Appl Microbiol Biotechnol (2008) 79:471–479 DOI 10.1007/s00253-008-1444-z

APPLIED GENETICS AND MOLECULAR BIOTECHNOLOGY

Corynebacterium glutamicum tailored for high-yield L-valine production Bastian Blombach & Mark E. Schreiner & Tobias Bartek & Marco Oldiges & Bernhard J. Eikmanns

Received: 31 January 2008 / Revised: 29 February 2008 / Accepted: 2 March 2008 / Published online: 1 April 2008 # Springer-Verlag 2008

Abstract We recently engineered the wild type of Corynebacterium glutamicum for the growth-decoupled production of L-valine from glucose by inactivation of the pyruvate dehydrogenase complex and additional overexpression of the ilvBNCE genes, encoding the L-valine biosynthetic enzymes acetohydroxyacid synthase, isomeroreductase, and transaminase B. Based on the first generation of pyruvate-dehydrogenase-complex-deficient C. glutamicum strains, a second generation of high-yield L-valine producers was constructed by successive deletion of the genes encoding pyruvate:quinone oxidoreductase, phosphoglucose isomerase, and pyruvate carboxylase and overexpression of ilvBNCE. In fed-batch fermentations at high cell densities, the newly constructed strains produced up to 410 mM (48 g/l) L-valine, showed a maximum yield of 0.75 to 0.86 mol/mol (0.49 to 0.56 g/g) of glucose in the production phase and, in contrast to the first generation strains, excreted neither pyruvate nor any other by-product tested.

B. Blombach : M. E. Schreiner : B. J. Eikmanns (*) Institute of Microbiology and Biotechnology, University of Ulm, 89069 Ulm, Germany e-mail: [email protected] T. Bartek : M. Oldiges Institute of Biotechnology 2, Research Center Jülich, 52425 Jülich, Germany Present address: M. E. Schreiner R&D Women’s Health, Europe, Johnson & Johnson GmbH, 42289 Wuppertal, Germany

Keywords Corynebacterium glutamicum . L-valine production . Pyruvate dehydrogenase complex . Pyruvate:quinone oxidoreductase . Phosphoglucose isomerase . Pyruvate carboxylase

Introduction Corynebacterium glutamicum is a Gram-positive soil bacterium that grows on a variety of sugars and organic acids. The organism is the workhorse for the production of the amino acids L-glutamate and L-lysine (Liebl 1991; Leuchtenberger et al. 2005; Takors et al. 2007). Furthermore, a few thousand tons per year of other amino acids like L-threonine, L-isoleucine, L-tryptophan, and also Lvaline are produced with C. glutamicum (Eggeling and Bott 2005). The latter branched-chain amino acid is essential for vertebrates and used for infusion solutions, for cosmetics, and as a precursor for the chemical synthesis of some herbicides (Leuchtenberger 1996; Eggeling 2001; Park et al. 2007). Because the yields and the productivities of amino acid production strains still are below the expected theoretical values, there is a large interest to further improve the performance of bacterial production strains (Takors et al. 2007). As shown in Fig. 1, L-valine is synthesized from pyruvate in a pathway comprising four reactions, catalyzed by acetohydroxyacid synthase (AHAS, ilvBN gene product), acetohydroxyacid isomeroreductase (AHAIR, the ilvC gene product), dihydroxyacid dehydratase (DHAD, ilvD gene product), and transaminase B (TA, ilvE gene product; Marienhagen et al. 2005) (for an overview see Patek 2007). The same four enzymes catalyze also the biosynthesis of

472 Fig. 1 Schematic representation of the central metabolism of C. glutamicum with the biosynthetic pathway of L-valine. Abbreviations: AHAIR, acetohydroxyacid isomeroreductase; AHAS, acetohydroxyacid synthase; AK, acetate kinase; DHAD, dihydroxyacid dehydratase; PCx, pyruvate carboxylase; PDHC, pyruvate dehydrogenase complex; PK, pyruvate kinase; PEP, phosphoenolpyruvate; PEPCk, PEP carboxykinase; PEPCx, PEP carboxylase; PGI, phosphoglucose isomerase; PQO, pyruvate:quinone oxidoreductase; PTA, phosphotransacetylase; TA, transaminase B

