Appl Microbiol Biotechnol (1999) 51: 325333

Springer-Verlag 1999

ORIGINAL PAPER

C. A. Shaw-Reid M. M. McCormick A. J. Sinskey G. Stephanopoulos

Flux through the tetrahydrodipicolinate succinylase pathway is dispensable for L-lysine production in Corynebacterium glutamicum

Received: 20 May 1998 / Received revision: 12 August 1998 / Accepted: 3 September 1998

Abstract The N-succinyl-LL-diaminopimelate desuccinylase gene (dapE) in the four-step succinylase branch of the L-lysine biosynthetic pathway of Corynebacterium glutamicum was disrupted via marker-exchange mutagenesis to create a mutant strain that uses only the onestep meso-diaminopimelate dehydrogenase branch to overproduce lysine. This mutant strain grew and utilized glucose from minimal medium at the same rate as the parental strain. In addition, the dapE ) strain produced lysine at the same rate as its parent strain. Transformation of the parental and dapE ) strains with the amplied meso-diaminopimelate dehydrogenase gene (ddh) on a plasmid did not aect lysine production in either strain, despite an eightfold amplication of the activity of the enzyme. These results indicate that the four-step succinylase pathway is dispensable for lysine overproduction in shake-ask culture. In addition, the one-step meso-diaminopimelate dehydrogenase pathway does not limit lysine ux in Corynebacterium under these conditions.

Introduction
The gram-positive microorganism Corynebacterium glutamicum is widely used for the overproduction of amino acids such as L-lysine. The development of a geC. A. Shaw-Reid1 G. Stephanopoulos (&) Department of Chemical Engineering, Room 56-469, Massachusetts Institute of Technology, Cambridge, MA 02139 USA Tel.: +1-617-253-4583 Fax: +1-617-253-3122 M. M. McCormick2 A. J. Sinskey Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA Present address: Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood Avenue, Boston, MA 02115, USA 2 McKinsey and Co., Houston, TX, 77010, USA
1

netic toolbox to study this bacterium has been progressing rapidly (Jetten and Sinskey 1995), but the interaction of many intermediate biochemical reactions in the lysine biosynthetic network is still not well understood. Product yield maximization requires optimal coordination of the C. glutamicum metabolic network. To this end, experiments were conducted to study the eect of biochemical variations in the reactions of the tetrahydrodipicolinate branchpoint on lysine ux. At the tetrahydrodipicolinate branchpoint of the L-lysine pathway of C. glutamicum, carbon can be shunted either through the one-step meso-diaminopimelate (Dap) dehydrogenase pathway (Misono et al. 1979), or through the four-step succinylase pathway (Yeh et al. 1988; Schrumpf et al. 1991) to form the cellwall component and immediate precursor to lysine, meso-Dap (Fig. 1). To assess the role of the dual pathway at this branchpoint, marker-exchange mutagenesis was employed to create a mutant strain that lacks the dapE gene and consequently is unable to synthesize Nsuccinyl-LL-Dap desuccinylase, the third enzyme in the four-step succinylase pathway. This mutant strain utilizes only the one-step meso-Dap dehydrogenase pathway for lysine synthesis. The dapE ) mutant was further characterized by determining pathway uxes when the strain was transformed with a multicopy plasmid, which amplied the ddh gene product, meso-Dap dehydrogenase. Although disruption of the meso-Dap dehydrogenase (ddh) gene in the one-step pathway has been characterized (Schrumpf et al. 1991), the lysine productivity of a strain harboring a disrupted gene in the four-step succinylase pathway has not yet been studied. The tetrahydrodipicolinate succinylase (dapD) and dapE genes in the four-step pathway have been cloned in Corynebacterium, and disruption of these genes has been shown to alter cell morphology (Wehrmann et al. 1998, 1994). In amplication experiments this group was able to achieve tenfold overexpression of the desuccinylase enzyme, but no data were provided on how its overexpression aected lysine ux (Wehrmann et al. 1994). The

