Transcription levels of key metabolic genes are the cause for different glucose utilization pathways in E. coli B (BL21) and E. coli K (JM109)

Journal of Biotechnology 109 (2004) 21–30

Transcription levels of key metabolic genes are the cause for different glucose utilization pathways in E. coli B (BL21) and E. coli K (JM109) Je-Nie Phue, Joseph Shiloach∗ Biotechnology Unit, NIH, NIDDK, Building 14A Rm 173, Bethesda, MD 20892, USA Received 16 November 2002; accepted 14 October 2003

Abstract Acetate accumulation is a common problem observed in aerobic high cell density cultures of Escherichia coli. It has been hypothesized in previous reports that the glyoxylate shunt is active in E. coli BL21, the low acetate producer, and inactive in E. coli JM109, the high acetate producer. This hypothesis was further strengthened by incorporating 13 C from uniformly labeled glucose into TCA cycle intermediates. Using northern blot analyses, the current report demonstrates that the reason for the inactivity of the glyoxylate pathway in E. coli JM109 is the no apparent transcription of isocitrate lyase (aceA) and malate synthase (aceB), and transcription of the isocitrate lyase repressor (iclR). The reverse is seen in E. coli BL21 where the glyoxylate pathway is active due to constitutive transcription of aceA and aceB and no transcription of the iclR. In addition, there is a difference between the two strains in the transcription of the acetyl-CoA synthetase (acs), phosphotransacetylase–acetate kinase (pta–ackA) pathway, and pyruvate oxidase (poxB), pathway. The transcript of acs is higher in E. coli BL21 and lower in the E. coli JM109, while the reverse is true for poxB transcription. Published by Elsevier B.V. Keywords: E. coli; Glucose; Acetate; Glyoxylate; Northern blot

1. Introduction Escherichia coli is an efficient source for recombinant protein production. Growth of the organism to high cell densities, which is the standard procedure (Lee, 1996; Riesenberg and Guthke, 1999), is achieved by controlling the supply of oxygen and glucose. When E. coli is grown in excess glucose, the ∗ Corresponding author. Tel.: +1-301-496-9719; fax: +1-301-451-5911. E-mail address: [email protected] (J. Shiloach).

0168-1656/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.jbiotec.2003.10.038

carbon flux through glycolysis exceeds the capacity of the TCA cycle and acetate accumulation occurs (Majewski and Domach, 1990). High concentrations of acetate have been shown to reduce growth rate and recombinant protein synthesis (Kleman and Strohl, 1994; Ko et al., 1995). As a result, several strategies have been developed to limit acetate accumulation in high cell density cultures, including gradual addition of glucose to the culture using fed-batch techniques and the development of mutant strains with altered metabolic patterns (Lee, 1996; Riesenberg and Guthke, 1999; Contiero et al., 2000).

22

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30 Aconitase

Glucose

Citrate

Isocitrate

Citrate synthase

Pyruvate

Isocitrate lyase repressor

Acetyl-CoA

Phosphotransacetylase

Acetyl-P Acetate kinase

Acetyl-CoA synthetase

Isocitrate lyase

FadR

Acetyl-AMP

Isocitrate dehydrogenase

FruR, IHF

α-Ketoglutarate Succinyl CoA synthetase

Acetyl-CoA synthetase

Glyoxylate

Acetate Pyruvate oxidase B

Oxaloacetate Malate dehydrogenase

Succinate Malate synthase Succinate dehydrogenase

Malate

Fumarase

Fumarate

Fig. 1. TCA cycle and glyoxylate shunt pathway in E. coli. FadR-regulator of fatty acid metabolism that activates isocitrate lyase repressor; FruR-fructose repressor and IHF-integration host factor.