Appl Microbiol Biotechnol (2008) 79:471–479

Glucose Glucose-6-P

Pentose-P NADP+ NADPH+H+ NADP+ NADPH+H+

PGI Fructose-6-P L-Valine Glyceraldehyde-3-P TA PEP

PEPCx

Acetolactate AHAS

AHAIR

DHAD

2-Oxoglutarate

L-Glutamate

NADPH+H+ NADP+

PK PEPCk

Pyruvate

PQO

Acetate AK

PDHC Acetyl-CoA

PTA

Acetylphosphate

PCx

Citrate

Oxaloacetate

NADP+

TCA cycle

NADPH+H+

2-Oxoglutarate

L-Glutamate

NADPH+H+ NADP+

L-isoleucine from pyruvate and 2-oxobutyrate. The latter enzyme is formed from L-threonine by the threonine dehydratase (TD, ilvA gene product). Ketoisovalerate, the last intermediate of L-valine synthesis, is also the precursor for L-leucine and D-pantothenate biosynthesis. In C. glutamicum strains lacking TD, it has been shown that plasmid-bound overexpression of the genes ilvBNCD or ilvBNCE was beneficial for L-valine production (Sahm and Eggeling 1999; Radmacher et al. 2002). Introduction of the genes encoding a feedback-resistant AHAS enzyme into L-valine-producing C. glutamicum strains led to further improvement of L-valine production (Elisakova et al. 2005). Moreover, Radmacher et al. (2002) showed that inactivation of D-pantothenate biosynthesis by deleting the panBC genes led to an increased L-valine production of C. glutamicum when cultivated under D-pantothenate-limiting conditions. This latter improvement of L-valine production probably can be explained by an increased precursor (pyruvate) availability (Bartek et al. 2008b) because D-pantothenate limitation leads to reduced coenzyme A availability for the reaction of the pyruvate dehydrogenase complex (PDHC). The importance of precursor availability was also highlighted for L-lysine production by increasing the pyruvate and/or oxaloacetate supply either by inactivation of the PDHC complex (Blombach et al. 2007a), overexpression of the pyruvate carboxylase (PCx) gene (Peters-Wendisch

et al. 2001), or inactivation of the phosphoenolpyruvate carboxykinase gene (Riedel et al. 2001). Aside from engineering the biosynthetic pathways and increasing the precursor availability, NADPH supply was also shown to be a critical factor for amino acid production, as has been in detail studied in the case of L-lysine accumulation with C. glutamicum (Marx et al. 2003; Kabus et al. 2007). C. glutamicum possess four enzymes for NADPH generation, notably, glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase of the oxidative pentose phosphate pathway (Moritz et al. 2000), malic enzyme (Gourdon et al. 2000), and isocitrate dehydrogenase (Eikmanns et al. 1995). However, carbon flux analysis in a L-lysine-producing C. glutamicum strain revealed that in minimal medium with glucose the latter two enzymes play a minor role for the generation of NADPH (Marx et al. 1996, 1997, 1999, 2003). Increased flux from glycolysis to the pentose phosphate pathway and, thus, an increased NADPH supply was achieved by introduction of a mutant allele encoding a feedback-resistant 6-phosphogluconate dehydrogenase (Ohnishi et al. 2005) and also by overexpression of fructose 1,6-biphosphatase (Georgi et al. 2005; Becker et al. 2005). Furthermore, Marx et al. (2003) reported that the disruption of the phosphoglucose isomerase (PGI) redirected the carbon flux towards the pentose phosphate pathway

Appl Microbiol Biotechnol (2008) 79:471–479

resulting in increased L-lysine formation. In contrast to L-lysine synthesis, which requires 4 mol NADPH per mol of L-lysine, for the generation of 1 mol L-valine only 2 mol of NADPH are oxidized, one for the conversion of acetolactate by the AHAIR, the other for the synthesis of the primary amino group donor glutamate (Fig. 1). However, based on a correlation between the cell-specific rate of L-valine production and the specific activity of the glucose6-phosphate dehydrogenase at different growth rates of a TD-deficient C. glutamicum strain, Ruklisha et al. (2007) suggested that an increase in NADPH generation might be advantageous for L-valine production. We recently demonstrated the ability of PDHC-deficient C. glutamicum strains to form L-valine, L-alanine, and pyruvate from glucose. Additional plasmid-bound overexpression of the L-valine biosynthesis genes ilvBNCE shifted the product spectrum towards L-valine and resulted in an Lvaline producer, excreting this amino acid in a growthdecoupled manner (Blombach et al. 2007b). In contrast to the previously constructed L-valine producer C. glutamicum ΔilvA ΔpanBC (pJC4ilvBNCD) (Radmacher et al. 2002; see above), the PDHC-deficient strains do not require supplements such as L-isoleucine and D-pantothenate. This fact and the further complication that L-valine was shown to inhibit L-isoleucine uptake of C. glutamicum ΔilvA ΔpanBC (pJC4ilvBNCD) (Lange et al. 2003) suggest that PDHCdeficient C. glutamicum strains are more favorable L-valine producers. Based on the first generation of PDHC-deficient L-valine producers, we here highlight the importance of the enzymes AHAS, AHAIR, and TA for L-valine overproduction by testing the effect of different combinations of plasmid-bound L-valine biosynthetic genes. Furthermore, we investigate the importance of an increased precursor supply by further inactivation of pyruvate-converting enzymes. Finally, we test whether an increase of the NADPH supply, brought about by forcing the carbon flux through the pentose phosphate pathway, is beneficial for L-valine production with C. glutamicum.

Materials and methods Bacterial strains, plasmids, and oligonucleotides All bacterial strains and plasmids and their relevant characteristics and sources are given in Table 1. The oligonucleotides used and their sequences are also listed in Table 1. DNA preparation and transformation The isolation of plasmids from Escherichia coli and C. glutamicum was performed as described before (Eikmanns

