on october 26, 2020 by guest · 4/28/2014 · 18 the chinese university of hong kong, shatin, nt,...
TRANSCRIPT
1
Pyruvate-associated acid resistance in bacteria 1
Running title: acid resistance 2
3
Jianting Wu1,2*, Yannan Li3*, Zhiming Cai1,2, Ye Jin4 4
5
1Guangdong and Shenzhen Key Laboratory of Male Reproductive Medicine and Genetics, 6
Institute of Urology, Peking University Shenzhen Hospital, Shenzhen PKU-HKUST Medical 7
Center, Shenzhen, China 8
9
2Shenzhen Second People's Hospital, the First Affiliated Hospital of Shenzhen University, 10
Shenzhen, Guangdong 518035, China 11
12
3Center for Synthetic Biology Engineering Research, Institute of Biomedicine and 13
Biotechnology, Shenzhen Institutes of Advanced Technology, Xueyuan avenue 1068, 14
Shenzhen University Town, Shenzhen, Guangdong 518055, People's Republic of China 15
16
4Department of Medicine and Therapeutics and State Key Laboratory of Digestive Disease, 17
The Chinese University of Hong Kong, Shatin, NT, Hong Kong 18
*These authors contributed equally to this work 19
20
Correspondence and requests for materials should be addressed to Ye Jin (Tel: 852-3763 6100; 21
Fax: 852-2144 5330; E-mail: [email protected]) or Zhiming Cai (Tel: 86 755-8336 5668; Fax: 22
AEM Accepts, published online ahead of print on 2 May 2014Appl. Environ. Microbiol. doi:10.1128/AEM.01001-14Copyright © 2014, American Society for Microbiology. All Rights Reserved.
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
2
86 755-83356952; E-mail: [email protected]) 23
24
25
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
3
Abstract 26
Glucose confers acid resistance to exponentially growing bacteria by repressing 27
formation of the cAMP-CRP complex and consequently activating acid resistance 28
genes. Therefore, in a glucose-rich growth environment, bacteria are capable of 29
resisting acidic stresses due to low levels of cAMP-CRP. Here we reveal a second 30
mechanism for the glucose-conferred acid resistance. We show that glucose induces 31
acid resistance in exponentially growing bacteria through pyruvate, the glycolysis 32
product. Pyruvate and/or the downstream metabolites induce expression of the small 33
noncoding RNA (sncRNA) Spot42 and the sncRNA in turn activates expression of the 34
master regulator of acid resistance, RpoS. In contrast to glucose, pyruvate has little 35
effects on levels of the cAMP-CRP complex and does not require the complex for its 36
effects on acid resistance. Another important difference between glucose and pyruvate 37
is that pyruvate can be produced by bacteria. This means that bacteria have the 38
potential to protect themselves from acidic stresses by controlling glucose-derived 39
generation of pyruvate, pyruvate-acetate efflux or reversion from acetate to pyruvate. 40
We tested this possibility by shutting down the pyruvate-acetate efflux and found that 41
the resulting accumulation of pyruvate elevated acid resistance. Many sugars can be 42
broken into glucose and the subsequent glycolysis generates pyruvate. Therefore, the 43
pyruvate-associated acid resistance is not confined to glucose-grown bacteria but 44
functional in bacteria grown on various sugars. 45
46
Keywords. Acid resistance, pyruvate, glucose, cAMP-CRP, Spot42 47
48
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
4
Introduction 49
Gastric acid (pH 2) in the stomach of the host efficiently kills or inhibits the growth of 50
bacterial pathogens. After being swallowed, enteric organisms have to overcome the 51
low pH of gastric acid. Acidic stresses also come from bacterial growth per se. It is 52
well established that bacteria like Escherichia coli utilize the phosphotransacetylase 53
(Pta)-acetate kinase (AckA) pathway to generate ATP from acetate production in 54
glucose-containing environments even in the presence of ample oxygen (1). The 55
phenomenon of overflow metabolism has been attributed to an imbalance between the 56
fluxes of glucose uptake and those for energy production and biosynthesis (2), 57
probably caused by improperly controlled glucose uptake and/or limited activity of 58
the tricarboxylic acid (TCA) cycle (3). As a weak acid, acetate is toxic to bacteria and 59
has been found to uncouple the transmembrane pH gradient (4, 5), acidify the 60
cytoplasm, and interfere with methionine biosynthesis (6-8). Therefore, acid 61
resistance (AR) is an important ability that E. coli possess to survive low pH and 62
flourish. Indeed, E. coli can be so resistant to low pH that they survive gastric acid, 63
colonize the gut and cause diseases even though small numbers of them (10 to 100) 64
are ingested (9, 10). 65
66
The cyclic AMP (cAMP) receptor protein (CRP) is a crucial regulator of AR in E. coli. 67
It has been demonstrated that CRP negatively regulates AR by repressing a set of AR 68
genes (11, 12). CRP has to form a complex with the signal metabolite cAMP to be 69
functional (13, 14), and cAMP synthesis requires adenylate cyclase (CyaA) that is 70
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
5
activated by the phosphorylated EIIA component of the glucose-specific 71
phosphotransferase system (phospho-EIIAglc)(15, 16). Glucose dephosphorylates the 72
