. 2c). Even when cell growth entered the stationary phase, generation of
200
riboflavin continued at a certain rate. In contrast, only a small amount of riboflavin was
201
produced in the exponential phase of MG1655 (Fig. 2c).
202
In vivo, riboflavin is metabolized into FMN and FAD, which are known as the ligand of
203
pyruvate dehydrogenase, succinate dehydrogenase, and other important central metabolic
204
enzymes. To find if riboflavin accumulation also caused the accumulation of FAD, the
205
intracellular concentration of FAD in both MG1655 and BL21 was measured. The result
206
showed that the intracellular FAD concentration in BL21 was about 1.5 times that in
207
MG1655 (Table 3). FAD accumulation may enhance cell respiration, thereby affecting
208
metabolic flux.
209
Exploring possible reasons for riboflavin accumulation
210
To understand the physiological causes for riboflavin accumulation in BL21, we analyzed
211
the transcription of ribF and three genes involved in riboflavin biosynthesis of strains
212
BL21 and MG1655. Of these three, ribB encodes 3,4-dihydroxy-2-butanone-4-phosphate
213
synthase, ribE encodes riboflavin synthase β chain, and ribC encodes riboflavin synthase 9
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
187
α chain. These three gene products catalyze a three-step reaction of ribose-5-phosphate,
215
thereby yielding riboflavin. Our results showed that the expression level of these genes in
216
BL21 is approximately two-fold higher than that in MG1655 (Fig. 2d). The transcription
217
of ribF in BL21 was also higher than that in MG1655, and comparable to that of genes
218
involved in biosynthesis.
219
By comparing the protein sequences of the riboflavin synthesis pathways of BL21 and
220
MG1655, we discovered that BL21 underwent a mutation in RibF (Fig. S5). Histidine
221
115 (polar) mutated to leucine (neutral). Leucine 116 is a conserved amino acid among all
222
flavokinases in different bacteria and is important in catalysis. To compare the enzymatic
223
differences between them, ribF genes from BL21 and MG1655 were cloned into
224
expression vector pET28a. The expression of RibF was confirmed by SDS-PAGE (Fig. 3).
225
The enzymatic analysis indicated that the RibF enzyme specific activity of BL21 was
226
534.03 nmol h-1 mg-1 of protein which is only 55% of that in MG1655 (954.19 nmol h-1
227
mg-1 of protein). Together with the result from transcription analysis, we suggest that
228
riboflavin accumulation in BL21 occurred most probably because of an up-regulation of
229
FAD synthesis pathway and the reduced RibF enzyme activity.
230 231
Discussion
232
Compared with K-12, E. coli B strains show a number of phenotypes, such as faster
233
growth in minimal media, lower acetate production, and higher expression levels of
234
recombinant proteins. BL21(DE3) is a specifically engineered B descendant harboring
235
the T7 RNA polymerase gene (Studier & Moffatt, 1986), and has become the most widely
236
used strain for biotechnological applications. An increasing number of researchers prefer
237
BL21 as the metabolic engineering host for production of chemicals and other
238
biomolecules (Choi, et al., 2006).
239
Understanding the regulation mechanism and metabolic differences between B and K-12
240
strains will inform improved usage of these strains in metabolic engineering. 10
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
214
Transcriptome analysis of B and K-12 strains has proven that genes involved in the
242
glyoxylate shunt, TCA cycle, fatty acid, gluconeogenesis and anaplerotic pathways are
243
expressed differently (Phue, et al., 2007). In the present study, we found that BL21 could
244
accumulate more riboflavin than MG1655. The derivatives of riboflavin, FMN and FAD,
245
which participate in the delivery of hydrogen to the electron transport chain, are
246
important cofactors in cellular respiration. Therefore, they are vital in cell physiology,
247
although they exist in small amounts within cells. To find the reason why BL21
248
accumulated excessive riboflavin, the gene and protein sequences of the riboflavin
249
synthesis pathways in BL21 and MG1655 were compared, which showed that the
250
bifunctional enzyme RibF in BL21 underwent a His115Leu mutation. We suggest that the
251
inadequate enzyme activity of RibF contributed to the riboflavin accumulation. qPCR
252
analysis showed that riboflavin biosynthetic genes are all up-regulated. Up-regulation of
253
the riboflavin and FAD biosynthesis pathways increased riboflavin and FAD
254
accumulation. Although there is a possibility that the His115Leu mutation is the reason
255
for increased riboflavin, we are inclined to believe that there may be a mechanism
256
up-regulating all genes of the riboflavin pathway and causing accumulation of the two
257
downstream products.
258
Acid secretion is a severe problem in E. coli fermentation because cells can barely
259
survive at pH lower than 5.5 (Eiteman & Altman, 2006). Our study revealed that BL21
260
secretes less acetic acid, consistent with the literature. Acid overflow is caused by an
261
imbalance in the metabolic flux distribution at a certain metabolic node. In MG1655, the
262
accumulation of acetic acid is caused by an imbalance in an active glycolytic pathway
263
and a less efficient TCA cycle. In BL21, the TCA cycle appears more efficient, therefore
264
less acetic acid was accumulated (Phue, 2005). One theory that can be inferred is that
265
cells can increase flow through the TCA cycle by up-regulating the expression of
266
cofactors matched with enzymes in the TCA cycle, and thereby strengthen cellular
267
respiration. Therefore, the accumulation of riboflavin and FAD is likely to affect cell 11
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
241
respiration and be one of the reasons for the metabolic differences between these two
269
strains.
270
E. coli B strains and derivatives have salient features, such as low acetate secretion, faster
271
growth in minimal media, protease deficiency (saiSree, et al., 2001), and simple cell
272
surfaces (Herrera, et al., 2002) that enhance permeability. Therefore, these strains have
273
been used as industrial hosts for protein and biomolecule production. The additional
274
up-regulated riboflavin and FAD biosynthesis pathway in BL21 suggests that this strain
275
may also be used for riboflavin production.