473

et al. 1994). Plasmid DNA transfer into C. glutamicum was carried out by electroporation and the recombinant strains were selected on Luria–Bertani Brain Heart Infusion (BHI) agar plates containing 0.5 M sorbitol, 85 mM potassium acetate (corresponds to 0.5% (w/v) acetate), and kanamycin (50 μg ml−1; van der Rest et al. 1999). The isolation of chromosomal DNA from C. glutamicum was performed as described previously (Eikmanns et al. 1994). Electroporation of E. coli was performed with competent cells according to the method of Dower et al. (1988). Construction of C. glutamicum ΔaceE Δpqo Δpgi and C. glutamicum ΔaceE Δpqo Δpgi Δpyc Inactivation of the chromosomal PGI gene (pgi) in C. glutamicum ΔaceE Δpqo (Schreiner et al. 2006) was performed using crossover PCR and the suicide vector pK18mobsacB. DNA fragments covering the 5′-end and the 3′-end of pgi were generated using the primer pairs pgi-d1–pgi-d2 and pgi-d3–pgi-d4, respectively. The two fragments were purified, mixed in equal amounts, and subjected to crossover PCR using primers pgi-d1 and pgi-d4. The resulting fusion product (containing the pgi gene with an internal deletion of 1,504 bp) was ligated into SmaI-restricted plasmid pK18mobsacB and transformed into E. coli. The recombinant plasmid was isolated from E. coli and electroporated into C. glutamicum ΔaceE Δpqo. By application of the method described by Schäfer et al. (1994), the intact chromosomal pgi gene in C. glutamicum ΔaceE Δpqo was replaced by the truncated pgi gene via homologous recombination (double crossover). The screening of the pgi mutants was done on 2xTY agar plates (Sambrook et al. 2001) containing 85 mM potassium acetate (corresponds to 0.5% (w/v) acetate) and 10% (w/v) sucrose. The replacement at the chromosomal locus was verified by PCR using primers pgiout1–pgiout2. Inactivation of the chromosomal PCx gene (pyc) in C. glutamicum ΔaceE Δpqo Δpgi was performed as described previously for C. glutamicum Δpyc (Peters-Wendisch et al. 1998), using the suicide vector pK19mobsacB Δpyc. The deletion at the chromosomal locus was verified by PCR using primers pycfow1–pycrev1. Culture conditions and fermentations E. coli was grown aerobically in 2xTY complex medium at 37°C as 50-ml cultures in 500-ml baffled Erlenmeyer flasks on a rotary shaker at 120 rpm. Precultures of the different C. glutamicum strains were grown in 2xTY medium containing 28 mM glucose (corresponds to 0.5% (w/v) glucose) and 85 mM potassium acetate. The plasmidcarrying strains were grown in the presence of kanamycin

474

Appl Microbiol Biotechnol (2008) 79:471–479

Table 1 Strains, plasmids, and oligonucleotides used in this study Strain, plasmid, or oligonucleotide Strains E. coli DH5α C. glutamicum ΔaceE C. glutamicum ΔaceE Δpqo C. glutamicum ΔaceE Δpqo Δpgi C. glutamicum ΔaceE Δpqo Δpgi Δpyc Plasmids pK18mobsacB pK18mobsacB pgidel pK19mobsacB Δpyc pJC4ilvBNC pJC4ilvBNCD

pJC4ilvBNCE

Oligonucleotides pgi-d1 pgi-d2

Relevant characteristic(s) or sequence

Source–reference or purpose

supE44 hsdR17 recA1 endA1 gyrA96 thi-1 relA1 C. glutamicum wild type (ATCC13032) with deletion of the E1p gene (aceE) of the PDHC ΔaceE strain with deletion of the pyruvate:quinone oxidoreductase gene pqo ΔaceE Δpqo strain with deletion of the phosphoglucose isomerase gene pgi ΔaceE Δpqo Δpgi strain with deletion of the pyruvate carboxylase gene pyc

Hanahan 1985 Schreiner et al. 2005

Kmr, mobilizable (oriT), oriV pK18mobsacB carrying a truncated pgi gene (shortened by 1,504 bp) pK19mobsacB carrying a truncated pyc gene (shortened by 2,421 bp) Plasmid carrying the ilvBNC genes encoding the L-valine biosynthetic enzymes acetohydroxyacid synthase and isomeroreductase Plasmid carrying the ilvBNCD genes encoding the L-valine biosynthetic enzymes acetohydroxyacid synthase, isomeroreductase and dihydroxyacid dehydratase Plasmid carrying the ilvBNCE genes encoding the L-valine biosynthetic enzymes acetohydroxyacid synthase, isomeroreductase and transaminase B

Schäfer et al. 1994 This work Peters-Wendisch et al. 1998 Sahm et al. 1999

5′-ACGACCTCACCTACGGCGAA-3′ 5′-TGGGGTCCAAATGTCCCTGGGTGGTCGAAATGTCC-3′

Primer for deletion of pgi Primer for deletion of pgi, crossover overlap underlined Primer for deletion of pgi, crossover overlap underlined Primer for deletion of pgi Primer to verify pgi deletion Primer to verify pgi deletion Primer to verify pyc deletion Primer to verify pyc deletion

pgi-d3

5′-GGACATTTCGACCCCAAATGACCTCGCTCCGG CT-3′

pgi-d4 pgiout1 pgiout2 pycfow1 pycrev1

5′-CCCCACACCAACCGAAGACT-3′ 5′-GGAACGACACCAGATAAG-3′ 5′-CACTCATTGGTCGTGATG-3′ 5′-GCAGATGCCATTTACCCG-3′ 5′-CGGTGACAGACTCAACG-3′

(50 μg ml−1). Culture conditions for shake-flask amino acid fermentations and fed-batch fermentations in an 1-l glass bioreactor (Biostat B; Braun, Melsungen, Germany) were performed as described before (Blombach et al. 2007b).

Schreiner et al. 2006 This work This work

Sahm et al. 1999

Radmacher et al. 2002

concentrations were determined enzymatically according to Bergmeyer (1983).