EIIA component, thereby inhibiting the cAMP synthesis and consequently 73
derepressing the cAMP-CRP dependent AR genes (17). Thus, glucose enhances the 74
ability of exponentially growing bacteria to survive low pH (18). 75
76
The CRP-dependent AR functions in the presence of glucose. When glucose is 77
exhausted, the formation of the cAMP-CRP complex is restored and the 78
CRP-repressed AR genes are inactivated. A question, therefore, arises as to how 79
bacteria resist acidic stresses in the presence of substantial cAMP-CRP complex. Here 80
we show that pyruvate and/or its downstream metabolites induce AR by a mechanism 81
that is independent of cAMP-CRP. We reveal that pyruvate and/or its downstream 82
metabolites enhance AR by activating the small noncoding RNA Spot42 and the 83
pyruvate activation of Spot42 does not require the cAMP-CRP complex. Spot42 in 84
turn confers AR by elevating expression of RpoS, a master regulator of stress 85
resistance (19, 20). Interestingly, Spot42 is repressed by the cAMP-CRP complex (21). 86
Thus, Spot42 is involved both in the CRP-independent AR and in the CRP-dependent 87
AR. Spot42 is widely present in medically important gammaproteobacteria genera 88
such as Escherichia, Pantoea, Xenorhabdus, Salmonella, Citrobacter, Yersinia, 89
Serratia, Edwardsiella, Dickeya, Photorhabdus, Enterobacter, Klebsiella, Rahnella 90
and Shigella)(22). In addition, pyruvate is a metabolic product of many sugars. 91
Therefore, the AR mechanisms revealed here are not limited to E. coli grown on 92
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
6
glucose but applicable to diverse bacteria living on various carbon sources. 93
94
95
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
7
Materials and Methods 96
97
E. coli strains and growth conditions 98
The E. coli strain MG1655 (from our laboratory stock) and its isogenic mutants 99
(constructed previously or in this study) were used for phenotypic examination in this 100
study. The MG1655 strain is a widely investigated laboratory strain with annotated 101
sequence and biochemical information. It has been completely sequenced so that 102
genetic engineering can be easily designed and performed. For these reasons, we 103
chose MG1655 for this study and our data could be compared with and integrated 104
with previous findings. The MG1655 strains were grown at 37° C in Luria-Bertani 105
(LB) medium (Affymetrix, Cleveland, OH USA) or on LB agar (Affymetrix, 106
Cleveland, OH USA) (23, 24). The antibiotics ampicillin (50 μg/ml) (Sigma, Saint 107
Louise, MO, USA) and chloramphenicol (12.5 μg/ml) (Sigma, Saint Louise, MO, 108
USA) were added to growth medium or agar when appropriate. 109
110
Gene deletion 111
Gene deletion was performed using the recombineering system as described 112
previously (25, 26). Briefly, the E. coli-K12 strain MG1655 was transformed with 113
plasmid pSim6 (a gift from Dr. Donald Court) from which the expression of the λ 114
recombination proteins was induced for 15 min at 42° C with shaking (220 rpm). The 115
MG1655 strain carrying pSim6 was incubated at 32°C with shaking at 220 rpm until 116
the optical density at 600 nm of the bacterial culture reached 0.4-0.6. Then, PCR 117
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
8
fragments encompassing a loxP-cat-loxP with homology (45 nt) to the regions 118
immediately flanking the deletion target were transformed via electroporation into the 119
MG1655 cells harboring pSim6. After induction of λred functions, recombinants were 120
selected for chloramphenicol resistance (encoded by the cat gene), and were further 121
verified by colony PCR and sequencing. 122
123
spf cloning 124
The selectable loxP-cat-loxP cassette was first inserted immediately after the stop 125
codon of the spf gene (encoding Spot42) the chromosome by recombineering (as 126
described above). We then inserted spf preceded by its native promoter and the 127
loxP-cat-loxP cassette in a pET32a expression vector by recombineering. Specifically, 128
we PCR amplified the spf gene (including its native promoter) and adjacent ‘floxed’ 129
cat cassette using primers that contained homology (45nt) to the plasmid insertion site. 130
The PCR product and expression vector were co-transformed into MG1655 carrying 131
pSim6 after induction of λred. Recombinants were selected for chloramphenicol 132
resistance and verified by PCR and sequencing. The recombinant plasmid was named 133
pSpot42 in which the native promoter of the spf gene drove the Spot42 134
overproduction without induction. To construct an empty plasmid, the loxP-cat-loxP 135
cassette was inserted into pET32a (Novagen, Inc., Madison, WI, USA) and selected 136
for chloramphenicol resistance. The resulting plasmid was named pCm. As pCm did 137
not express Spot42 but was otherwise similar to pSpot42, it served as a negative 138
control. 139
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
9
140
Construction of a chromosomal spf-lacZ transcriptional fusion 141
The loxP-cat-loxP selectable cassette was inserted immediately after the stop codon of 142
the lacZ gene on the MG1655 chromosome using recombineering (as described 143
above). Next, the lacZ-loxP-cat-loxP cassette was PCR amplified and inserted 144