276 277
Acknowledgments
278
This research was financially supported by grants from the National Natural Science
279
Foundation of China (31200033), National Basic Research Program of China
280
(2012CB725202, 2011CB707405). The authors do not have any conflicts of interest.
12
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
268
281
References
282
Bacher A, Eberhardt S, Fischer M, Kis K & Richter G (2000) Biosynthesis of vitamin B2 (riboflavin). Annu
283 284 285 286
Rev Nutr 20: 153-167.
287 288
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
1-49. Blattner FR, Plunkett G, 3rd, Bloch CA, et al. (1997) The complete genome sequence of Escherichia coli K-12. Science 277: 1453-1462. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254. Chen Q, Wang Q, Wei G, Liang Q & Qi Q (2011) Production in Escherichia coli of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) with differing monomer compositions from unrelated carbon sources. Appl Environ Microbiol 77: 4886-4893. Choi JH, Lee SJ & Lee SY (2003) Enhanced production of insulin-like growth factor I fusion protein in Escherichia coli by coexpression of the down-regulated genes identified by transcriptome profiling. Appl Environ Microbiol 69: 4737-4742. Choi JH, Keum KC & Lee SY (2006) Production of recombinant proteins by high cell density culture of Escherichia coli Chemical Engineering Science 61: 876-885. Covert MW, Knight EM, Reed JL, Herrgard MJ & Palsson BO (2004) Integrating high-throughput and computational data elucidates bacterial networks. Nature 429: 92-96. Depeint F, Bruce WR, Shangari N, Mehta R & O'Brien PJ (2006) Mitochondrial function and toxicity: role of the B vitamin family on mitochondrial energy metabolism. Chem Biol Interact 163: 94-112. Eiteman MA & Altman E (2006) Overcoming acetate in Escherichia coli recombinant protein fermentations. Trends Biotechnol 24: 530-536. Feist AM, Henry CS, Reed JL, et al. (2007) A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for 1260 ORFs and thermodynamic information. Mol Syst Biol 3: 121. Franchini AG & Egli T (2006) Global gene expression in Escherichia coli K-12 during short-term and long-term adaptation to glucose-limited continuous culture conditions. Microbiology 152: 2111-2127. Han MJ & Lee SY (2006) The Escherichia coli proteome: past, present, and future prospects. Microbiol Mol Biol Rev 70: 362-439. Hayashi K, Morooka N, Yamamoto Y, et al. (2006) Highly accurate genome sequences of Escherichia coli K-12 strains MG1655 and W3110. Mol Syst Biol 2: 2006 0007. Herrera G, Martinez A, Blanco M & O'Connor JE (2002) Assessment of Escherichia coli B with enhanced permeability to fluorochromes for flow cytometric assays of bacterial cell function. Cytometry 49: 62-69. Ishii N, Nakahigashi K, Baba T, et al. (2007) Multiple high-throughput analyses monitor the response of E. coli to perturbations. Science 316: 593-597. Lin Z, Xu Z, Li Y, Wang Z, Chen T & Zhao X (2014) Metabolic engineering of Escherichia coli for the production of riboflavin. Microb Cell Fact 13: 104. Mack M, van Loon AP & Hohmann HP (1998) Regulation of riboflavin biosynthesis in Bacillus subtilis is affected by the activity of the flavokinase/flavin adenine dinucleotide synthetase encoded by ribC. J Bacteriol 180: 950-955. Martinez A, Grabar TB, Shanmugam KT, Yomano LP, York SW & Ingram LO (2007) Low salt medium for 13
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
289 290 291
Bacher A, Eberhardt S, Eisenreich W, et al. (2001) Biosynthesis of riboflavin. Vitamins & Hormones 61:
322 323 324 325 326
lactate and ethanol production by recombinant Escherichia coli B. Biotechnol Lett 29: 397-404. Massey V (2000) The chemical and biological versatility of riboflavin. Biochem Soc Trans 28: 283-296. Nandakumar MP, Cheung A & Marten MR (2006) Proteomic analysis of extracellular proteins from Escherichia coli W3110. J Proteome Res 5: 1155-1161. Nielsen P, Rauschenbach P & Bacher A (1983) Phosphates of riboflavin and riboflavin analogs: a
327 328
reinvestigation by high-performance liquid chromatography. Anal Biochem 130: 359-368.