Results Analytics production by C. glutamicum ΔaceE strains overexpressing different ilv genes

L-Valine

For quantification of substrate consumption and product formation, 1-ml samples were taken from the cultures and centrifuged at 13,000 rpm (10 min) and the supernatant was used for determination of amino acids, glucose, and/or organic acid concentrations in the culture fluid. The amino acid concentrations were determined by reversed-phase high-pressure liquid chromatography as described before (Blombach et al. 2007b). Glucose, acetate, and lactate concentrations were determined by enzymatic tests from Roche Diagnostics. The pyruvate

Previously, we demonstrated the ability of C. glutamicum ΔaceE to form pyruvate, L-alanine, and L-valine. Plasmidbound overexpression of the ilvBNCE genes in C. glutamicum ΔaceE directed the carbon flux from pyruvate and L-alanine to L-valine (Blombach et al. 2007b). To study the relevance of ilvD as compared to ilvE overexpression on L-valine production, plasmids pJC4ilvBNC, pJC4ilvBNCD, and pJC4ilvBNCE were transformed into C. glutamicum

Appl Microbiol Biotechnol (2008) 79:471–479

475

Relevance of pyruvate:quinone oxidoreductase for L-valine production with C. glutamicum ΔaceE Previously, we showed the suitability of C. glutamicum ΔaceE (pJC4ilvBNCE) for a L-valine production process by fed-batch fermentation (Blombach et al. 2007b). To study the relevance of the pyruvate:quinone oxidoreductase (PQO; pqo gene product) for L-valine accumulation, we transformed C. glutamicum ΔaceE Δpqo (Schreiner et al. 2006) with plasmid pJC4ilvBNCE and performed comparative fed-batch fermentations with C. glutamicum ΔaceE (pJC4ilvBNCE) and C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE). These fermentations were carried out in CGXII medium initially containing about 4% (w/v) glucose, 1.3% (w/v) acetate, and

60

250

To investigate the relevance of PGI and, thus, the influence of an increased NADPH supply for L-valine production, we constructed plasmid pK18mobsacB pgidel

40 150 30 100 20 50

10 0

0 0

10

20

30

60

250

50

200

40

Time [h]

Fig. 2 L-valine accumulation during representative fed-batch fermentations of a C. glutamicum ΔaceE (pJC4ilvBNCE) and b C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) in CGXII medium initially containing glucose, acetate and BHI. Empty diamond, growth; filled

Growth [OD600] Glucose [g/l] Acetate [g/l]

Growth [OD600] Glucose [g/l] Acetate [g/l]

50

Relevance of PGI for L-valine production with C. glutamicum ΔaceE Δpqo

b

Pyruvate [mM] L-Valine [mM]

a

0.5% (w/v) BHI. To allow growth to a high cell density, adequate amounts of a 50% (w/v) acetate stock solution were repeatedly added to the growing cells (Fig. 2a,b). Within 11 h, C. glutamicum ΔaceE (pJC4ilvBNCE) grew in a first exponential growth phase to an OD600 of about 39. In a second growth phase (t=9 h to t=28 h), C. glutamicum ΔaceE (pJC4ilvBNCE) showed linear growth up to an OD600 of 54 (Fig. 2a). In contrast C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) grew exponentially to an OD600 of 54 within 11 h and stopped to grow immediately after having consumed the last acetate pulse completely (Fig. 2b). With depletion of acetate, both strains started to excrete L-valine. As shown in Fig. 2a, C. glutamicum ΔaceE (pJC4ilvBNCE) accumulated about 195 mM L-valine within 25 h with a volumetric productivity of 7.9 mmol l−1 h−1 and a substratespecific product yield (YP/S) in the production phase of 0.39 mol L-valine per mol of glucose. C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) accumulated about 225 mM L-valine within 24 h (Fig. 2b). The volumetric productivity was 9.5 mmol l−1 h−1 with a YP/S in the production phase of 0.52 mol L-valine per mol of glucose. In addition to L-valine, both strains also excreted small amounts of pyruvate (about 5 mM) into the medium, indicating that L-valine production by both C. glutamicum ΔaceE (pJC4ilvBNCE) and C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) can be further increased. Taken together, these results show that the absence of PQO in C. glutamicum ΔaceE (pJC4ilvBNCE) resulted in about 30% increased YP/S in the production phase.

200

40 150 30 100 20

Pyruvate [mM] L-Valine [mM]

ΔaceE and shake-flask fermentations with the resulting strains carried out in CGXII medium with 0.5% (w/v) BHI, 4% (w/v) glucose, and 1% (w/v) acetate. All cultures grew within 8 h to an OD600 of about 20, consuming acetate completely and glucose to a minor part (decrease from 4.0% to 3.5%). The cells then stopped growing; however, they further consumed glucose and accumulated L-valine (data not shown). After complete consumption of the glucose (i.e., at t=72 h), C. glutamicum ΔaceE (pJC4ilvBNC) and C. glutamicum ΔaceE (pJC4ilvBNCD) produced 67±15 and 73±4 mM L-valine, respectively, whereas overexpression of the ilvBNCE genes resulted in a significantly higher L-valine accumulation of 106±9 mM L-valine. All three strains excreted additionally 20 to 30 mM pyruvate into the medium, indicating that there is still precursor available for further L-valine production. These results show that, aside from AHAS (ilvBN gene product) and AHAIR (ilvC gene product), the TA (ilvE gene product) might be a major bottleneck for efficient L-valine overproduction by C. glutamicum ΔaceE.