When growing E. coli on glucose, acetate is produced from pyruvate by pyruvate oxidase (PoxB) and from acetyl-CoA by phosphotransacetylase (Pta) and acetate kinase (AckA; El-Mansi and Holms, 1989; Kleman and Strohl, 1994). The accumulated acetate is converted back to acetyl-CoA by acetyl-CoA synthetase (Acs), and by reversing the phosphotransacetylase–acetate kinase (Pta–AckA) pathway (Fig. 1). The formed acetyl-CoA is further metabolized through the TCA cycle and the glyoxylate shunt pathway by isocitrate lyase (ICL) and malate synthase (MalS; El-Mansi and Holms, 1989; Kornberg, 1966). When acetate is the sole carbon source, the glyoxylate shunt is the main pathway for acetate utilization (Oh et al., 2002; Kornberg, 1966). This pathway is regulated by isocitrate dehydrogenase (IDH) and by the activation of the glyoxylate shunt operon. IDH is a TCA cycle enzyme that competes with isocitrate lyase on isocitrate utilization. The enzyme is regulated by phosphorylation and dephosphorylation through IDH kinase/phosphatase (AceK; Cozzone, 1998; Borthwick et al., 1984; Stueland et al., 1988). The aceK gene is located on the glyoxylate operon, which

contains also the genes for isocitrate lyase (aceA) and malate synthase (aceB; Chung et al., 1988). The expression of the aceBAK is under the control of two regulatory proteins: isocitrate lyase repressor (IclR), which inhibits the expression of the glyoxylate operon, and the regulator of fatty acid metabolism (FadR), which activates the expression of the iclR gene (Gui et al., 1996; Sunnarborg et al., 1990). We have previously reported that E. coli BL21, a derivative of E. coli B, secrets very little acetate even in the presence of high levels of glucose (Shiloach et al., 1996; van de Walle and Shiloach, 1998). On the other hand, E. coli JM109, a K12 derivative, secrets acetate even under fed-batch conditions. Since both strains produce acetate at a similar rate, the existence of an acetate control mechanism was suggested (Shiloach et al., 1996). This mechanism is thought to operate in E. coli BL21 when the acetate concentration reaches beyond 1 g/l. Due to this mechanism, the bacteria starts to consume acetate and, therefore, its accumulation decreases. Activation of the glyoxylate shunt has been proposed as a route for acetyl-CoA consumption with a higher ICL activity observed for BL21 in comparison to JM109 in batch cultures with

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

high levels of glucose (van de Walle and Shiloach, 1998). This observation was confirmed by the determination of the carbon flux through the TCA cycle and the glyoxylate shunt in E. coli BL21 and JM 109 using 13 C NMR and mass spectrometric analysis (Noronha et al., 2000). These studies showed that the glyoxylate shunt in E. coli BL21 is active at 22% of the flux through the TCA cycle, but there is no flux through the glyoxylate shunt in E. coli JM109 (Noronha et al., 2000). Although the enzyme activity, the metabolic flux measurements and the 13 C NMR/MS analyses provide an accurate reflection of the TCA cycle and the glyoxylate shunt activities, they cannot fully explain the reason for the difference between the two strains. It is not clear which enzymes are responsible for this difference, and if the cause for the difference is the inhibition of specific gene expression or inhibition of already expressed proteins. To answer this question, direct analysis of gene transcription is needed. In this work, we attempt to clarify this point by performing northern blot analysis of the specific enzymes and regulatory proteins associated with the glyoxylate shunt operation. In addition, to obtain more comprehensive picture we examined the gene expression of the enzymes involved in acetate uptake and production.