329 330 331 332
application to the central carbon metabolism of Escherichia coli BL21 and JM109. Genomics 89: 300-305. Phue JN, S. B. Noronha, R. Hattacharyya, A. J. Wolfe, and J. Shiloach. (2005) Glucose metabolism at high density growth of E. coli B and E. coli K: differences in metabolic pathways are responsible for efficient glucose utilization in E. coli B as determined by microarrays and Northern blot analyses. Biotechnol. Bioeng. 90: 805-820. Richter G, Volk R, Krieger C, Lahm HW, Rothlisberger U & Bacher A (1992) Biosynthesis of riboflavin: cloning, sequencing, and expression of the gene coding for 3,4-dihydroxy-2-butanone 4-phosphate synthase of Escherichia coli. J Bacteriol 174: 4050-4056. Richter G, Fischer M, Krieger C, Eberhardt S, Luttgen H, Gerstenschlager I & Bacher A (1997) Biosynthesis of riboflavin: characterization of the bifunctional deaminase-reductase of Escherichia coli and Bacillus subtilis. J Bacteriol 179: 2022-2028. saiSree L, Reddy M & Gowrishankar J (2001) IS186 insertion at a hot spot in the lon promoter as a basis for lon protease deficiency of Escherichia coli B: identification of a consensus target sequence for IS186 transposition. J Bacteriol 183: 6943-6946. Sauer U, Lasko DR, Fiaux J, et al. (1999) Metabolic flux ratio analysis of genetic and environmental modulations of Escherichia coli central carbon metabolism. J Bacteriol 181: 6679-6688. Schneider D, Duperchy E, Depeyrot J, Coursange E, Lenski R & Blot M (2002) Genomic comparisons among Escherichia coli strains B, K-12, and O157:H7 using IS elements as molecular markers. BMC Microbiol 2: 18. Stahmann KP, Revuelta JL & Seulberger H (2000) Three biotechnical processes using Ashbya gossypii, Candida famata, or Bacillus subtilis compete with chemical riboflavin production. Appl Microbiol Biotechnol 53: 509-516. Studier FW & Moffatt BA (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189: 113-130. Swartz JR (1996) Escherichia coli recombinant DNA technology. Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, DC 1693-1711. Vitreschak AG, Rodionov DA, Mironov AA & Gelfand MS (2002) Regulation of riboflavin biosynthesis and transport genes in bacteria by transcriptional and translational attenuation. Nucleic Acids Res 30: 3141-3151. Waegeman H, Beauprez J, Moens H, et al. (2011) Effect of iclR and arcA knockouts on biomass formation and metabolic fluxes in Escherichia coli K12 and its implications on understanding the metabolism of Escherichia coli BL21 (DE3). BMC Microbiol 11: 70. Yoon SH, Han MJ, Jeong H, et al. (2012) Comparative multi-omics systems analysis of Escherichia coli strains B and K-12. Genome Biol 13: R37. 14
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362
Phue JN, Kedem B, Jaluria P & Shiloach J (2007) Evaluating microarrays using a semiparametric approach:
363
364
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
15
364
Table 1 Strains and plasmids used in this study Strains and plasmids
Relevant characteristics
Sources
E．coli K-12 MG1655
F-,λ-, ilvG-, rfb-50, rph-1
CGSC no: 6300
E．coli B BL21(DE3)
F– ompT gal dcm lon hsdSB(rB- mB-) λ(DE3 [lacI TransGen lacUV5-T7 gene 1 ind1 sam7 nin5]) pMB1 origin, protein expression vector
Laboratory stock
pRibbhis
pET28a containing ribF from BL21
This study
pRibmhis
pET28a containing ribF from MG1655
This study
365 366
16
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
pET28a
366
Table 2 Comparison of the physiological parameters between BL21 and MG1655 Growth rate
Glucose uptake
Lactate secretion
Acetate secretion
(h-1)
(mmol g-1 h-1)
(mmol g-1 h-1)
(mmol g-1 h-1)
BL21
0.48±0.01
9.14±0.92
0.84±0.03
2.85±0.13
MG1655
0.67±0.01
9.68±0.79
0
6.11±0.10
367 Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
368 369
17
369
Table 3 Intracellular FAD levels of MG1655 and BL21
370 FAD concentration (pmol/g cells) E. coli BL21
E. coli MG1655
Mid exponential phase
0.081±0.003
0.053±0.002
Late stationary phase
0.053±0.001
0.014±0.004 Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
371
18
Legends to Figures:
371 372
Fig. 1
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
373 374 375
Fig. 1 Identification of riboflavin by HPLC, (a) Riboflavin standard sample and (b)
376
fermentation broth of BL21 (the arrows point at the peak of riboflavin); (c) Mass spectra
377
of the Riboflavin standard sample, (d) Mass spectra of the identified Riboflavin peak in
378
BL21 fermentation broth.
379
19
379
Fig. 2 (b)
(a)
(d)
3.0 MG1655 BL21
Relative expression level
2.5 2.0 1.5 1.0 0.5 0.0
ribB
ribE
ribC
ribF
380 381
Fig. 2 Growth, glucose consumption and acetate and lactate secretion in (a) MG1655 and
382
(b) BL21 in AM1 medium; (c) riboflavin secretion in MG1655 and BL21 in AM1
383
medium; (d) comparison of associated gene expression of the riboflavin synthesis
384
pathway between MG1655 and BL21.
385 386
20
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
(c)
386
Fig. 3
66.2 kDa 45.0
35.0
RibF (34.7 kDa)
387
1
2
3
388 389
Fig. 3 The expression of RibFs. Lane 1, protein ladder; Lane 2, RibF expression from
390
MG1655; lane 3, RibF expression from BL21.
21
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
25.0
1
Identification of riboflavin: revealing different metabolic
2
characteristics between Escherichia coli BL21(DE3) and
3
MG1655
4 5
Running title: Riboflavin accumulation in Escherichia coli BL21(DE3)
6
Xinran Wang, Qian Wang*, Qingsheng Qi
8
National Glycoengineering Research Center, State Key Laboratory of Microbial
9
Technology,
10
Shandong University, Jinan 250100, P. R. China
11 12
* Corresponding author: Tel & Fax: +86-531-88365628;
13
E-mail address: [email protected]
14 15
1
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
7
15
Abstract
17
There are many physiological differences between Escherichia coli B and K-12 strains,
18
owing to their different origins. Deeper insight into the metabolic and regulative
19
mechanisms of these strains will inform improved usage of these industrial workhorses.
20
In the present study, we observed that BL21 fermentation broth gradually turned yellow
21
during cultivation. By spectral analysis and liquid chromatography-mass spectroscopy
22
identification, we confirmed for the first time that the yellow substance accumulated in
23
the fermentation broth is riboflavin. Comparing the enzyme sequences involved in
24
riboflavin metabolism between BL21 and MG1655, we identified a site mutation on the
25
115 residue of bifunctional riboflavin kinase/FMN adenylyltransferase (RibF) in BL21.
26
This His115Leu mutation was found to reduce enzyme activity to 55% that of MG1655,
27
which is probably one reason for riboflavin accumulation in BL21. Quantitative PCR
28
analysis showed that genes of the entire branch of the riboflavin and FAD biosynthesis
29
pathways in BL21 were up-regulated. Several physiological and metabolic characteristics
30
of BL21 and MG1655 were found to be different, and may also be related to the
31
riboflavin accumulation.