50

10 0

0 0

10

20

30

40

Time [h]

square, glucose; empty square, acetate; empty triangle, pyruvate; filled circle, L-valine. Three independent fed-batch fermentations were performed; all three showing comparable results

476

Appl Microbiol Biotechnol (2008) 79:471–479

and inactivated the chromosomal PGI gene (pgi) in C. glutamicum ΔaceE Δpqo. First, we compared the growth of C. glutamicum ΔaceE Δpqo with that of C. glutamicum ΔaceE Δpqo Δpgi in minimal medium containing either glucose plus acetate or fructose plus acetate. In minimal medium containing 0.5% (w/v) glucose and 0.5% (w/v) acetate, C. glutamicum ΔaceE Δpqo grew with a growth rate of 0.35 h−1 to an OD600 of about 9 after 24 h. C. glutamicum ΔaceE Δpqo Δpgi grew to nearly the same final OD600, however, showed a lower growth rate of 0.27 h−1 (data not shown). In minimal medium with 0.5% (w/v) fructose and 0.5% (w/v) acetate, C. glutamicum ΔaceE Δpqo grew with a slightly higher growth rate of 0.39 h−1 to an OD600 of about 9 after 24 h. In contrast, C. glutamicum ΔaceE Δpqo Δpgi showed only poor growth with a growth rate of 0.17 h−1 and a final OD600 of 4.5 after 24 h, indicating that PGI is essential for optimal growth of C. glutamicum ΔaceE Δpqo on fructose plus acetate. This phenotype was expected, as C. glutamicum possess no PGI isoenzyme (Marx et al. 2003) for the formation of glucose6-P during growth in minimal medium with fructose and acetate. For testing the relevance of PGI for L-valine production, C. glutamicum ΔaceE Δpqo Δpgi was transformed with plasmid pJC4ilvBNCE and fed-batch fermentations with the resulting strain were carried out in CGXII medium containing 0.5% (w/v) BHI, 4.5% glucose, and an acetate concentration of 1.5% (w/v). To allow growth to a high cell density, adequate amounts of a 50% (w/v) acetate stock solution were added twice to the growing cells (Fig. 3). Using this technique, an OD600 of about 37 was obtained after 15 h (Fig. 3). After having consumed the last pulse of acetate, the cells started to excrete L-valine and 15, 29.5, and 49 h after inoculation, we added glucose again to obtain concentrations of about 5% (w/v). As shown in Fig. 3, C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) accumulated about 412 mM L-valine within 74 h with a volumetric productivity

Relevance of PCx for L-valine production with C. glutamicum ΔaceE Δpqo Δpgi The observation that C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) did not secrete pyruvate into the culture broth suggested effective C-flux from pyruvate to acetolactate (and further to L-valine) and, thus, that in this strain the availability of pyruvate for L-valine production might become limiting. To test for this hypothesis and to possibly increase the pyruvate availability, we deleted the PCx gene ( pyc) in C. glutamicum ΔaceE Δpqo Δpgi and transformed the resulting strain C. glutamicum ΔaceE Δpqo Δpgi Δpyc with plasmid pJC4ilvBNCE. For L-valine accumulation we carried out fed-batch fermentations in CGXII medium as described above for C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE). After 18 h, having consumed the second of two acetate pulses completely, C. glutamicum ΔaceE Δpqo Δpgi Δpyc (pJC4ilvBNCE) reached an OD600 of about 36 and started to form L-valine from glucose. Within 46 h, C. glutamicum ΔaceE Δpqo Δpgi Δpyc (pJC4ilvBNCE) then produced 240 mM Lvaline with a volumetric productivity of 5.2 mmol l−1 h−1 and a YP/S in the production phase of 0.86 mol L-valine per mol (0.56 g/g) of glucose. Neither acetate nor pyruvate was detectable in the growth medium. Comparison of L-valine accumulation and substrate-specific L-valine yields Figure 4 summarizes L-valine accumulation and substratespecific L-valine yields of the production phases of the C. glutamicum ΔaceE derivatives. C. glutamicum ΔaceE (pJC4ilvBNCE) accumulated about 200 mM L-valine with 450

60

400 50

350 300

40

250 30

200

20

150 100

10

50 0

0 0

10

20

30

40

50

Time [h]

60

70

80

90

Pyruvate [mM] L-Valine [mM]

Growth [OD600] Glucose [g/l] Acetate [g/l]

Fig. 3 L-valine accumulation during a representative fedbatch fermentation of C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) in CGXII medium initially containing glucose, acetate, and BHI. Empty diamond, growth; filled square, glucose; empty square, acetate; empty triangle, pyruvate; filled circle, L-valine. Two independent fed-batch fermentations were performed, both showing comparable results

of 5.6 mmol l−1 h−1 and a YP/S in the production phase of 0.75 mol L-valine per mol of glucose. No pyruvate or any other of the tested byproducts were excreted into the medium.