2. Materials and methods 2.1. Bacterial strains The two E. coli strains used were BL21(␭DE3) (F− , ompT, hsdSB (rB− , mB+ ), dcm, gal (DE3), Cmr ) and JM109(DE3) (endA1, recA1, gyrA96, thi, hsdR17 (rk − , mk + ), relA1, supE44, ␭− , (lac-proAB), [F , traD36, proAB, lacIq ZM15], ␭DE3). Both strains were obtained from Promega (Madison, WI). 2.2. Fermentation and sample preparation Both strains were grown at 37 ◦ C in modified LB medium containing 10 g/l tryptone, yeast extract (5 g/l for BL21, 15 g/l for JM109), 5 g/l NaCl, and 5 g/l K2 HPO4 . After sterilization, 10 mM MgSO4 , 1 ml/l trace metal solution, and glucose at a set concentration for each experiment (40 g/l for batch fermentations and

23

2.0 g/l for the fed-batch fermentations) were added. Overnight cultures grown at 37 ◦ C were used to inoculate 4.0 l of medium in a B. Braun fermentor equipped with data acquisition and control system. The cultures were grown to high cell density, the pH was controlled at 7.0 by the addition of 50% NH4 OH, and dissolved oxygen was kept above 30% of saturation at all times. Glucose feeding was conducted in response to the dissolved oxygen concentration as described previously (Shiloach et al., 1996). Samples for acetic acid analysis were collected at regular intervals, centrifuged and the supernatant was kept at −20 ◦ C for further analysis. Samples for total RNA purification were collected and centrifuged at 14,000 × g for 1 min at 4 ◦ C, the supernatant was removed and the cell pellets were quickly frozen in dry ice and stored at −80 ◦ C. 2.3. Analytical methods Acetic acid in the culture supernatant was detected using a Boehringer Mannheim Kit 148621. Acetate determination is based on the formation of NADH, while acetate is converted to citrate and acetyl-CoA in the presence of Acs. Glucose in the culture supernatant was determined using a YSI glucose analyzer (YSI Inc., Yellow Spring OH). 2.4. Preparation of probe DNA E. coli BL21 genomic DNA was used as a template for amplifying fragments of aceA, aceB, acs, ackA, icdA, iclR, poxB, and pta. The gene sequence was identified using BLAST search of E. coli K12 genome, and the following primers were constructed: • aceA: 5 -CCAGTTCATCACCCTGGCAGGTATC3 (forward), 5 -GATTCTTCAGTGGAGCCGGTC3 (reverse); • aceB: 5 -CACATGGATCGCTCACCCAG-3 (forward), 5 -GAAATTTCAGCCGTCGCCG CATCTTC-3 (reverse); • ackA: 5 -CAGCTGACTGCTATCGGTCAC-3 (forward), 5 -CAGGTTGTAAGGCAGGGCGTA GAG-3 (reverse); • acs: 5 -CCGCCATCTGCAAGAAAACGGC-3 (forward), 5 -CACACCTTCGTCGGAAGTGATC3 (reverse);

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

3. Results 3.1. Growth and acetate production Growth profile of E. coli BL21 and E. coli JM109 at low (fed-batch; 2.0 g/l initial concentration) and high (batch; 40 g/l initial concentration) glucose and the acetic acid production pattern are seen in Fig. 2. The growth and acetate production profiles confirmed previously published results (Shiloach et al., 1996). During the growth at high glucose, E. coli JM109 accumulated acetate throughout the fermentation until b

70

10

Cell OD(BL21) Cell OD(JM109)

60

9

a

8

Acetate(BL21)

50

d

7

Acetate(JM109)

40

6

c

5

30

4

20

3 2

10

2.5. Northern blot analysis

1

0

0 0

(A)

1

2

3

4

5

6

7

8

9

Fermentation Time (hr) b

80

10 Cell OD(BL21)

Cell OD (600nm)