immediately prior to the poly-T string of the spf gene on the chromosome. The 145
inserted lacZ together with its native ribosome binding site (15 base pairs upstream of 146
the start codon) was co-transcribed with spf. 147
148
Beta-galactosidase assay 149
Overnight cultures were diluted 1:500 in fresh LB medium and incubated at 37°C 150
with shaking at 220 rpm for 4 h. Expression of lacZ fusions was quantified using a 151
commercially available beta-galactosidase assay kit (Pierce Biotechnology, Rockford, 152
IL, USA) as described previously (23, 27). Levels of beta-galactosidase were 153
calculated using the following formula: 154
units (MU) = OD420 × 1000/(OD600 × hydrolysis time × volume of lysate) 155
156
Acid resistance assay 157
Overnight cultures were diluted 1:500 in fresh LB medium (pH 6.8) and incubated at 158
37°C with shaking at 220 rpm for 4 h. Cells were then challenged to pH2.0 by 159
adjusting the LB medium with HCl. Cells were treated statically with acid for 2 h and 160
then washed with phosphate buffered saline (PBS) at pH 7.2 to remove acid. Cells 161
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
10
were serially diluted in PBS and 20 μl of each diluted suspension was plated on LB 162
agar. After overnight culture at 37°C, viable cells were determined by counting colony 163
forming units (CFU) of acid-treated cells. As untreated controls, cells before the acid 164
treatment were also quantified by counting CFU as described above. Acid survival (%) 165
was calculated with the following formula: 166
Acid survival (%) = 100 × (CFU of treated/CFU of untreated) 167
All assays were carried out in quadruplicate on two different occasions. 168
169
Growth curve 170
For simultaneous measurement of acid resistance and corresponding cell growth, 100 171
μl of overnight cultures were diluted in 50 ml LB medium in 250 ml flasks, followed 172
by incubation with shaking (220 rpm) for 12 h at 37 °C. Every 2 h, an aliquot was 173
sampled from the culture and optical density at 600 nm was measured after dilution 174
when necessary. 175
176
Determination of extracellular cAMP, pyruvate and acetate 177
To quantify extracellular cAMP, overnight cultures were diluted 1:500 in fresh LB 178
medium and incubated at 37°C with shaking at 220 rpm for 4 h. Cell suspensions 179
were then centrifuged at 12,000 g for 5 min at 4° C. The resulting supernatants were 180
then subjected to cAMP determination using a cAMP Direct Immunoassay kit 181
(BioVision, Mountain View, CA, USA) as directed by the manufacture. To quantify 182
extracellular pyruvate, overnight cultures were diluted 1:500 in fresh LB medium and 183
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
11
incubated at 37°C with shaking at 220 rpm for 4 h before the quantification assay. For 184
time-course measurement of extracellular pyruvate, cells were incubated or 6 h and 185
pyruvate was determined at 1 h-intervals. Pyruvate in the supernatants was assayed 186
using an EnzyChrom Pyruvate Assay Kit (BioAssay Systems, Hayward, CA, USA). 187
Acetate in the supernatants was quantified using an acetate detection kit (Megazme, 188
Bray, Ireland) according to the manufacturer's instructions. 189
190
Western blot 191
Overnight cultures were diluted 1:500 in fresh LB medium and incubated at 37°C 192
with shaking at 220 rpm for 4 h. The bacterial cells were then centrifuged at 12,000 g 193
for 5 min at 4° C and the resulting cell pellets were mixed with sample loading buffer 194
(300 mM Tris-HCl, 6% SDS, 30% glycerol, 0.6% bromphenol blue, 1.2 M 195
beta-mercaptoethanol), boiled for 10 min, and subjected to 12% SDS-PAGE in 196
duplicate. Proteins of one set of gels was then electrotransferred onto an immobilon-P 197
membrane (Millipore, MA, USA) and detected with an E. coli CRP monoclonal 198
antibody (NeoClone Biotechnology, Madison, WI, USA) followed by an anti-mouse 199
horseradish peroxidase (HRP)-conjugated antibody (Invitrogen). Proteins were 200
visualized using the SuperSignal West Pico Chemiluminescent Substrate from Pierce 201
(Rockford, IL, USA). The other set of SDS-PAGE gels was stained with coomassie 202
brilliant blue to demonstrate equal loading. 203
204
Statistical analysis 205
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
12
Independent t tests were used to compare means obtained from pyruvate, cAMP, 206
beta-galactosidase activity and acid resistance assays. P values of < 0.05 were 207
considered statistically significant. 208
209
210
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
13
Results and Discussion 211
212
Pyruvate and/or its downstream metabolites induce acid resistance 213
It is well known that pyruvate is a product of glycolysis. Our data showed that added 214
glucose (20mM), lactose (1%) or glycerol (5%) increased extracellular levels of 215
pyruvate (all P < 0.05) (Figure 1A), indicating that many sugars are converted to 216
pyruvate. Under normal conditions without exogenous pyruvate, intracellular 217
pyruvate levels are 45-fold higher than extracellular pyruvate levels (28). The 218
increased extracellular pyruvate observed with cells grown on various sugars (Figure 219