477

450

1.0

400

0.9

350

0.8 0.7

300

0.6 250

0.5 200

0.4

150

YP/S [mol/mol]

L-Valine [mM]

Appl Microbiol Biotechnol (2008) 79:471–479

0.3

100

0.2

50

0.1 0.0

0

A

B

C

D

Fig. 4 L-valine accumulation (black bars) and substrate-specific product yields (YP/S in mol L-valine per mole glucose; grey bars) at the end of the production phases of representative fed-batch fermentations of (A) C. glutamicum ΔaceE (pJC4ilvBNCE), (B) C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE), (C) C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE), and (D) C. glutamicum ΔaceE Δpqo Δpgi Δpyc (pJC4ilvBNCE) in CGXII medium containing glucose, acetate, and BHI. Means are from at least two independent experiments

a YP/S of 0.39 mol L-valine per mol glucose in the production phase. PQO deficiency in this strain resulted in an about 13% higher L-valine accumulation and an about 30% increased YP/S. Additional deletion of the PGI gene resulted in a drastic increase of L-valine accumulation (more than 400 mM) and a YP/S of 0.75 mol L-valine per mol of glucose. Subsequent inactivation of the PCx in C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) resulted in the maximal theoretical YP/S of 0.86 mol L-valine per mol glucose however a lower accumulation of about 240 mM Lvaline (Fig. 4). Taken together, these results show that pyruvate supply on the one hand, an increased NADPH availability on the other and, finally increasing the intracellular content of L-valine biosynthetic enzymes AHAS, AHAIR, and TA are highly beneficial for L-valine overproduction with C. glutamicum.

Discussion Radmacher et al. (2002) showed that overexpression of ilvBNC in combination with ilvD or ilvE in strains with deleted ilvA and/or panBC (see “Introduction”) resulted in increased L-valine accumulation. Our findings here corroborate the importance of increasing the intracellular content of L-valine biosynthetic enzymes for L-valine production with C. glutamicum. Plasmid-bound expression of ilvBNC, ilvBNCD, or ilvBNCE shifted the product spectrum of C. glutamicum ΔaceE significantly from pyruvate and Lalanine towards L-valine. However, the fermentations with C. glutamicum ΔaceE harboring each of the three different plasmids revealed differences in L-valine accumulation and identified, aside from AHAS and AHAIR, the TA enzyme as

more important than DHAD for efficient L-valine production. This result is surprising because in C. glutamicum ΔilvA ΔpanBC the overexpression of ilvBNCD resulted in about 10% more L-valine than overexpression of ilvBNCE (91.9 vs. 81.2 mM; Radmacher et al. 2002). However, our results underline the importance of increasing TA activity for Lvaline production with a PDHC-deficient C. glutamicum strain and based on these findings, C. glutamicum ΔaceE (pJC4ilvBNCE) represented the ideal platform for further strain improvement. Previously, Schreiner et al. (2006) observed no significant differences in growth of C. glutamicum ΔaceE and C. glutamicum ΔaceE Δpqo under several conditions in shake flasks. These results indicated that PQO is not relevant for growth of these strains under the given conditions. However, the authors also showed that overexpression of pqo in C. glutamicum ΔaceE resulted in linear growth in complex medium without acetate, indicating that PQO activity at least partially can compensate for PDHC activity in C. glutamicum (Schreiner et al. 2006). We here show that C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) stopped growing immediately after consumption of the last pulse of acetate whereas C. glutamicum ΔaceE (pJC4ilvBNCE) still showed linear growth after acetate was completely consumed. These results indicate that PQO and the acetate-activating enzymes acetate kinase and phosphotransacetylase can apparently catalyze the formation of acetyl-CoA to bypass the PDHC reaction and so contribute to growth of C. glutamicum ΔaceE (pJC4ilvBNCE) at high cell densities. Hence, inactivation of the PQO avoided withdrawal of pyruvate for growth purposes and resulted in an about 30% increased YP/S. In fed-batch fermentations, both C. glutamicum ΔaceE (pJC4ilvBNCE) and C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) still excreted pyruvate into the medium, indicating that L-valine production in these strains is limited by the reactions from pyruvate to L-valine. Stoichiometric network modeling and computational analysis of the central metabolism and of the L-valine biosynthetic pathway in C. glutamicum indicated that NADPH supply is critical for Lvaline production from glucose (Bartek et al. 2008a). The modeling predicted an optimal NADPH supply and a maximal YP/S of 0.86 mol L-valine per mol of glucose when the carbon flux from glucose is completely directed into the pentose phosphate pathway instead of into glycolysis (Bartek et al. 2008a). In accordance, the fedbatch fermentations with C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) revealed a higher conversion of glucose to L-valine with a YP/S of 0.75 mol L-valine per mol of glucose when compared to the respective strain with intact PGI. Furthermore, the PGI-deficient strain produced up to 400 mM L-valine and excreted no pyruvate into the medium. These results indicated that L-valine production