326

Fig. 1 Genes and corresponding enzymatic reactions at the tetrahydrodipicolinate branchpoint in the L-lysine pathway of Corynebacterium glutamicum. Derivatives of the precursors pyruvate and oxaloacetate condense to form the branchpoint intermediate tetrahydrodipicolinate. The one-step pathway transforms tetrahydrodipicolinate to meso-diaminopimelate (meso-DAP) via the activity of (1) meso-a-e-diaminopimelate D-dehydrogenase (DDH; EC 1.4.1.16) encoded by the ddh gene. The four-step succinylase pathway contains the following enzymes: (2) tetrahydrodipicolinate N-succinyltransfer-

ase (DAPD; EC 2.3.1.89) encoded by the dapD gene, (3) N-succinylLL-diaminopimelate aminotransferase (DAPC; EC 2.6.1.17) encoded by the dapC gene, (4) N-succinyl-LL-diaminopimelate desuccinylase (DAPE; EC 3.5.1.18) encoded by the dapE gene, and (5) LLdiaminopimelate epimerase (DAPF; EC 5.1.1.7) encoded by the dapF gene. The common product of the dual pathway, meso-diaminopimelate, is transformed to L-lysine by the activity of (6) meso-a-ediaminopimelate decarboxylase (LYSA; EC 4.1.1.20) encoded by the lysA gene

327

dapE gene was chosen for disruption in this report because of evidence suggesting that it may have a ratecontrolling role in the four-step pathway (see below). N-Succinyl-Dap desuccinylase has been puried to homogeneity in Escherichia coli (Lin et al. 1988). The dapE gene has been compared to the argE gene, which encodes acetylornithine deacetylase in the ornithine biosynthesis pathway of E. coli, and it is hypothesized that these deacylases may have a common evolutionary origin because of similarities in their amino acid sequences (26% homology) and because both gene products require a cobalt cofactor (Boyen et al. 1992). Wu et al. (1992) found that the msgB gene, which encodes a protein that suppresses the temperature-sensitive grpE heat-shock gene, has complete amino acid sequence homology to the E. coli dapE gene. There has been no analogous research published on either the argE or msgB genes in Corynebacterium. The meso-Dap dehydrogenase (ddh) gene in the onestep pathway has been cloned and sequenced in C. glutamicum (Ishino et al. 1987; Yeh et al. 1988). Previous experiments (Yeh et al. 1988; Cremer et al. 1991) have shown that overexpression of the ddh gene on a plasmid results in 18- to 20-fold amplication of meso-Dap dehydrogenase activity, but the increase in activity of this enzyme has no eect on lysine production in wild-type Corynebacterium containing both the succinylase and meso-Dap dehydrogenase pathways. The eect of mesoDap dehydrogenase amplication on lysine production in a Corynebacterium strain containing only the mesoDap dehydrogenase pathway has not been described. The present study extends the characterization of lysineproduction pathways by investigating the eect of varying meso-Dap dehydrogenase activity in dapEdisruption strains that cannot utilize the succinylase pathway to produce lysine. Previous research indicates that when the ddh gene is disrupted in C. glutamicum, cell growth is not aected, but the lysine titer decreases two- to fourfold (Schrumpf et al. 1991). Schrumpf et al. also found that ddh) mutants accumulated up to 25 mM cytosolic N-succinyl-LLDap, while the wild type only accumulated trace amounts. This observation led to the hypothesis that N-succinyl-Dap desuccinylase activity in the four-step pathway may limit lysine production. A subsequent publication (Sonntag et al. 1993) with the same ddh) strain revealed that lysine production varied signicantly with culture conditions. The accumulation of intracellular N-succinyl-Dap is not unique to the C. glutamicum ddh) strain. Gilvarg (1959) found that N-succinyl-Dap accumulates in a Dap-requiring mutant of E. coli. In addition, Sundharadas and Gilvarg (1967) found that Dap auxotrophs of Bacillus megaterium accumulate N-acetyl-Dap in their cytosol. Both E. coli and B. megaterium utilize only the four-step pathway to synthesize lysine. These results support the hypothesis that the N-succinyl-LLDap desuccinylase reaction may be a rate-controlling step in the four-step pathway to meso-Dap and lysine