1A), therefore, indicates a more significant increase in intracellular pyruvate levels. 220
Our time-course assays revealed that E. coli quickly consumed pyruvate added to the 221
growth medium so that extracellular pyruvate decreased to low levels after 5 h of 222
incubation with shaking in LB at 37°C (Figure 1B), suggesting that pyruvate has 223
effects, if any, on biological processes primarily in the log phase. 224
225
Next, we examined if pyruvate affected the ability of bacteria to resist acid. For acid 226
resistance (AR) assays, bacterial cells in the early log phase were treated with 227
acidified LB at pH 2.0 for 2 h and then survival percentage was determined. As a 228
control, glucose at 20 mM dramatically increased AR by a factor of over 1700 (P = 229
0.016)(Figure 1C). This is expected as glucose represses the formation of the 230
cAMP-CRP complex, a transcriptional regulator known to negatively regulate AR by 231
repressing AR genes (11, 12). The requirement of the glucose activation of AR for the 232
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
14
cAMP-CRP complex was confirmed by our AR assays in which the survival 233
percentage of a crp null mutant was 2.59 (±1.5)% after acid challenge (Figure 1C, 234
lower panel), 2500 times higher than that of the wild-type strain (Figure 1C, upper 235
panel)(P = 0.0096). Moreover, added glucose at 20 mM did not increase AR with the 236
crp null mutant (P = 0.933) (Figure 1C, lower panel). Interestingly, added pyruvate at 237
20 mM significantly increased AR as observed with glucose (P = 0.0246) (Figure 1C, 238
upper panel). In contrast to glucose, however, pyruvate did not lose the AR-inducing 239
ability with the crp null mutant (P = 0.0054) (Figure 1C, lower panel). cAMP assays 240
and western blot assays revealed that added pyruvate had no effects on either cAMP 241
(P = 0.818) or CRP levels (Figure 1D). Collectively, these results indicate that 242
pyruvate induces AR by a mechanism independent of the cAMP-CRP complex. 243
244
It is noteworthy that 20 mM was the concentration of pyruvate added to the growth 245
medium. Due to the consumption of pyruvate, the concentration of pyruvate 246
dramatically decreased after 4 h-incubation when the bacteria were subjected to the 247
AR assays. As revealed by the time-course measurement of extracellular pyruvate 248
(Figure 1B), the extracellular pyruvate concentration was less than 47.5 (±3.3) μM 4 h 249
after the concentration dropped to 20 mM. It is intracellular pyruvate that mediates 250
AR. However, exogenous pyruvate at high concentrations just mildly increased 251
intracellular levels of pyruvate (28) since pyruvate uptake is an energy dependent 252
active transport process (29). For these reasons, the concentration of added pyruvate 253
has to be high to cause remarkable effects on AR and regulation of target genes in this 254
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
15
study. In contrast to added pyruvate, added glucose and other carbon sources result in 255
endogenous generation of pyruvate through glycolysis. After growth in LB plus 20 256
mM glucose for 4 h, bacteria displayed extracellular pyruvate concentrations of 40.7 257
(±7.9) μM (Figure 1A), close to the extracellular pyruvate concentrations of the 258
bacteria grown in LB supplemented with 20 mM pyruvate. This indicates that 259
pyruvate generated from glucose is sufficient to have significant effects on AR. 260
261
Increased AR as a result of the addition of pyruvate can result from pyruvate or 262
downstream metabolites such as acetyl-coenzyme A (Ac-coA), acetyl phosphate and 263
acetate which are reversibly interconvert (1). To see if these metabolites are 264
responsible for the elevated AR, we tested the effects of Ac-coA and acetate on 265
log-phase AR. Addition of neither of them had any effect on log-phase AR (both P > 266
0.05) (Figure 1C), excluding roles of Ac-coA and acetate. Conversion from Ac-coA to 267
acetate is mainly achieved by the AckA-Pta pathway (acetate 268
kinase-phosphotransacetylase) (1, 30, 31). As the acetate outflow is an outlet for 269
pyruvate discharge, we reasoned that shutting off this pathway could elevate levels of 270
pyruvate, thereby increasing AR. To test this possibility, we deleted the entire 271
ackA-pta operon. We then monitored extracellular acetate over time of the resulting 272
null mutant (ΔackA-pta) and the wild-type strain to confirm the role of this pathway in 273
acetate outflow. With the wild-type strain, extracellular acetate levels increased during 274
the log phase and dropped after 4 h of incubation (Figure 2A). This phenomenon has 275
previously been termed “acetate switch” which occurs when glucose is exhausted and 276
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
16
bacteria transit from acetate excretion to acetate uptake (1). In contrast, ΔackA-pta 277
produced little acetate throughout the incubation (Figure 2A), confirming that deleting 278
the ackA-pta operon effectively shuts off acetate excretion. We then examined if 279
shutting down the acetate outflow caused pyruvate accumulation and had effects on 280
AR. As shown in Figure 2B, deleting the ackA-pta operon elevated extracellular levels 281