Total RNA was isolated with MasterPure RNA Purification Kit (Epicentre Technologies, Madison WI.) according to the manufacturer’s protocol (Kit MCR 85102). The concentration of RNA was determined by measuring the absorbance at 260 nm (A260 ). The purity of RNA was determined by running a 1% agarose–formadehyde denaturing gel and measuring the ratio of the readings at 260 and 280 nm (A260 /A280 ). The isolated RNA had an A260 /A280 ratio of 1.85–1.95. The isolated RNA (10 ␮g per well) was separated using a 1% agarose–formadehyde denaturing gel at 75 V. The gels were blotted on Nytran Super Charge membrane (11 cm × 14 cm; Schleicher & Schuell, NH, USA) at room temperature in 20× SSC. The membranes were fixed by UV-induced cross-linking. The hybridization with 32 P-labeled DNA probes were performed with Quickhyb solution (Stratagene, La Jolla, CA) as recommended by the

Acetate Amount (g/L)

DNA amplification was performed in 50 ␮l mixtures using 5 U of Taq DNA polymerase (Boehringer Mannheim), 20 ng of genomic DNA/␮l, 100 ␮M of each deoxynucleoside triphosphate, 1 ␮M of each primer, and 1× Taq buffer containing MgCl2. PCR was carried out for 30 cycles, each comprising 1 min of denaturing at 94 ◦ C, 1 min of annealing at 64 ◦ C, and 3 min of extension at 72 ◦ C. The PCR products were purified by using gel extraction kit (Qiagen, Valencia, CA). The PCR products were labeled with 32 P using ready-to-go DNA-labeling beads (Amersham Pharmacia Biotech, Piscataway, NJ) and were purified by Probe Quant G-50 Micro column (Amersham Pharmacia Biotech, Piscataway, NJ).

manufacturer’s protocol. Northern blots were repeated to verify reproducibility of results.

70

Cell OD(JM109)

60

Acetate(BL21)

9 8 7 6

Acetate(JM109)

a

50

5

40 30 d

20 c

10 0

(B)

4 3 2 1 0

Acetate Amount (g/L)

• icdA: 5 -CACCAAACGTCTGGTTCGTGCAGC G-3 (forward), 5 -CTGCCAGGGCGTCAGAAATG-3 (reverse); • iclR: 5 -GTACGCATCTGATGCGAATGTCCG-3 (forward), 5 -GTACGCCAGCGTCACTTCCTTC3 (reverse); • poxB: 5 -CGATGGAGATGAAAGCTGGT-3 (forward), 5 -CTGAAACCTTTGGCCTGTTC-3 (reverse); • pta: 5 -CTGATCCCTACCGGAACCAG-3 (forward), 5 -GCAGACCTTCAACGTAGCTC-3 (reverse).

Cell OD (600nm)

24

0 1 2 3 4 5 6 7 8 9 10 11 12 Fermentation Time (hr)

Fig. 2. Growth and acetate production during the growth of E. coli BL21 and JM109: (A) high glucose and (B) low glucose.

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

its concentration reached 7.5 g/l when the bacteria entered the stationary phase. In contrast, E. coli BL21, at the same conditions accumulated acetate only to a maximum concentration of about 2.5 g/l, then decreased to almost zero. At the growth at low glucose, the accumulated acetate in E. coli JM109, was decreasing when the bacteria reached the stationary phase. The acetate accumulation pattern in E. coli BL21 was similar to the one obtained a high glucose but its concentration was lower. 3.2. Transcripts of genes involved in the glyoxylate shunt pathway Transcripts of isocitrate lyase, malate synthase, and isocitrate lyase repressor (which are part of the glyxoylate shunt pathway), and IDH (which is at the cross-point between the TCA cycle and the glyoxylate shunt; Fig. 1) were detected by northern blot analysis. The overall scheme is shown in Fig. 1 and the results are shown in Fig. 3 and Table 1.