478

with C. glutamicum ΔaceE Δpqo (pJC4ilvBNCE) was limited in NADPH supply and not by the content of the enzymes of the L-valine biosynthetic pathway. Supporting our results, Bartek et al. (2008a) showed that the concentration of NADPH in C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) was much higher than in C. glutamicum ΔaceE (pJC4ilvBNCE; i.e., 412 vs. 58 μM). NADPH supply was previously also discussed to be a critical factor for L-lysine production with C. glutamicum (Becker et al. 2005; Kabus et al. 2007; Georgi et al. 2005) and Marx et al. (2003) reported about a PGI-deficient mutant of a L-lysine producer, which showed increased and more efficient Llysine formation than the respective strain with intact PGI. The observation that C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) excreted no pyruvate into the medium prompted us to test for the relevance of PCx for L-valine production. C. glutamicum ΔaceE Δpqo Δpgi Δpyc (pJC4ilvBNCE) reached in fed-batch fermentations the maximal theoretical YP/S of 0.86 mol L-valine per mol of glucose. Although no data about the intracellular content of pyruvate for C. glutamicum ΔaceE Δpqo Δpgi Δpyc (pJC4ilvBNCE) were available, the about 15% higher YP/S compared to C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE) shows that inactivation of the PCx is beneficial for effective L-valine production. It is noteworthy to mention that the given YP/S values for all fermentations presented here (Fig. 4) correspond to the production phase in the absence of cellular growth. Because all strains did not produce L-valine in the growth phase, the overall YP/S values are lower (e.g., 0.36 vs. 0.61 mol-C/mol-C for C. glutamicum ΔaceE Δpqo Δpgi (pJC4ilvBNCE)), due to consumption of acetate and glucose for the formation of biomass. The finding that all engineered C. glutamicum ΔaceE derivatives showed Lvaline production only in a growth-decoupled manner (see also Blombach et al. 2007b) suggests a further potential to improve L-valine production by overcoming the regulatory and/or metabolic reasons for the nonproduction phenotype in the growth phase at the beginning of the fermentations. Acknowledgement We thank Lothar Eggeling for providing plasmids pJC4ilvBNC, pJC4ilvBNCD, and pJC4ilvBNCE and Brigitte Bathe (Evonik Degussa) for providing plasmid pK18mobsacB pgidel. We are grateful to Andreas Karau and Robert Gerstmeir (Degussa AG) for valuable discussions. We thank Konstanze Fleischer for technical assistance. The support of the Fachagentur Nachwachsende Rohstoffe of the BMVEL (grant 04NR004/22000404) is gratefully acknowledged.

References Bartek T, Blombach B, Zönnchen E, Makus P, Wahl A, Lang S, Eikmanns BJ, Oldiges M (2008a) Importance of NADPH supply for improved L-valine formation based on stoichiometric modelling. Submitted to AMB

Appl Microbiol Biotechnol (2008) 79:471–479 Bartek T, Makus P, Klein B, Lang S, Oldiges M (2008b) Influence of L-isoleucine and pantothenate auxotrophy for l-valine formation in Corynebacterium glutamicum revisited by metabolome analyses. Bioprocess Biosyst Eng DOI 10.1007/s00449-008-0202-z Becker J, Klopprogge C, Zelder O, Heinzle E, Wittmann C (2005) Amplified expression of fructose 1,6-bisphosphatase in Corynebacterium glutamicum increases in vivo flux through the pentose phosphate pathway and lysine production on different carbon sources. Appl Environ Microbiol 71:8587–8596 Bergmeyer HU (1983) Methods of enzymatic analysis vol. VI, 3rd edn. Verlag Chemie, Weinheim, pp 59–66 Blombach B, Schreiner ME, Moch M, Oldiges M, Eikmanns BJ (2007a) Effect of pyruvate dehydrogenase complex deficiency on L-lysine production with Corynebacterium glutamicum. Appl Microbiol Biotechnol 76:615–623 Blombach B, Schreiner ME, Holatko J, Bartek T, Oldiges M, Eikmanns BJ (2007b) L-Valine production with pyruvate dehydrogenase complex-deficient Corynebacterium glutamicum. Appl Environ Microbiol 73:2079–2084 Dower WJ, Miller JF, Ragsdale CW (1988) High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res 16:6127–6145 Eggeling L (2001) Amino acids. In: Ratledge C, Kristiansen B (eds) Basic biotechnology. Cambridge University Press, London, pp 281–303 Eggeling L, Bott M (2005) Handbook of Corynebacterium glutamicum. CRC, Boca Raton Eikmanns BJ, Thum-Schmitz N, Eggeling L, Lüdtke KU, Sahm H (1994) Nucleotide sequence, expression and transcriptional analysis of the Corynebacterium glutamicum gltA gene encoding citrate synthase. Microbiology 140:1817–1828 Eikmanns BJ, Rittmann D, Sahm H (1995) Cloning, sequence analysis, expression, and inactivation of the Corynebacterium glutamicum icd gene encoding isocitrate dehydrogenase and biochemical characterization of the enzyme. J Bacteriol 177:774–782 Elisakova V, Patek M, Holatko J, Nesvera J, Leyval D, Goergen JL, Delaunay S (2005) Feedback-resistant acetohydroxyacid synthase increases valine production in Corynebacterium glutamicum. Appl Environ Microbiol 71:207–213 Georgi T, Rittmann D, Wendisch VF (2005) Lysine and glutamate production by Corynebacterium glutamicum on glucose, fructose and sucrose: roles of malic enzyme and fructose-1,6-bisphosphatase. Metab Eng 7:291–301 Gourdon P, Baucher MF, Lindley ND, Guyonvarch A (2000) Cloning of the malic enzyme gene from Corynebacterium glutamicum and role of the enzyme in lactate metabolism. Appl Environ Microbiol 66:2981–2987 Hanahan D (1985) Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166:557–580 Kabus A, Georgi T, Wendisch VF, Bott M (2007) Expression of the Escherichia coli pntAB genes encoding a membrane-bound transhydrogenase in Corynebacterium glutamicum improves L-lysine. Appl Microbiol Biotechnol 75:47–53 Lange C, Rittmann D, Wendisch VF, Bott M, Sahm H (2003) Global expression profiling and physiological characterization of Corynebacterium glutamicum grown in the presence of Lvaline. Appl Environ Microbiol 69:2521–2532 Leuchtenberger W (1996) Amino acids—technical production and use. In: Rehm H-J, Reed G, Pühler A, Stadler P (eds) Biotechnology, vol. vol. 6. VCH, Weinheim, pp 465–502 Leuchtenberger W, Huthmacher K, Drauz K (2005) Biotechnological production of amino acids and derivates: current status and prospects. Appl Microbiol Biotechnol 69:1–8 Liebl W (1991) The genus Corynebacterium—nonmedical. In: Balows A, Trüper HG, Dworkin M, Harder W, Schleifer KH (eds) The Prokaryotes, vol. 2, Springer, New York, pp 1157–1171