32
2
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
16
32
Introduction
34
Escherichia coli has been a workhorse, not only in fundamental biological studies but
35
also in various biotechnological applications. Almost all laboratory strains of E. coli are
36
derivatives of nonpathogenic K-12 or B strains. Since two K-12 strains - MG1655 and
37
W3110 are sequenced in 1997 (Blattner, et al., 1997, Hayashi, et al., 2006), Increasing
38
omics analyses (Choi, et al., 2003, Franchini & Egli, 2006, Han & Lee, 2006,
39
Nandakumar, et al., 2006, Ishii, et al., 2007) and in silico metabolic modeling (Covert, et
40
al., 2004, Feist, et al., 2007) have been performed in K-12. In contrast, there have been
41
few studies for B strains, although they have been widely used for large-scale production
42
of recombinant proteins, ethanol, and other biomolecules (Choi, et al., 2006). Recently,
43
the genome sequences of two B strains, REL606 and BL21(DE3), have been determined.
44
Comparative omics analysis of B and K-12 show that they are very similar (Schneider, et
45
al., 2002) and closely related (Yoon, et al., 2012), although the strains show many
46
important behavioral differences in cell metabolism and physiology (Swartz, 1996). The
47
B strains have a lower acetate production than K-12 derivatives. A common explanation
48
for that is the activated glyoxylate shunt, which is the main pathway for acetate utilization.
49
Deletions of arcA and iclR in E. coli MG1655 endowed the host with central metabolic
50
fluxes similar to that of E. coli BL21(DE3) under glucose abundant conditions
51
(Waegeman, et al., 2011).
52
Riboflavin (vitamin B2), the precursor of coenzymes flavin adenine dinucleotide (FAD)
53
and flavin mononucleotide (FMN), is an essential component of basic metabolism
54
(Massey, 2000). Riboflavin is required for a wide variety of cellular processes and has a
55
key role in the metabolism of fats, ketone bodies, carbohydrates, and proteins (Depeint, et
56
al., 2006). Many microorganisms including fungi, and also plants, synthesize riboflavin,
57
while humans and animals need external riboflavin supply to avoid symptoms of
58
deficiency (Stahmann, et al., 2000). 3
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
33
Riboflavin biosynthesis has been studied in both gram-positive and gram-negative
60
bacteria, such as Bacillus subtilis and E. coli (Richter, et al., 1992, Richter, et al., 1997).
61
Riboflavin is formed from one GTP molecule and two molecules of ribulose 5-phosphate
62
(Ru-5-P), through a seven-step pathway (Bacher, et al., 2000). In B. subtilis, the encoding
63
genes of riboflavin biosynthesis reside within one rib operon (Vitreschak, et al., 2002). In
64
contrast, these genes in E. coli do not converge to a single operon but are scattered on the
65
chromosome (Bacher, et al., 2001). It has been reported that reducing the enzymatic
66
activity of bifunctional riboflavin kinase/FMN adenylyltransferase (encoded by the ribC
67
gene) could result in riboflavin overproduction in B. subtilis (Mack, et al., 1998). In E.
68
coli, this enzyme is encoded by ribF, which is an essential gene for survival. By
69
modulating the expression of ribF, riboflavin production was improved in E. coli (Lin, et
70
al., 2014). However, the exact regulation mechanism of riboflavin in E. coli has not been
71
revealed.
72
In the present study, we showed that E. coli BL21 was able to generate riboflavin under
73
normal cultivation conditions. Comparative analysis of E. coli strains BL21 and MG1655
74
indicated that riboflavin accumulation may be a reason for their metabolic difference. By
75
transcription and enzymatic analysis, we found that the riboflavin accumulation in BL21
76
is caused by not only a single mutation of ribF but also a certain regulation mechanism
77
which up-regulates the whole FAD synthesis pathway.
78 79
4
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
59
79
Material and methods
81
Bacterial strains and plasmids
82
The strains of E. coli and plasmids used in the study are summarized in Table 1. E. coli
83
DH5α was used as the host for molecular cloning. E. coli BL21(DE3) was used as the
84
host for protein expression. Plasmid pET28a was used as the expression vector for the
85
enzyme activity assay of different RibFs. Genomic DNA was isolated from wild-type E.
86
coli MG 1655 and BL21(DE3) with a TIANamp Bacterial DNA kit (Tiangen Biotech).
87
The two ribF genes were respectively amplified from the genome DNA of E. coli MG
88
1655
primer
pair
PribF
FP
89
(GCTTCCATGGCTATGAAGCTGATACGCGGCATACA)
and
PribF
RP
90
(ACTCCTCGAGAGCCGGTTTTGTTAGCCC). To construct the expression plasmids,
91
PCR fragments were digested with the restriction enzymes NcoI and XhoI and then
92
ligated into the corresponding sites of pET28a, which was cut with the same restriction
93
enzymes to generate plasmids pRibmhis (ribF from MG1655) and pRibbhis (ribF from
94
BL21(DE3)).
95
Culture conditions
96
LB medium (10 g/L tryptone, 5 g/L yeast extract and 10 g/L NaCl) was used for all DNA
97
manipulations and protein expression. During cultivation and fermentation, AM1 mineral
98
salts medium (Martinez, et al., 2007) supplemented with 10 g/L glucose was used.
99
Kanamycin (25 mg/ml) was added to provide selective pressure during cultivation when
and
BL21(DE3)
To
induce
using
the
100
necessary.
expression
101
isopropyl-β-D-thiogalactopyranoside (IPTG) was added to cultures, with a final
102
concentration of 0.2 mM.
103
Analysis methods during fermentation
104
Cell growth was detected by measuring optical density at 600 nm with a
105
spectrophotometer (Thermo Scientific Instruments LLC, USA). To detect the yellow 5
of
plasmid-borne
genes,
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
80
fluorescent substance in the fermentation broth, 2 μl of culture supernatant was spotted
107
onto filter papers and exposed to ultraviolet light, to check for the fluorescence. The
108
excitation-emission spectra were measured with a fluorescence spectrophotometer
109
(RF-5301PC, Shimadzu Corp., Japan).