25

3.2.1. Batch fermentation at high glucose concentration At high glucose, the gene transcription detected by northern blot (Fig. 3A) correlated to the acetate production pattern of both E. coli BL21 and JM109 (Fig. 2A). There was almost no expression of isocitrate lyase and malate synthase in E. coli JM109, both at the mid log phase and at the late log phase. And at the same time, there was transcript of isocitrate lyase repressor. A reverse pattern was seen in E. coli BL21, there was transcript of isocitrate lyase and malate synthase at the mid and the late log phase, though early in the fermentation the transcripts at the mid log phase was lower, this points to the possibility that the aceBAK was not fully activated. The transcript of isocitrate lyase repressor was visible at the mid log phase but was not detectable at the late log phase. As for IDH, its transcript in BL21 was higher at the mid log phase and lower at the late log phase when the acetate concentration was close to zero, indicating higher glyoxylate pathway activity at this point.

Fig. 3. Northern blot analysis of isocitrate lyase, malate synthase, isocitrate lyase repressor, and isocitrate dehydrogenase in E. coli BL21 and JM 109: (A) high glucose and (B) low glucose. The a, c samples taken at mid log phase and the b, d samples taken at late log phase.

26

Strain

Fermentation

Isocitrate lyase

Malate synthase

Isocitrate lyase repressor

Isocitrate dehydrogenase

Mid log phase

Late log phase

Mid log phase

Late log phase

Mid log phase

Late log phase

Mid log phase

Late log phase

BL21

High glucose batch Low glucose fed-batch

++ +++

+++ +++

++ ++

+++ +++

+ ∗

– ∗

+++ +++

+ +++

JM109

High glucose batch Low glucose fed-batch

– –

∗ ∗

– –

∗ ∗

++ ++

++ ∗

+++ +++

+++ ++

Acetyl-CoA synthetase

Acetate kinase

Phosphotransacetylase

Pyruvate oxidase

BL21

High glucose batch Low glucose fed-batch

+ +++

++ +++

+ +++

– ++

++ ++

– +

– +

++ ∗

JM109

High glucose batch Low glucose fed-batch

– –

∗ ∗

++ ++

++ ++

++ +

++ +

+++ ++

++ ∗

–, not detectable; ∗, trace; +, low expression; ++, mid expression; +++, high expression.

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

Table 1 Expression summary of the TCA cycle and the glyoxylate pathway enzymes

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

In JM109, the transcript of IDH was similar at both phases. 3.2.2. Fed-batch fermentation at low glucose concentration The transcripts of the four genes at the low glucose (fed-batch) fermentation (Fig. 3B) showed similar results to those obtained from the high glucose fermentation with the exception that less isocitrate lyase repressor was observed in JM109 at late log phase. Further, only traces of isocitrate lyase and malate synthetase were observed. This indicates low activity of the glyoxylate pathway. IDH transcript was similar for both strains at the mid log phase, but at the late log phase its transcript was lower in JM109 than in BL21. The combined results are summarized in Table 1. 3.3. Transcripts of acetate consumption and production genes Transcripts of acetyl-CoA synthetase (the gene responsible for metabolizing acetate by converting it to acetyl-CoA), and phosphotransacetylase and acetate

27

kinase (two genes responsible for converting acetylCoA to acetate) and pyruvate oxidase B (the gene responsible for directly converting pyruvate to acetate), were detected by northern analysis is seen in Fig. 4 and Table 1. 3.3.1. Batch fermentation at high glucose concentration As shown in Fig. 4A, transcript of acetyl-CoA synthetase at high glucose was high in BL21 but not detectable in JM109. Transcript of acetate kinase and phosphotransacetylase were visible in both strains, but, were lower in BL21 at the mid log phase and barely detectable at the late log phase. The transcripts of pyruvate oxidase B, at the late log phase were similar in the two strains, but at the mid log phase, the transcript was not detectable in BL21 while it was very high in JM109. 3.3.2. Fed-batch fermentation at low glucose concentration At low glucose concentration the transcript of acetyl-CoA synthetase was high in BL21 both at mid

Fig. 4. Northern blot analysis of acetyl-CoA synthase, acetate kinase, phosphotransacetylase, and pyruvate oxidase in E. coli BL21 and JM 109: (A) high glucose and (B) low glucose. The a, c samples taken at mid log phase and the b, d samples taken at late log phase.