biosynthesis, but they do not imply that the four-step pathway limits lysine production when a cell also has access to the one-step meso-Dap dehydrogenase pathway. The objectives of this work were to elucidate the dependence of lysine synthesis on the two pathways by constructing dapE-disruption mutant strains and establishing the eect of dapE disruption on cellular growth, glucose uptake and lysine production. The phenotype of the dapE-disruption mutant was characterized in the presence of varying levels of meso-Dap dehydrogenase activity. Our results indicate that the dapE gene is dispensable for lysine overproduction in shake-ask cultures, and amplication of the ddh gene has no eect on lysine production in Corynebacterium that utilizes only the one-step meso-Dap dehydrogenase pathway.

Materials and methods

Bacterial strains, plasmids and culture conditions The strains and plasmids used or constructed in this study are listed in Table 1. This parental strain, C. glutamicum ATCC 21253, is a homoserine, or threonine plus methionine, auxotrophic strain that overproduces and excretes lysine after the depletion of threonine from the extracellular medium. C. glutamicum was routinely grown at 30 C in LB (per liter: 10 g tryptone, 5 g NaCl, and 5 g yeast extract) supplemented with 5 g D-glucose/l. Minimal medium modified from Kiss and Stephanopoulos (1991) contained, per liter, 20 g D-glucose, 200 mg MgSO4 7H2O, 1 g NaCl, 75 mg Na2 EDTA 2H2O, 25 mg FeSO4 7H2O, 50 mg CaCl2 2H2O, 1.2 g trisodium citrate 2H2O, 10 ml trace mineral salts, 8 g K2HPO4, 1 g KH2PO4, 150 mg L-threonine, 80 mg L-methionine, 100 mg L-leucine, 15 g (NH4)2SO4, 1 mg biotin, and 1 mg thiamine HCl. The 100 trace mineral salts solution contained, per liter, 200 mg MnSO4 H2O, 200 mg FeCl3 6H2O, 50 mg ZnSO4 7H2O, 20 mg Na2B4O7 10H2O, 20 mg CuCl2 2H2O, and 10 mg (NH4)6Mo7O24 4H2O. Cells used for enzyme assays were grown to mid-exponential phase in 50100 ml of either complex or minimal medium. Growth and amino acid production studies were performed exclusively in 50 ml minimal medium. E. coli strains were grown in LB at 37 C. When necessary, 50 mg/l ampicillin, 25 mg/l kanamycin, or 20 mg/l chloramphenicol antibiotics were added to the medium for selection pressure. DNA techniques Standard protocols (Sambrook et al. 1989) were used for the construction, purication, and analysis of plasmid DNA. Restriction enzymes, T4 DNA ligase, and other reagents were purchased from New England Biolabs or Boehringer Mannheim. DNA fragments were isolated from 1% agarose gels using the QIAEX II Gel Extraction Kit for DNA purication (Qiagen). Corynebacterium was transformed via electroporation as described in Jetten et al. (1995). The polymerase chain reaction (PCR) was used to amplify the N-succinyl-LL-Dap desuccinylase (dapE) gene as a 1.4-kb fragment of DNA. Flanking primers (Gibco BRL) were selected on the basis of the sequence published by Wehrmann et al. (1995). The upstream primer sequence was 5-CTGCAGCGATATTTGCGATTCCAACTGCT-3, while the downstream primer sequence was 5-CTGCAGCGCATTTCTAGCAAGCGT-3. An articial PstI site (in bold type) was inserted at the 5-terminal ends of both primers to facilitate subsequent cloning steps. The amplied dapE