of pyruvate by a factor of 126 (P = 0.0079). Similar observations have also been 282
reported previously (32, 33). As predicted, deleting the ackA-pta operon increased 283
AR by a factor of 577 (P = 0.0236) (Figure 2C), confirming the role of pyruvate for 284
the elevated log-phase AR. 285
286
Pyruvate and/or the downstream metabolites enhance acid resistance by 287
inducing Spot42 288
We next asked how pyruvate and/or the downstream metabolites conferred on bacteria 289
the ability to resist low pH. The following findings linked the pyruvate-induced AR 290
with a small non-coding RNA (sncRNA) named Spot42. Our previous work on a 291
knockout library of sncRNAs has revealed that some sncRNAs upregulated AR. One 292
unpublished AR-regulating sncRNA is Spot42 that is encoded by the spf gene. We 293
found that deleting the spf gene reduced AR during the first 8 h-incubation (all P < 294
0.05) (Figure 3A). We then cloned the spf gene into a multicopy pET32a vector, 295
generating pSpot42 that overexpresses Spot42. Consistent with the knockout data, 296
overproduction of Spot42 increased AR through the culture (all P < 0.05) (Figure 3B). 297
Specifically, bacterial cells were grown in LB and aliquots were measured for both 298
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
17
AR and optical density at OD600 at different time points. By this means, AR as a 299
function of cell growth phase could be evaluated. Cells overproducing Spot42 was 300
4000- to 15000-fold more acid resistant than those lacking the sncRNA after 2-6 h of 301
incubation. This figure dropped to 7-fold after 8 h of incubation and further reduced to 302
1.5-fold after 16 h (Figure 3B). Cell growth curves revealed that 2-6 h corresponded 303
to the log phase (Figure 3A, 3B). These results indicate that AR regulation by Spot42 304
is most evident in the early log phase. 305
306
Spot42 has been identified as a downstream target of the cAMP-CRP complex and is 307
activated by glucose (21). This together with the Spot42 regulation of log-phase AR 308
encouraged us to ask if Spot42 has a role in the pyruvate-conferred log-phase AR. To 309
facilitate evaluation of the possible role of Spot42, we constructed a chromosomally 310
located spf-lacZ transcriptional fusion in E. coli MG1655 so that Spot42 transcription 311
could be easily quantified by measuring beta-galactosidase activity. Subsequent 312
beta-galactosidase assays showed that pyruvate increased the Spot42 expression (P = 313
0.00015) (Figure 4A). Activation of Spot42 by pyruvate maintained in cells deleted 314
for crp (Δcrp) (P < 0.05 at all pyruvate concentrations) (Figure 4A). Thus, in contrast 315
to glucose, pyruvate induces the Spot42 expression by a CRP-independent mechanism. 316
The induction of Spot42 expression by pyruvate and upregulation of log-phase AR by 317
Spot42 suggest that pyruvate and/or the downstream metabolites confers log-phase 318
AR at least in part through Spot42. It is noteworthy that deletion of the spf gene 319
encoding Spot42 did not abolish the pyruvate-conferred AR but the effects of 320
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
18
pyruvate on AR were not statistically significant (P = 0.08) (Figure 4B), confirming 321
the critical role of Spot42 for this AR system. 322
323
RpoS participates in pyruvate/Spot42-conferred acid resistance 324
The phase specificity of Spot42-mediated AR prompted us to link it to growth 325
phase-specific biological processes. One such process is rpoS expression that is not 326
induced until bacteria enter the stationary phase. RpoS is a master regulator of stress 327
resistance (19, 20). This led us to speculate that RpoS may have a role in the 328
Spot42-mediated AR. To quantify the RpoS expression, we employed a previously 329
constructed MG1655 isogenic mutant carrying an rpoS-lacZ translational fusion on 330
the chromosome (23). In support of the above speculation, deleting the spf gene 331
reduced the rpoS-lacZ fusion expression regardless of the pyruvate addition (both P < 332
0.05) (Figure 5A) and overexpressing the sncRNA increased the expression (P = 333
0.0018) (Figure 5B). Given the activating effects of pyruvate on Spot42, we predicted 334
that pyruvate enhanced RpoS expression. As expected, pyruvate increased rpoS-lacZ 335
expression (P = 0.0001) (Figure 5A). Thus pyruvate enhances Spot42, which in turn 336
activates RpoS. However, deleting the spf gene did not abolish the effects of pyruvate 337
on RpoS (P = 0.00036) (Figure 5A), indicating existence of a Spot42-indepenent 338
mechanism. 339
340
To provide more evidence for the role of RpoS in the pyruvate and Spot42 regulation 341
of AR, we tested if removing RpoS abolished the regulation. An rpoS null mutant 342
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
19
(ΔrpoS) showed an extremely low survival rate after 2 h of acid treatment, and Spot42 343
overproduction failed to increase it (Figure 5C). We then employed a previously 344
constructed rpoS mutant d567-1342, which loses the natural regulation of rpoS and 345
constitutively expresses rpoS (23). Removing the natural regulation of rpoS 346
significantly diminished the effects of Spot42 on AR, as Spot42 overproduction 347