28

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

log and late log phase, but there was no transcript or very low transcript in JM109 (Fig. 4B). Acetate kinase and phosphotransacetylase transcripts in BL21 were higher at the mid log phase and lower at the late log phase, while their transcript in JM109 were similar at both the mid log and the late log phase. Pyruvate oxidase transcript was higher in JM109 especially at the mid log phase. The combined results are summarized in Table 1.

4. Discussion 4.1. Glyoxylate shunt genes The glyoxylate shunt pathway is a vital part of E. coli metabolism (Kornberg and Krebs, 1957). The pathway allows the bacteria to utilize acetate as the sole carbon source to generate the energy and the intermediates required for biosynthesis from this two-carbons compound. The glyoxylate shunt is part of the TCA cycle and its activity must be based on the carbon source and the requirements of the organism for energy and biosynthesis. The metabolic and regulatory proteins of the glyoxylate shunt pathway (isocitrate lyase, malate synthase and IDH kinase/phosphates) reside in the same operon (Chung et al., 1988; Sunnarborg et al., 1990). The expression of this operon is, in part, under the control of the isocitrate lyase repressor (IclR). IclR expression seems to be in response to acetate, to FadR (regulator of fatty acid metabolism) and to a wide array of conditions that reflect the general metabolic state of the cell (Chin et al., 1989; Gui et al., 1996; Ramseier et al., 1995; Resnik et al., 1996). The constitutive activation of the glyoxylate shunt pathway may be due to low transcription of iclR in BL21. In contrast, iclR is highly expressed in JM109 and the pathway is inactive even when acetate concentration was above 5 g/l. IclR is under the control of the transcriptional repressor FadR. It is possible that in BL21 FadR activity is very low or, that positive factors such as fructose repressor (FruR) and integration host factor (IHF) stimulate the transcription of aceBAK. (Ramseier et al., 1995; Resnik et al., 1996). The TCA enzyme, IDH also plays a part in the glyoxylate pathway activation. The enzyme, which determined how much isocitrate will be available for

isocitrate lyase, is regulated by phosphorylation (Stueland et al., 1988). The northern blot analysis indicated that the enzyme is regulated at the transcription level since there was lower amount of icdA in E. coli BL21 late in the log phase. As a result more acetate is being directed to the glyoxylate shunt, and acetate consumption is increased. Our findings support the information and conclusions obtained in previous studies, which were based on enzyme activity, metabolic flux analysis and, 13 C flux analyses using NMR and MS (Shiloach et al., 1996; van de Walle and Shiloach, 1998; Noronha et al., 2000). Those studies concluded that in E. coli K (JM 109), the glyoxylate shunt is not active and that only the TCA cycle is active; while in E. coli B, both TCA cycle and the glyoxylate shunt are active. 4.2. Acetate production and consumption genes In addition to the glyoxylate shunt and the TCA cycle, acetate concentration is also controlled by acetyl-CoA synthetase (which convert acetate to acetyl-CoA) and phosphotransacetylase and acetate kinase (which produces acetate from acetyl-CoA). In addition, acetate can be produced directly from pyruvate by the action of pyruvate oxidase B. The phosphotransacetylase–acetate kinase pathway is considered to be constitutive (Brown et al., 1977), while acetyl-CoA synthetase is under the control of the sigma factors rpoS and rpoD, and requires cAMP receptor protein, and the oxygen regulator (Fnr) (Kumari et al., 2000a,b) for full expression. The acetyl-CoA synthetase pathway is induced by acetate for the purpose of scavenging extracellular acetate (Brown et al., 1977) In contrast, the phosphotransacetylase–acetate kinase pathway functions primarily in a catabolic role, excreting acetate and generating ATP during aerobic growth on excess glucose, while its low-affinity reverse pathway activates acetate consumption only when acetate is present extracellulary in large quantity (Brown et al., 1977). It is not clear what is the role of pyruvate oxidase B, It is believed that pyruvate oxidase may function as a safety valve by converting excess pyruvate to acetate rather than to acetyl-CoA, thus maintaining the intracellular CoA pool for other metabolic functions (Abdel-Hamid et al., 2001). The transcript pattern of acetyl-CoA synthetase is similar to that of isocitrate lyase and malate synthase,