328 Table 1 Bacterial strains and plasmids. ATCC American Type Culture Collection, Rockville, Md, USA; BGSC Bacillus Genetic Stock Center, Columbus, Ohio, USA. Kmr, Apr, Cmr, Specr, Rf Strain or plasmid Corynebacterium glutamicum 21253 212DAPEM ASO19-E12 Escherichia coli DH5a S17-1 Bacillus sphaericus 13A5 Plasmids pBluescript KS pPCR-Script pSUP301 pMT1 pMG110 pAL214 pDDHHind pCSR1 pDapEScript pDapESm pDapSupSm pDapCmSup Relevant characteristics
L-Lysine-overproducing )

resistance against kanamycin, ampicillin, chloramphenicol, spectinomycin and rifampicin respectively Source or reference ATCC This work Follettie 1989 Hanahan, 1983 Schafer et al. 1990 BGSC Stratagene Stratagene Schafer et al. 1990 Follettie et al. 1993 Gubler et al. 1994 This work This work This work This work This work This work This work

strain, hom) dapE derivative of 21253, Cmr Restriction-decient and Rfr derivative of ATCC 13059

lacZ DM15, recA), hsdR) hsdR, recA, tra from RP4-2 integrated in chromosome ddh+, dapE ) Cloning vector, Apr Cloning vector, Apr pACY177 derivative, mob, Apr, Kmr E. coli (Apr)/C. glutamicum (Kmr) shuttle vector pUC4 derivative, Apr, pneo::cat Cmr pBluescript KS derivative, Specr, Apr pBluescript KS derivative, ddh+, Apr pMT1 derivative containing ddh from pDDHHind pCR-Script derivative, dapE ), Apr pDapEScript derivative, dapE ), Apr, Specr pSUP301 derivative, dapE ), Specr pSUP301 derivative, dapE ), Cmr

PCR product was inserted into an Srf I-digested pCR-Script cloning vector according to the ligation protocol outlined in the Stratagene pCR-Script Amp SK(+) Cloning Kit (Stratagene Cloning Systems, La Jolla, Calif.). The resulting 4.4-kb plasmid was named pDapEScript. A similar PCR strategy was used to amplify the meso-Dap dehydrogenase (ddh) gene as a 1.8-kb fragment of DNA. The following primers were designed for the ddh gene amplication using the sequence published by Ishino et al. (1987): 5CCTCTTGCTGCAGCGAAGACAC-3 and 5-CGGGATTTTCTGCAGCCCAGGT-3. The amplied ddh PCR product was inserted via PstI digestion into a pBluescript vector that was modied by removal of the HindIII site (pDDHHind). PCR amplication was performed using the TwinBlock System microprocessor-controlled temperature cycler. The 100-ll reaction mixtures contained 10 mM TRIS/HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.01% gelatin, 100 pmol primers, 20 nmol each deoxynucleoside triphosphate, 2.5 units Taq DNA polymerase (Boehringer Mannheim), and 1 lg genomic template DNA. The QIAamp Tissue Kit for chromosomal DNA preparation (Qiagen) was used to isolate C. glutamicum genomic DNA from 5-ml overnight cultures. After a hot start at 94 C for 1 min, amplication was performed during 30 cycles of denaturation at 94 C for 1 min, annealing at 58 C (dapE) or 68 C (ddh) for 2 min, and extension at 72 C for 3 min. The nal cycle extension time at 72 C was 10 min. The strategy for the construction of the mobilizable plasmid pDapCmSup used for disruption of the dapE gene is given in the legend of Fig. 2. Mobilization of the gene-disruption plamid pDapCmSup from E. coli to C. glutamicum via marker-exchange mutagenesis was performed as described previously (Schafer et al. 1990; Schwarzer and Puhler 1991; Gubler et al. 1994). The donor:recipient ratio was 1:1, and the recipients were heat-treated at 49 C for 9 min. C. glutamicum transconjugants were selected on LB/agar containing, per liter, 25 mg nalidixic acid and 10 mg chloramphenicol. Positive double-crossover candidates (dapE)) were distinguished from single-crossovers (dapE+) by their inability to grow on 6 mg/l kanamycin. Transconjugants were further veried by Southern blot analysis using the Genius System from Boehringer Mannheim following their Users' Guide for lter hybridization. The plasmid pCSR1 contains the intact meso-diaminopimelate dehydrogenase (ddh) gene on a vector that replicates in Coryne-