increased AR ~500-fold in the wild-type strain but only ~27-fold in d567-1342 (P = 348
0.012)(Figure 5C). SncRNAs regulate gene expression either by directly binding to 349
their mRNA targets or indirectly by acting on transcriptional or translational factors 350
that in turn regulate the gene expression (34). If Spot42 regulate rpoS by a direct 351
mechanism, there should be extensive complementarities between the two RNAs. 352
However, complementarity analysis using the program TargetRNA 353
(http://snowwhite.wellesley.edu/targetRNA) (35) did not reveal any extensive 354
complementarities between rpoS and Spot42, suggesting that Spot42 regulates RpoS 355
through an indirect mechanism. 356
357
The AR system reported in this study is summarized in Figure 6. It appears to be a 358
new AR system different from the currently known systems that are primarily 359
restricted to the stationary phase (24, 36, 37). Acid resistance system 1 (AR1) requires 360
the induction of RpoS but is repressed by glucose, which is contrast to our system. 361
AR2 is glutamate dependent, which requires the presence of glutamate decarboxylase 362
and a putative glutamate:GABA antiporter. Our AR system is not part of AR2 since 363
AR2 is absent from log-phase cells grown in LB plus 4% glucose (38). AR3 is 364
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
20
arginine dependent, requires the presence of arginine decarboxylase (AdiA), and 365
provides a modest level of protection. AR3 is induced by low pH under anaerobic 366
conditions, which is different from our AR system. In contrast to these 367
thoroughly-investigated stationary AR, much less are known about how log-phase 368
cells resist acid since effort towards this may have been discouraged by the fact that 369
log-phase cells are acid sensitive. This study reveals for the first time how 370
exponentially-growing bacteria like E. coli resist low pH through pyruvate and/or its 371
downstream metabolites. Bacteria use pyruvate and/or its downstream metabolites as 372
key effector molecules to induce AR by activating expression of the sncRNA Spot42 373
which in turn up-regulates RpoS. Any biological processes that affect pyruvate levels 374
would, therefore, have an impact on AR. This is confirmed by our observation that 375
increasing pyruvate accumulation by shutting down pyruvate discharge remarkably 376
elevates AR. Another source of pyruvate is sugar metabolism. Bacteria grown on 377
sugars like glucose produce pyruvate inside the cells through glycolysis without a 378
need for active uptake of extracelluar pyruvate. We show that addition of glucose at 379
20 mM resulted in a 240-fold increase in extracellular levels of pyruvate, and 380
enhanced AR accordingly. Concentration of glucose in food and drink easily reaches 381
20 mM. For instance, glucose concentration of orange juices range from 22 to 218 382
mM (39). Thus, this study reveals that bacteria in sugar-rich food or drink are capable 383
of acquiring the ability to resist acid at least partially through the 384
pyruvate/Spot42-dependent AR system. The resulting acid-resistant bacteria could 385
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
21
survive gastric acid and colonize the intestine to become commensal habitants or 386
cause diseases depending on virulence of the bacteria. 387
388
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
22
Acknowledgments 389
This work was supported by the National Natural Science Foundation of China 390
(C010301) and National Program on Key Basic Research Project (2014CB745200). 391
392
393
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
23
References 394
1. Wolfe AJ. 2005. The acetate switch. Microbiology and molecular biology reviews : MMBR 395
69:12-50. 396
2. Farmer WR, Liao JC. 1997. Reduction of aerobic acetate production by Escherichia coli. 397
Appl Environ Microbiol 63:3205-3210. 398
3. Veit A, Polen T, Wendisch VF. 2007. Global gene expression analysis of glucose overflow 399
metabolism in Escherichia coli and reduction of aerobic acetate formation. Appl Microbiol 400
Biotechnol 74:406-421. 401
4. Repaske DR, Adler J. 1981. Change in intracellular pH of Escherichia coli mediates the 402
chemotactic response to certain attractants and repellents. J Bacteriol 145:1196-1208. 403
5. Kihara M, Macnab RM. 1981. Cytoplasmic pH mediates pH taxis and weak-acid repellent 404
taxis of bacteria. J Bacteriol 145:1209-1221. 405
6. Roe AJ, O'Byrne C, McLaggan D, Booth IR. 2002. Inhibition of Escherichia coli growth by 406
acetic acid: a problem with methionine biosynthesis and homocysteine toxicity. Microbiology 407
148:2215-2222. 408
7. Roe AJ, McLaggan D, Davidson I, O'Byrne C, Booth IR. 1998. Perturbation of anion 409
balance during inhibition of growth of Escherichia coli by weak acids. J Bacteriol 410
180:767-772. 411
8. Russell JB, Diez-Gonzalez F. 1998. The effects of fermentation acids on bacterial growth. 412
Adv Microb Physiol 39:205-234. 413
9. Schmid-Hempel P, Frank SA. 2007. Pathogenesis, virulence, and infective dose. PLoS 414
Pathog 3:1372-1373. 415
10. Griffin PM, Tauxe RV. 1991. The epidemiology of infections caused by Escherichia coli 416
O157:H7, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. 417