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

There was transcription in BL21 and none in JM109, both at the low and high glucose growth strategies. Its absence in JM109 contributes to the acetate accumulation in this strain, and the constitutive expression of acs in BL21 contributes to the decrease in acetate concentration. As for the pta–ackA, it seems that this pathway may not significantly affect acetate accumulation since there was no clear difference between the two strains, except that ackA transcription was higher in JM109 at high glucose growth. Pyruvate oxidase, on the other hand, seems to play a stronger role in acetate accumulation since its transcript was higher in JM109. 4.3. Concluding remarks The emerging picture from the northern blot analyses points to the possibility that the glyoxylate shunt and acetyl-CoA synthethase are constitutively expressed in E. coli BL21, while the same functions are repressed in E. coli JM109, this general scheme is seen both in low glucose and high glucose growth. It was anticipated that the glyoxylate shunt will be active when acetate is the carbon source or when glucose concentration is high. However, the performance of these two strains is not according to this pattern: the glyoxylate shunt was repressed in E. coli JM109 even when acetate concentration was above 5 g/l, and was expressed in E. coli BL21 when the acetate concentration is only 1 g/l. It is possible that the transcription of the glyoxylate shunt pathway is correlated with the acetyl-CoA synthetase because the glyoxylate shunt requires additional acetyl-CoA that is supplied through acetate uptake by acetyl-CoA synthetase. It seems that at the same glucose concentration, the two strains utilize a different metabolism, which includes glyoxylate pathway, acetyl-CoA synthetase and phosphotransacetylase–acetate kinase. Perhaps northern blot analysis of the glyoxylate shunt pathway and acetyl-CoA synthetase can be a potential tool for selecting a host cell for high cell density culture and high expression of recombinant protein.

References Abdel-Hamid, A.M., Attwood, M.M., Guest, J.R., 2001. Pyruvate oxidase contributes to the aerobic growth efficiency of Escherichia coli. Microbiology 147, 1483–1493.