110
Glucose and organic acids were determined by high-performance liquid chromatography
111
(HPLC) (Shimadzu) coupled with a refractive index detector (RID-10A), as reported
112
previously (Chen, et al., 2011). Samples were filtered through a 0.22 μm syringe filter
113
and analyzed on an HPX-87H column (Bio-Rad, USA) at 65 °C with 5 mM H2SO4 as
114
mobile phase and flow rate 0.6 ml/min. Physiological parameters of specific growth rate,
115
specific glucose uptake rate, and specific product secretion rates were determined for the
116
exponentially growing cells, according to methods described by (Sauer, et al., 1999).
117
Intracellular FAD concentration was measured using an FAD Colorimetric/Fluorometric
118
Assay Kit (Cat# K357-100; BioVision Inc., USA).
119
HPLC and LC-MS analysis of riboflavin
120
Supernatants were subsequently filtered through a 0.22-μm syringe filter, and a 20 μl
121
aliquot of the filtered solution was injected into an HPLC system (Shimadzu) equipped
122
with a UV-VIS detector (SPD-10A; excitation 470 nm; emission 530 nm). Riboflavin
123
was analyzed on an Agilent ZORBAX SB-C18 column (4.6 mm ID × 250 mm) with 30%
124
(v/v) methanol-0.1% (v/v) acetic acid as a mobile phase and flow rate 0.5 ml/min.
125
Riboflavin was qualitatively analyzed by liquid chromatography-mass spectroscopy
126
(LC-MS) using a TurboMass Gold mass spectrometer (PerkinElmer Inc., USA).
127
Riboflavin was identified by LC-MS based on retention time and fragmentation patterns
128
of known standards.
129
Protein expression of RibF
130
E. coli BL21(DE3) was transformed with plasmids pRibbhis and pRibmhis, and the
131
resulting strains were aerobically cultivated overnight at 37 °C in LB medium. For shake
132
flask culture, 2 ml of the overnight culture was added to a 300 ml Erlenmeyer flask 6
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
106
containing 50 ml LB medium at 37 °C. Gene expression was stimulated by adding 0.2
134
mM IPTG after the culture had reached an optical density at 600 nm of 0.8–1.0. Flasks
135
were transferred to 16 °C for soluble protein expression. After 20 h aerobic cultivation,
136
cells were harvested by centrifugation at 14,000 ×g at 4 °C for 10 min.
137
SDS-PAGE analysis
138
Harvested cells were washed three times with PBS buffer (pH 7.5) and resuspended in the
139
same buffer, followed by ultrasonication at 4 °C (40% output, 10 min). After
140
centrifugation at 14,000 ×g at 4 °C for 10 min, the resulting supernatant was analyzed by
141
12% (v/v) SDS-PAGE. The gels were stained with Coomassie Brilliant Blue G-250.
142
Protein was estimated by the method of Bradford (Bradford, 1976) using a Bio-Rad
143
protein assay and bovine serum albumin as a standard.
144
Preparation of enzyme solutions
145
All procedures were carried out at 0–4 °C. Cells were washed with cold PBS buffer (pH
146
7.5). After centrifugation, cells were resuspended in 10–15 ml of the aforementioned
147
buffer solution. After ultrasonication, cells were centrifuged for 20 min to remove cell
148
debris and unbroken cells. The resulting supernatants were collected and used to measure
149
protein content and flavokinase activity. Total protein content was determined using an
150
Easy Protein Quantitative Kit (TransGen Biotech Inc., Beijing, China).
151
Enzyme activity assay of RibF
152
Flavokinase activity is expressed as nanomoles of FMN formed from riboflavin and ATP.
153
Flavokinase activity was measured in a final volume of 1 ml with 600 μl protein soluble
154
supernatants and 400 μl PBS buffer (pH 7.5) containing 50 mM riboflavin, 3 mM ATP,
155
15 mM MgCl2, and 10 mM Na2SO3. The mixture were pre-incubated at 37 °C for 5 min,
156
then the reaction was stopped by heating at 100 °C for 5 min (Nielsen, et al., 1983).
157
Proteins were removed from the substrates and product solution by centrifugation, and
158
supernatant compositions were analyzed by HPLC.
159
Quantitative real-time PCR (qRT-PCR) 7
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
133
160
The gapA gene (glyceraldehyde-3-phosphate dehydrogenase A), whose expression level
161
is relatively constant, was used as the control gene. After strains were cultivated in a
162
shake flask for 3 h, total cellular RNA was extracted using an RNeasy minikit (Tiangen).
163
cDNA was generated by reverse transcription PCR using random 6-mers and oligo dT
164
primers with a PrimeScript RT reagent kit (TaKaRa). qRT-PCR was carried out using the
165
LightCycler 480 Real-Time PCR system (Roche) with SYBR Premix ExTaq II (TaKaRa).
166
Measurement of each gene was repeated three times.
168
Results
169
Identification of riboflavin in fermentation broth of E. coli BL21(DE3)
170
During the comparative experiments of strains BL21(DE3) and MG1655, we noticed that
171
the color of the BL21 fermentation broth gradually turned yellow. This phenomenon
172
occurred from the late exponential phase of BL21, whereas it did not arise in MG1655
173
(see Fig. S1). The observed yellow substance emitted blue-violet fluorescence when
174
exposed to UV light (Fig. S2). Thus, we inferred that such obvious color change in the
175
BL21 culture medium was attributable to the accumulation of a fluorescent substance. To
176
identify the yellow substance in the fermentation broth, its excitation-emission spectra
177
were measured with a fluorescence spectrophotometer. The results showed that the BL21
178
fermentation broth had a strong absorption peak at 350–450 nm and strong emission peak
179
at 450–520 nm, which are proximate to the excitation and emission peaks of vitamin B2
180
or riboflavin. MG1655 did not have such obvious peaks in this wavelength range (Fig.