Epidemiol Rev 13:60-98. 418
11. Castanie-Cornet MP, Foster JW. 2001. Escherichia coli acid resistance: cAMP receptor 419
protein and a 20 bp cis-acting sequence control pH and stationary phase expression of the 420
gadA and gadBC glutamate decarboxylase genes. Microbiology 147:709-715. 421
12. Gosset G, Zhang Z, Nayyar S, Cuevas WA, Saier MH, Jr. 2004. Transcriptome analysis of 422
Crp-dependent catabolite control of gene expression in Escherichia coli. J Bacteriol 423
186:3516-3524. 424
13. Tagami H, Aiba H. 1998. A common role of CRP in transcription activation: CRP acts 425
transiently to stimulate events leading to open complex formation at a diverse set of promoters. 426
The EMBO journal 17:1759-1767. 427
14. Busby S, Ebright RH. 1999. Transcription activation by catabolite activator protein (CAP). 428
Journal of molecular biology 293:199-213. 429
15. Park YH, Lee BR, Seok YJ, Peterkofsky A. 2006. In vitro reconstitution of catabolite 430
repression in Escherichia coli. The Journal of biological chemistry 281:6448-6454. 431
16. Bettenbrock K, Sauter T, Jahreis K, Kremling A, Lengeler JW, Gilles ED. 2007. 432
Correlation between growth rates, EIIACrr phosphorylation, and intracellular cyclic AMP 433
levels in Escherichia coli K-12. Journal of bacteriology 189:6891-6900. 434
17. Hogema BM, Arents JC, Bader R, Eijkemans K, Yoshida H, Takahashi H, Aiba H, 435
Postma PW. 1998. Inducer exclusion in Escherichia coli by non-PTS substrates: the role of 436
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
24
the PEP to pyruvate ratio in determining the phosphorylation state of enzyme IIAGlc. 437
Molecular microbiology 30:487-498. 438
18. Rowbury RJ, Goodson M. 1998. Glucose-induced acid tolerance appearing at neutral pH in 439
log-phase Escherichia coli and its reversal by cyclic AMP. J Appl Microbiol 85:615-620. 440
19. Battesti A, Majdalani N, Gottesman S. 2011. The RpoS-mediated general stress response in 441
Escherichia coli. Annual review of microbiology 65:189-213. 442
20. Lacour S, Landini P. 2004. SigmaS-dependent gene expression at the onset of stationary 443
phase in Escherichia coli: function of sigmaS-dependent genes and identification of their 444
promoter sequences. J Bacteriol 186:7186-7195. 445
21. Polayes DA, Rice PW, Garner MM, Dahlberg JE. 1988. Cyclic AMP-cyclic AMP receptor 446
protein as a repressor of transcription of the spf gene of Escherichia coli. Journal of 447
bacteriology 170:3110-3114. 448
22. Hansen GA, Ahmad R, Hjerde E, Fenton CG, Willassen NP, Haugen P. 2012. Expression 449
profiling reveals Spot 42 small RNA as a key regulator in the central metabolism of Aliivibrio 450
salmonicida. BMC Genomics 13:37. 451
23. Jin Y, Watt RM, Danchin A, Huang JD. 2009. Small noncoding RNA GcvB is a novel 452
regulator of acid resistance in Escherichia coli. BMC Genomics 10:165. 453
24. Lin J, Smith MP, Chapin KC, Baik HS, Bennett GN, Foster JW. 1996. Mechanisms of 454
acid resistance in enterohemorrhagic Escherichia coli. Appl Environ Microbiol 62:3094-3100. 455
25. Sharan SK, Thomason LC, Kuznetsov SG, Court DL. 2009. Recombineering: a 456
homologous recombination-based method of genetic engineering. Nat Protoc 4:206-223. 457
26. Yu D, Ellis HM, Lee EC, Jenkins NA, Copeland NG, Court DL. 2000. An efficient 458
recombination system for chromosome engineering in Escherichia coli. Proc Natl Acad Sci U 459
S A 97:5978-5983. 460
27. Jin Y, Watt RM, Danchin A, Huang JD. 2009. Use of a riboswitch-controlled conditional 461
hypomorphic mutation to uncover a role for the essential csrA gene in bacterial 462
autoaggregation. J Biol Chem 284:28738-28745. 463
28. Yang YT, Bennett GN, San KY. 2001. The effects of feed and intracellular pyruvate levels on 464
the redistribution of metabolic fluxes in Escherichia coli. Metab Eng 3:115-123. 465
29. Lang VJ, Leystra-Lantz C, Cook RA. 1987. Characterization of the specific pyruvate 466
transport system in Escherichia coli K-12. J Bacteriol 169:380-385. 467
30. Hesslinger C, Fairhurst SA, Sawers G. 1998. Novel keto acid formate-lyase and propionate 468
kinase enzymes are components of an anaerobic pathway in Escherichia coli that degrades 469
L-threonine to propionate. Molecular microbiology 27:477-492. 470
31. Klein AH, Shulla A, Reimann SA, Keating DH, Wolfe AJ. 2007. The intracellular 471
concentration of acetyl phosphate in Escherichia coli is sufficient for direct phosphorylation of 472
two-component response regulators. Journal of bacteriology 189:5574-5581. 473
32. Dittrich CR, Vadali RV, Bennett GN, San KY. 2005. Redistribution of metabolic fluxes in 474
the central aerobic metabolic pathway of E. coli mutant strains with deletion of the ackA-pta 475
and poxB pathways for the synthesis of isoamyl acetate. Biotechnology progress 21:627-631. 476
33. Sadykov MR, Thomas VC, Marshall DD, Wenstrom CJ, Moormeier DE, Widhelm TJ, 477
Nuxoll AS, Powers R, Bayles KW. 2013. Inactivation of the Pta-AckA pathway causes cell 478
death in Staphylococcus aureus. J Bacteriol 195:3035-3044. 479
34. Repoila F, Darfeuille F. 2009. Small regulatory non-coding RNAs in bacteria: physiology and 480
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
25