29

Borthwick, A.C., Holms, W.H., Nimmo, H.G., 1984. Isolation of active and inactive forms of isocitrate dehydrogenase from Escherichia coli ML308. Eur. J. Biochem. 141, 393–400. Brown, T.D.K., Jones-Mortimer, M.C., Kornberg, H.L., 1977. The enzymatic interconversion of acetate and acetyl-coenzyme A in Escherichia coli. J. Gen. Microbiol. 102, 327–336. Chin, A.M., Feldheim, D.A., Saier Jr, M.H., 1989. Altered transcriptional patterns affecting several metabolic pathways in strains of Salmonella typhimurium which overexpress the fructose region. J. Bacteriol. 171, 2424–2434. Chung, T., Klumpp, D.J., LaPorte, D.C., 1988. Glyoxylate bypass operon of Escherichia coli: cloning and determination of functional map. J. Bacteriol. 170, 386–392. Contiero, J., Beatty, C., Kumari, S., DeSanti, C.L., Strohl, W.R., Wolfe, A.J., 2000. Effects of mutants in acetate metabolism on high-cell-density growth of Escherichia coli. J. Ind. Microbiol. 24, 421–430. Cozzone, A.J., 1998. Regulation of acetate metabolism by protein phosphorylation in enteric bacteria. Annu. Rev. Microbiol. 52, 127–164. El-Mansi, E.M.T., Holms, W.H., 1989. Control of carbon flux to acetate excretion during growth of E. coli in batch and continuous culture. J. Gen. Microbiol. 135, 2875–2883. Gui, L., Sunnaborg, A., LaPorte, D.C., 1996. Regulated expression of a repressor protein: FadR activates iclR. J. Bacteriol. 178, 4704–4709. Kleman, G.L., Strohl, W.R., 1994. Acetate metabolism by Escherichia coli in high-cell-density fermentation. Appl. Env. Microbiol. 60, 3952–3958. Kornberg, H.L., Krebs, H.A., 1957. Synthesis of cell constituents from C2 units by modified tricarboxylic acid cycle. Nature 179, 988–991. Kornberg, H.L., 1966. The role and control of the glyoxylate cycle in Escherichia coli. J. Biochem. 99, 1–11. Ko, Y.F., Bentley, W.E., Weigand, W.A., 1995. The effects of cellular energetics on foreign protein production. Appl. Biochem. Biotechnol. 50, 145–159. Kumari, S., Beatty, C.M., Browning, D.F., Busby, S.J.W., Simel, E.J., Hovel-Miner, G., Wolfe, A.J., 2000a. Regulation of acetyl Coenzyme A synthetase in Escherichia coli. J. Bacteriol. 182, 4173–4179. Kumari, S., Simel, E.J., Wolfe, A.J., 2000b. ␴70 is the principal sigma factor responsible for transcription of acs, which encodes acetyl Coenzyme A synthetase in Escherichia coli. J. Bacteriol. 182, 551–554. Lee, S.Y., 1996. High cell density culture of Escherichia coli. Trends Biotechnol. 14, 98–105. Majewski, R.A., Domach, M.M., 1990. Simple constrained optimization view of acetate overflow in E. coli. Biotechnol. Bioeng. 35, 732–738. Noronha, S.B., Yeh, H.J.C., Spande, T.F., Shiloach, J., 2000. Investigation of the TCA cycle and the glyoylate shunt in Escherichia coli BL21 and JM109 using 13 C-NMR/MS. Biotechnol. Bioeng. 68, 316–327. Oh, M.K., Rohlin, L., Kao, K.C., Liao, J.C., 2002. Global expression profiling of acetate-grown E. coli. J. Biol. Chem. 227, 13175–13183.

30

J.-N. Phue, J. Shiloach / Journal of Biotechnology 109 (2004) 21–30

Ramseier, T.M., Bledig, S., Michotey, V., Feghai, R., Saier Jr, M.H., 1995. The global regulatory protein, FruR, modulates the direction of carbon flow in Escherichia coli. Mol. Microbiol. 16, 1157–1169. Resnik, E., Pan, B., Ramani, N., Freundlich, M., LaPorte, D.C., 1996. Integration host factor amplifies the induction of aceBAK operon of Escherichia coli by relieving IclR repression. J. Bacteriol. 178, 2715–2717. Riesenberg, D., Guthke, R., 1999. High-cell-density cultivation of microorganisms. Appl. Microbiol. Biotechnol. 51, 422– 430. Shiloach, J., Kaufman, J., Guillard, A.S., Fass, R., 1996. Effect of glucose supply strategy on acetate accumulation, growth,

and recombinant protein production by Escherichia coli BL21 (DE3) and Escherichia coli JM109. Biotechnol. Bioeng. 49, 421–428. Stueland, C.S., Gorden, K., LaPorte, D.C., 1988. The isocitrate dehydrogenase phosphorylation cycle: identification of the primary rate-limiting step. J. Biol. Chem. 283, 19474–19475. Sunnarborg, A., Klumpp, D., Chung, T., Laporte, D.C., 1990. Regulation of the glyoxylate bypass operon: cloning and characterization of iclR. J. Bacteriol. 172, 2642–2649. van de Walle, M., Shiloach, J., 1998. Proposed mechanism of acetate accumulation in two recombinant Escherichia coli strains during high density fermentation. Biotechnol. Bioeng. 57, 71–78.