Appl Microbiol Biotechnol (2008) 79:471–479 Marienhagen J, Kennerknecht N, Sahm H, Eggeling L (2005) Functional analysis of all aminotransferase proteins inferred from the genome sequence of Corynebacterium glutamicum. J Bacteriol 187:7639–7646 Marx A, de Graaf AA, Wiechert W, Eggeling L, Sahm H (1996) Determination of the fluxes in the central metabolism of Corynebacterium glutamicum by nuclear magnetic resonance spectroscopy combined with metabolite balancing. Biotechnol Bioeng 49:111–129 Marx A, Striegel K, de Graaf AA, Eggeling L (1997) Response of central metabolism of Corynebacterium glutamicum to different flux burdens. Biotechnol Bioeng 56:168–180 Marx A, Eikmanns BJ, Sahm H, de Graaf AA, Eggeling L (1999) Response of central metabolism of Corynebacterium glutamicum to the use of an NADH-dependent glutamate dehydrogenase. Metab Eng 1:35–48 Marx A, Hans S, Mockel B, Bathe B, de Graaf AA (2003) Metabolic phenotype of phosphoglucose isomerase mutants of Corynebacterium glutamicum. J Biotechnol 104:185–197 Moritz B, Striegel K, de Graaf AA, Sahm H (2000) Kinetic properties of the glucose-6-phosphate and 6-phosphogluconate dehydrogenases from Corynebacterium glutamicum and their application for predicting pentose pathway flux in vivo. Eur J Biochem 267:3442–3452 Ohnishi J, Katahira R, Mitsuhashi S, Kakita S, Ikeda M (2005) A novel gnd mutation leading to increased L-lysine production in Corynebacterium glutamicum. FEMS Microbiol Lett 242:265–274 Park JH, Lee KH, Kim TY, Lee SY (2007) Metabolic engineering of Escherichia coli for the production of L-valine based on transcriptome analysis and in silico knockout simulation. Proc Natl Acad Sci U S A 104:7797–7802 Patek M (2007) Branched-chain amino acids. In: Steinbüchel A (ed) Microbiol monographs, vol. 5, Springer, Berlin/Heidelberg, pp 129–162 Peters-Wendisch PG, Kreutzer C, Kalinowski J, Pátek M, Sahm H, Eikmanns BJ (1998) Pyruvate carboxylase from Corynebacterium glutamicum: characterization, expression and inactivation of the pyc gene. Microbiology 144:915–927 Peters-Wendisch PG, Schiel B, Wendisch VF, Katsoulidis E, Mockel B, Sahm H, Eikmanns BJ (2001) Pyruvate carboxylase is a major bottleneck for glutamate and lysine production by Corynebacterium glutamicum. J Mol Microbiol Biotechnol 3:295–300

479 Radmacher E, Vaitsikova A, Burger U, Krumbach K, Sahm H, Eggeling L (2002) Linking central metabolism with increased pathway flux: L-valine accumulation by Corynebacterium glutamicum. Appl Environ Microbiol 68:2246–2250 Riedel C, Rittmann D, Dangel P, Möckel B, Sahm H, Eikmanns BJ (2001) Characterization, expression, and inactivation of the phosphoenolpyruvate carboxykinase gene from Corynebacterium glutamicum and significance of the enzyme for growth and amino acid production. J Mol Microbiol Biotechnol 3:573–583 Ruklisha M, Paegle L, Denina I (2007) L-Valine biosynthesis during batch and fed-batch cultivations of Corynebacterium glutamicum: relationship between changes in bacterial growth rate and intracellular metabolism. Process Biochem 42:634–640 Sahm H, Eggeling L (1999) D-Pantothenate synthesis in Corynebacterium glutamicum and use of panBC and genes encoding L-valine synthesis for D-pantothenate overproduction. Appl Environ Microbiol 65:1973–1979 Sambrook J, Russel DW, Irwin N, Janssen UA (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor Schäfer A, Tauch A, Jäger W, Kalinowski J, Thierbach G, Pühler A (1994) Small mobilizable multi-purpose cloning vectors derived from the E. coli plasmids pK18 and pK19: selection of defined deletions in the chromosome of Corynebacterium glutamicum. Gene 145:69–73 Schreiner ME, Fiur D, Holátko J, Pátek M, Eikmanns BJ (2005) E1 enzyme of the pyruvate dehydrogenase complex in Corynebacterium glutamicum: Molecular analysis of the gene and phylogenetic aspects. J Bacteriol 187:6005–6018 Schreiner ME, Riedel C, Holatko J, Patek M, Eikmanns BJ (2006) Pyruvate:quinone oxidoreductase in Corynebacterium glutamicum: molecular analysis of the pqo gene, significance of the enzyme, and phylogenetic aspects. J Bacteriol 188: 1341–1350 Takors R, Bathe B, Rieping M, Hans S, Kelle R, Huthmacher K (2007) Systems biology for industrial strains and fermentation processes—example: amino acids. J Biotechnol 129:181–190 van der Rest ME, Lange C, Molenaar D (1999) A heat shock following electroporation induces highly efficient transformation of Corynebacterium glutamicum with xenogenic plasmid DNA. Appl Microbiol Biotechnol 52:541–545