181
S3). Thus, we preliminarily inferred that the yellow substance in the BL21 fermentation
182
broth was riboflavin.
183
To verify this, the fermentation broths of both E. coli strains were collected and analyzed
184
with HPLC and LC-MS. A peak at 21.37 min was identified in BL21 but not MG1655.
185
This peak was subsequently proven to be riboflavin via LC-MS (Fig. 1). Upon diluting
186
the riboflavin to the same concentration as in the fermentation supernatant, it displayed a 8
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
167
similar light yellow color (Fig. S4). Therefore, we concluded that the fluorescent yellow
188
substance in the BL21 fermentation broth was riboflavin.
189
Analysis of relationship between riboflavin accumulation and cell physiology
190
changes
191
Growth behavior was compared for BL21 and MG1655. When cultivated in AM1
192
medium in the presence of 10 g/L glucose, both strains showed a similar glucose
193
consumption rate, but the growth rate of MG1655 was 1.5 times faster than that of BL21
194
(Table 2). MG1655 produced 1.2 g/L acetic acid after 14 h cultivation (Fig. 2a). Acetic
195
acid is the main byproduct of MG1655, which is consistent with the literature. However,
196
the acetic acid secretion rate of BL21 was 50% lower than that of MG1655 (Table 2).
197
Additionally, lactic acid was detected in the BL21 culture, which reached its maximum
198
yield of 1.7 g/L at 14 h (Fig. 2b). In BL21, riboflavin accumulated together with cell
199
growth (Fig. 2c). Even when cell growth entered the stationary phase, generation of
200
riboflavin continued at a certain rate. In contrast, only a small amount of riboflavin was
201
produced in the exponential phase of MG1655 (Fig. 2c).
202
In vivo, riboflavin is metabolized into FMN and FAD, which are known as the ligand of
203
pyruvate dehydrogenase, succinate dehydrogenase, and other important central metabolic
204
enzymes. To find if riboflavin accumulation also caused the accumulation of FAD, the
205
intracellular concentration of FAD in both MG1655 and BL21 was measured. The result
206
showed that the intracellular FAD concentration in BL21 was about 1.5 times that in
207
MG1655 (Table 3). FAD accumulation may enhance cell respiration, thereby affecting
208
metabolic flux.
209
Exploring possible reasons for riboflavin accumulation
210
To understand the physiological causes for riboflavin accumulation in BL21, we analyzed
211
the transcription of ribF and three genes involved in riboflavin biosynthesis of strains
212
BL21 and MG1655. Of these three, ribB encodes 3,4-dihydroxy-2-butanone-4-phosphate
213
synthase, ribE encodes riboflavin synthase β chain, and ribC encodes riboflavin synthase 9
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
187
α chain. These three gene products catalyze a three-step reaction of ribose-5-phosphate,
215
thereby yielding riboflavin. Our results showed that the expression level of these genes in
216
BL21 is approximately two-fold higher than that in MG1655 (Fig. 2d). The transcription
217
of ribF in BL21 was also higher than that in MG1655, and comparable to that of genes
218
involved in biosynthesis.
219
By comparing the protein sequences of the riboflavin synthesis pathways of BL21 and
220
MG1655, we discovered that BL21 underwent a mutation in RibF (Fig. S5). Histidine
221
115 (polar) mutated to leucine (neutral). Leucine 116 is a conserved amino acid among all
222
flavokinases in different bacteria and is important in catalysis. To compare the enzymatic
223
differences between them, ribF genes from BL21 and MG1655 were cloned into
224
expression vector pET28a. The expression of RibF was confirmed by SDS-PAGE (Fig. 3).
225
The enzymatic analysis indicated that the RibF enzyme specific activity of BL21 was
226
534.03 nmol h-1 mg-1 of protein which is only 55% of that in MG1655 (954.19 nmol h-1
227
mg-1 of protein). Together with the result from transcription analysis, we suggest that
228
riboflavin accumulation in BL21 occurred most probably because of an up-regulation of
229
FAD synthesis pathway and the reduced RibF enzyme activity.
230 231
Discussion
232
Compared with K-12, E. coli B strains show a number of phenotypes, such as faster
233
growth in minimal media, lower acetate production, and higher expression levels of
234
recombinant proteins. BL21(DE3) is a specifically engineered B descendant harboring
235
the T7 RNA polymerase gene (Studier & Moffatt, 1986), and has become the most widely
236
used strain for biotechnological applications. An increasing number of researchers prefer
237
BL21 as the metabolic engineering host for production of chemicals and other
238
biomolecules (Choi, et al., 2006).
239
Understanding the regulation mechanism and metabolic differences between B and K-12
240
strains will inform improved usage of these strains in metabolic engineering. 10
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
214
Transcriptome analysis of B and K-12 strains has proven that genes involved in the
242
glyoxylate shunt, TCA cycle, fatty acid, gluconeogenesis and anaplerotic pathways are
243
expressed differently (Phue, et al., 2007). In the present study, we found that BL21 could
244
accumulate more riboflavin than MG1655. The derivatives of riboflavin, FMN and FAD,
245
which participate in the delivery of hydrogen to the electron transport chain, are
246
important cofactors in cellular respiration. Therefore, they are vital in cell physiology,
247
although they exist in small amounts within cells. To find the reason why BL21
248
accumulated excessive riboflavin, the gene and protein sequences of the riboflavin
249
synthesis pathways in BL21 and MG1655 were compared, which showed that the
250
bifunctional enzyme RibF in BL21 underwent a His115Leu mutation. We suggest that the
251
inadequate enzyme activity of RibF contributed to the riboflavin accumulation. qPCR
252
analysis showed that riboflavin biosynthetic genes are all up-regulated. Up-regulation of
253
the riboflavin and FAD biosynthesis pathways increased riboflavin and FAD
254
accumulation. Although there is a possibility that the His115Leu mutation is the reason
255
for increased riboflavin, we are inclined to believe that there may be a mechanism
256
up-regulating all genes of the riboflavin pathway and causing accumulation of the two
257
downstream products.