mechanistic aspects. Biol Cell 101:117-131. 481
35. Tjaden B. 2008. TargetRNA: a tool for predicting targets of small RNA action in bacteria. 482
Nucleic Acids Res 36:W109-113. 483
36. Foster JW. 2004. Escherichia coli acid resistance: tales of an amateur acidophile. Nat Rev 484
Microbiol 2:898-907. 485
37. Lin J, Lee IS, Frey J, Slonczewski JL, Foster JW. 1995. Comparative analysis of extreme 486
acid survival in Salmonella typhimurium, Shigella flexneri, and Escherichia coli. J Bacteriol 487
177:4097-4104. 488
38. Castanie-Cornet MP, Penfound TA, Smith D, Elliott JF, Foster JW. 1999. Control of acid 489
resistance in Escherichia coli. J Bacteriol 181:3525-3535. 490
39. Li WJ, Goovaerts P, Meurens M. 1996. Quantitative analysis of individual sugars and acids 491
in orange juices by near-infrared spectroscopy of dry extract. Journal of Agricultural and Food 492
Chemistry 44:2252-2259. 493
494
495
496
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
26
Figure legends 497
498
Figure 1. Pyruvate induces acid resistance independently of the cAMP-CRP complex. 499
(A) Levels of extracellular pyruvate of the E. coli MG1655 strain grown in LB 500
medium alone, LB plus 20 mM glucose (Glc), LB plus 1% lactose (Lac) and LB plus 501
5% glycerol (Gly) for 4 h. (B) Time course of extracellular pyruvate levels of 502
MG1655 grown in LB plus 40 mM pyruvate. (C) Acid survival (%) at pH 2.0 of the 503
wild-type MG1655 (wt) grown in LB alone, LB plus 20 mM glucose (Glc), LB plus 504
20 mM pyruvate (Pyr), LB plus 10 mM acetyl-coenzyme A (coA), and LB plus 20 505
mM acetate (Ac); and of a crp null mutant (Δcrp) grown in LB alone, LB plus 20 mM 506
glucose (Glc) and LB plus 20 mM pyruvate (Pyr). Cells were grown for 4 h and then 507
treated with acidified LB (pH 2.0) for 2 h. Survival percentage was then determined 508
by counting colony forming units. (D) Extracellular cAMP and corresponding CRP 509
levels of MG1655 grown in LB without pyruvate supplementation and LB plus 20 510
mM pyruvate. Cells were grown for 4 h before quantification of cAMP and CRP. 511
512
Figure 2. Elevation of acid resistance as a result of shutting down acetate production. 513
(A) Extracellular levels of acetate in the wild-type MG1655 (wt) and the ackA-pta 514
null mutant (ΔackA-pta) as a function of time. (B) Extracellular levels of pyruvate in 515
wt and ΔackA-pta grown in LB for 4 h. (C) Acid survival (%) at pH 2.0 of wt and 516
ΔackA-pta that were grown in LB for 4 h before 2 h of acid treatment. 517
518
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
27
Figure 3. Effects of Spot42 on bacterial acid resistance. (A) Cell growth (OD600) and 519
acid survival (%) at pH 2.0 of the wild-type E. coli MG1655 (wt) and a spf null 520
mutant (Δspf), as a function of time. (B) Cell growth (OD600) and acid survival (%) 521
at pH 2.0 of Δspf carrying a control vector (pCm) or pSpot42 that overproduces 522
Spot42, as a function of time. Cells were grown for 4 h before the acid resistance 523
assay. OD600 nm was measured as follows. 100 μl of overnight cultures were diluted 524
in 50 ml LB medium in 250 ml flasks, and then incubated at 37 °C. Every 2 h, an 525
aliquot was determined for OD600 after dilution when necessary. 526
527
Figure 4. Induction of Spot42 transcription by pyruvate. (A) Beta-galactosidase 528
activity of mutants carrying a chromosomal spf-lacZ transcriptional fusion with the 529
crp gene (crp+) or without crp (crp-). Cells were grown in LB supplemented with 530
increasing concentrations of glucose (Glc) or pyruvate (Pyr) for 4 h before 531
quantification of beta-galactosidase activity. (B) Pyruvate dramatically increased acid 532
resistance of the wild-type MG1655 (wt) but affected acid resistance of the spf null 533
mutant (Δspf) to a much less extent. However, deletion of spf did not abolish the 534
pyruvate-conferred acid resistance. 535
536
Figure 5. Role of RpoS for pyruvate/Spot42-conferred acid resistance. (A) 537
Beta-galactosidase activity of mutants carrying a chromosomal rpoS-lacZ 538
translational fusion with the small noncoding RNA Spot42 (Spot42+) or without 539
Spot42 (Spot42-). Cells were grown in LB or LB plus 20 mM pyruvate (Pyr) for 4 h 540
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from
28
before measurement of beta-galactosidase activity. (B) Effects of Spot42 541
overproduction from pSpot42 on expression of the rpoS-lacZ translational fusion. The 542
fusion strain carrying an empty vector pCm served as a negative control. (C) Acid 543
survival (%) of various strains carrying pCm or pSpot42. These include the wild type 544
E. coli MG1655 (wt), an isogenic rpoS null mutant (ΔrpoS), and an isogenic mutant 545
d567-1342 where the native regulation of rpoS is removed. All cells were grown in 546
LB for 4 h before the acid resistance assay. 547
548
Figure 6. Pathways of the acid resistance system mediated by pyruvate and Spot42. 549
P-EIIAglc, the phosphorylated EIIA component of the glucose-specific 550
phosphotransferase system; EIIAglc, dephosphorylated EIIA component of the 551
glucose-specific phosphotransferase system; Glc, glucose; Pyr, pyruvate; AR, acid 552
resistance 553
554
on April 25, 2021 by guest
http://aem.asm
.org/D
ownloaded from