258
Acid secretion is a severe problem in E. coli fermentation because cells can barely
259
survive at pH lower than 5.5 (Eiteman & Altman, 2006). Our study revealed that BL21
260
secretes less acetic acid, consistent with the literature. Acid overflow is caused by an
261
imbalance in the metabolic flux distribution at a certain metabolic node. In MG1655, the
262
accumulation of acetic acid is caused by an imbalance in an active glycolytic pathway
263
and a less efficient TCA cycle. In BL21, the TCA cycle appears more efficient, therefore
264
less acetic acid was accumulated (Phue, 2005). One theory that can be inferred is that
265
cells can increase flow through the TCA cycle by up-regulating the expression of
266
cofactors matched with enzymes in the TCA cycle, and thereby strengthen cellular
267
respiration. Therefore, the accumulation of riboflavin and FAD is likely to affect cell 11
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
241
respiration and be one of the reasons for the metabolic differences between these two
269
strains.
270
E. coli B strains and derivatives have salient features, such as low acetate secretion, faster
271
growth in minimal media, protease deficiency (saiSree, et al., 2001), and simple cell
272
surfaces (Herrera, et al., 2002) that enhance permeability. Therefore, these strains have
273
been used as industrial hosts for protein and biomolecule production. The additional
274
up-regulated riboflavin and FAD biosynthesis pathway in BL21 suggests that this strain
275
may also be used for riboflavin production.
276 277
Acknowledgments
278
This research was financially supported by grants from the National Natural Science
279
Foundation of China (31200033), National Basic Research Program of China
280
(2012CB725202, 2011CB707405). The authors do not have any conflicts of interest.
12
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
268
281
References
282
Bacher A, Eberhardt S, Fischer M, Kis K & Richter G (2000) Biosynthesis of vitamin B2 (riboflavin). Annu
283 284 285 286
Rev Nutr 20: 153-167.
287 288
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
1-49. Blattner FR, Plunkett G, 3rd, Bloch CA, et al. (1997) The complete genome sequence of Escherichia coli K-12. Science 277: 1453-1462. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254. Chen Q, Wang Q, Wei G, Liang Q & Qi Q (2011) Production in Escherichia coli of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) with differing monomer compositions from unrelated carbon sources. Appl Environ Microbiol 77: 4886-4893. Choi JH, Lee SJ & Lee SY (2003) Enhanced production of insulin-like growth factor I fusion protein in Escherichia coli by coexpression of the down-regulated genes identified by transcriptome profiling. Appl Environ Microbiol 69: 4737-4742. Choi JH, Keum KC & Lee SY (2006) Production of recombinant proteins by high cell density culture of Escherichia coli Chemical Engineering Science 61: 876-885. Covert MW, Knight EM, Reed JL, Herrgard MJ & Palsson BO (2004) Integrating high-throughput and computational data elucidates bacterial networks. Nature 429: 92-96. Depeint F, Bruce WR, Shangari N, Mehta R & O'Brien PJ (2006) Mitochondrial function and toxicity: role of the B vitamin family on mitochondrial energy metabolism. Chem Biol Interact 163: 94-112. Eiteman MA & Altman E (2006) Overcoming acetate in Escherichia coli recombinant protein fermentations. Trends Biotechnol 24: 530-536. Feist AM, Henry CS, Reed JL, et al. (2007) A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for 1260 ORFs and thermodynamic information. Mol Syst Biol 3: 121. Franchini AG & Egli T (2006) Global gene expression in Escherichia coli K-12 during short-term and long-term adaptation to glucose-limited continuous culture conditions. Microbiology 152: 2111-2127. Han MJ & Lee SY (2006) The Escherichia coli proteome: past, present, and future prospects. Microbiol Mol Biol Rev 70: 362-439. Hayashi K, Morooka N, Yamamoto Y, et al. (2006) Highly accurate genome sequences of Escherichia coli K-12 strains MG1655 and W3110. Mol Syst Biol 2: 2006 0007. Herrera G, Martinez A, Blanco M & O'Connor JE (2002) Assessment of Escherichia coli B with enhanced permeability to fluorochromes for flow cytometric assays of bacterial cell function. Cytometry 49: 62-69. Ishii N, Nakahigashi K, Baba T, et al. (2007) Multiple high-throughput analyses monitor the response of E. coli to perturbations. Science 316: 593-597. Lin Z, Xu Z, Li Y, Wang Z, Chen T & Zhao X (2014) Metabolic engineering of Escherichia coli for the production of riboflavin. Microb Cell Fact 13: 104. Mack M, van Loon AP & Hohmann HP (1998) Regulation of riboflavin biosynthesis in Bacillus subtilis is affected by the activity of the flavokinase/flavin adenine dinucleotide synthetase encoded by ribC. J Bacteriol 180: 950-955. Martinez A, Grabar TB, Shanmugam KT, Yomano LP, York SW & Ingram LO (2007) Low salt medium for 13
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
289 290 291
Bacher A, Eberhardt S, Eisenreich W, et al. (2001) Biosynthesis of riboflavin. Vitamins & Hormones 61:
322 323 324 325 326
lactate and ethanol production by recombinant Escherichia coli B. Biotechnol Lett 29: 397-404. Massey V (2000) The chemical and biological versatility of riboflavin. Biochem Soc Trans 28: 283-296. Nandakumar MP, Cheung A & Marten MR (2006) Proteomic analysis of extracellular proteins from Escherichia coli W3110. J Proteome Res 5: 1155-1161. Nielsen P, Rauschenbach P & Bacher A (1983) Phosphates of riboflavin and riboflavin analogs: a
327 328
reinvestigation by high-performance liquid chromatography. Anal Biochem 130: 359-368.
329 330 331 332
application to the central carbon metabolism of Escherichia coli BL21 and JM109. Genomics 89: 300-305. Phue JN, S. B. Noronha, R. Hattacharyya, A. J. Wolfe, and J. Shiloach. (2005) Glucose metabolism at high density growth of E. coli B and E. coli K: differences in metabolic pathways are responsible for efficient glucose utilization in E. coli B as determined by microarrays and Northern blot analyses. Biotechnol. Bioeng. 90: 805-820. Richter G, Volk R, Krieger C, Lahm HW, Rothlisberger U & Bacher A (1992) Biosynthesis of riboflavin: cloning, sequencing, and expression of the gene coding for 3,4-dihydroxy-2-butanone 4-phosphate synthase of Escherichia coli. J Bacteriol 174: 4050-4056. Richter G, Fischer M, Krieger C, Eberhardt S, Luttgen H, Gerstenschlager I & Bacher A (1997) Biosynthesis of riboflavin: characterization of the bifunctional deaminase-reductase of Escherichia coli and Bacillus subtilis. J Bacteriol 179: 2022-2028. saiSree L, Reddy M & Gowrishankar J (2001) IS186 insertion at a hot spot in the lon promoter as a basis for lon protease deficiency of Escherichia coli B: identification of a consensus target sequence for IS186 transposition. J Bacteriol 183: 6943-6946. Sauer U, Lasko DR, Fiaux J, et al. (1999) Metabolic flux ratio analysis of genetic and environmental modulations of Escherichia coli central carbon metabolism. J Bacteriol 181: 6679-6688. Schneider D, Duperchy E, Depeyrot J, Coursange E, Lenski R & Blot M (2002) Genomic comparisons among Escherichia coli strains B, K-12, and O157:H7 using IS elements as molecular markers. BMC Microbiol 2: 18. Stahmann KP, Revuelta JL & Seulberger H (2000) Three biotechnical processes using Ashbya gossypii, Candida famata, or Bacillus subtilis compete with chemical riboflavin production. Appl Microbiol Biotechnol 53: 509-516. Studier FW & Moffatt BA (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189: 113-130. Swartz JR (1996) Escherichia coli recombinant DNA technology. Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, DC 1693-1711. Vitreschak AG, Rodionov DA, Mironov AA & Gelfand MS (2002) Regulation of riboflavin biosynthesis and transport genes in bacteria by transcriptional and translational attenuation. Nucleic Acids Res 30: 3141-3151. Waegeman H, Beauprez J, Moens H, et al. (2011) Effect of iclR and arcA knockouts on biomass formation and metabolic fluxes in Escherichia coli K12 and its implications on understanding the metabolism of Escherichia coli BL21 (DE3). BMC Microbiol 11: 70. Yoon SH, Han MJ, Jeong H, et al. (2012) Comparative multi-omics systems analysis of Escherichia coli strains B and K-12. Genome Biol 13: R37. 14
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362
Phue JN, Kedem B, Jaluria P & Shiloach J (2007) Evaluating microarrays using a semiparametric approach:
363
364
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
15
364
Table 1 Strains and plasmids used in this study Strains and plasmids
Relevant characteristics
Sources
E．coli K-12 MG1655
F-,λ-, ilvG-, rfb-50, rph-1
CGSC no: 6300
E．coli B BL21(DE3)
F– ompT gal dcm lon hsdSB(rB- mB-) λ(DE3 [lacI TransGen lacUV5-T7 gene 1 ind1 sam7 nin5]) pMB1 origin, protein expression vector
Laboratory stock
pRibbhis
pET28a containing ribF from BL21
This study
pRibmhis
pET28a containing ribF from MG1655
This study
365 366
16
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
pET28a
366
Table 2 Comparison of the physiological parameters between BL21 and MG1655 Growth rate
Glucose uptake
Lactate secretion
Acetate secretion
(h-1)
(mmol g-1 h-1)
(mmol g-1 h-1)
(mmol g-1 h-1)
BL21
0.48±0.01
9.14±0.92
0.84±0.03
2.85±0.13
MG1655
0.67±0.01
9.68±0.79
0
6.11±0.10
367 Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
368 369
17
369
Table 3 Intracellular FAD levels of MG1655 and BL21
370 FAD concentration (pmol/g cells) E. coli BL21
E. coli MG1655
Mid exponential phase
0.081±0.003
0.053±0.002
Late stationary phase
0.053±0.001
0.014±0.004 Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
371
18
Legends to Figures:
371 372
Fig. 1
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
373 374 375
Fig. 1 Identification of riboflavin by HPLC, (a) Riboflavin standard sample and (b)
376
fermentation broth of BL21 (the arrows point at the peak of riboflavin); (c) Mass spectra
377
of the Riboflavin standard sample, (d) Mass spectra of the identified Riboflavin peak in
378
BL21 fermentation broth.
379
19
379
Fig. 2 (b)
(a)
(d)
3.0 MG1655 BL21
Relative expression level
2.5 2.0 1.5 1.0 0.5 0.0
ribB
ribE
ribC
ribF
380 381
Fig. 2 Growth, glucose consumption and acetate and lactate secretion in (a) MG1655 and
382
(b) BL21 in AM1 medium; (c) riboflavin secretion in MG1655 and BL21 in AM1
383
medium; (d) comparison of associated gene expression of the riboflavin synthesis
384
pathway between MG1655 and BL21.
385 386
20
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
(c)
386
Fig. 3
66.2 kDa 45.0
35.0
RibF (34.7 kDa)
387
1
2
3
388 389
Fig. 3 The expression of RibFs. Lane 1, protein ladder; Lane 2, RibF expression from
390
MG1655; lane 3, RibF expression from BL21.
21
Downloaded from http://femsle.oxfordjournals.org/ at Pennsylvania State University on May 25, 2015
25.0