bacterium. This plasmid was constructed by excising the PCRamplied ddh gene from the pBluescript-derived vector pDDHHind and inserting the gene into the KpnI-XbaI-digested E. coli-C. glutamicum shuttle vector, pMT1. Enzyme assays Assays were carried out with cell-free extracts, which were prepared as described by Jetten and Sinskey (1993), with the exception that the homogenization buer consisted of 100 mM HEPES (pH 7.8), 200 mM KCl and 1 mM 2-mercaptoethanol was added to protect sulfhydryl enzymes. The protein concentration of the extracts was determined by the method of Bradford (1976). All enzyme activities were determined at room temperature in a Hewlett-Packard 8452A diode-array spectrophotometer. The chemicals used for enzymatic assays were purchased from Sigma Chemical Co., with the exception of the substrate N-succinyl-LL-Dap. This was synthesized by anhydride acylation and partially purified by ion-exchange chromatography according to the method described by Bouvier et al. (1992). The enzymatic activity of N-succinyl-LL-Dap desuccinylase (EC 3.5.1.18) was measured with a ninhydrin assay as described by Kindler and Gilvarg (1960). The assay reaction mixtures contained 10.4 lmol N-succinyl-LL-Dap, 1 lmol CoCl2, 20 lmol NH4OH, 200 mM TRIS/HCl buer at pH 8.0, and crude extract in a total volume of 550 ll. The reaction was initiated with crude extract and incubated at 37 C. The ninhydrin assay was carried out on initial samples at t 0 and samples taken after 30 min as a nal assay time. The oxidative deamination enzymatic activity of the reversible meso-Dap dehydrogenase (EC 1.4.1.16) was determined by measuring the increase in absorbance of NADPH at 340 nm as meso-Dap is converted to tetrahydrodipicolinate. The reaction conditions were those described by Misono et al. (1979). Determination of amino acids and glucose Amino acids were analyzed as ortho-phthaldialdehyde derivatives by reversed-phase chromatography on an AminoQuant column at 40 C in a Hewlett-Packard series 1050 high-pressure liquid chromatography (HPLC) system. Glucose was determined by HPLC on

329 Fig. 2 Cloning strategy for the construction of the mobilizable plasmid used for dapE disruption. A 1.3-kb spectinomycin marker was removed from pAL214 via PstI-PvuII digestion, and this marker was used to replace an 85-bp NsiI-StuI internal fragment in the dapE gene of pDapEScript, thus rendering the dapE gene nonfunctional in plasmid pDapESm. The mobilizable plasmid pDapSupSm was constructed by ligating the 2.6-kb PstI fragment from pDapESm into the PstI site of the vector pSUP301. When initial unsuccessful transconjugation attempts were conclusively attributed to the presence of spectinomycin resistance in C. glutamicum wildtype strains, the spectinomycin marker in pDapSupSm was replaced by chloramphenicol. The mobilizable plasmid pDapCmSup was constructed by ligating the 1.1-kb SmaI-SalI fragment from pMG110 carrying the chloramphenicol-resistance marker into the partially digested NruI-SalI sites of pDapSupSm. The plasmid pDapCmSup was used for successful transconjugation of C. glutamicum 21253 and subsequent disruption of the dapE gene to form the strain 212DAPEM

330 a BioRad Aminex HPX-87H reverse-phase column at 45 C with a mobile phase consisting of 5 mM H2SO4.

Results
Site-specic disruption of the dapE gene in C. glutamicum In order to disrupt dapE with a selectable marker, the cat cartridge from pMG110 (Fig. 2) was used to substitute for an 85-bp DNA fragment of the dapE gene in plasmid pDapEScript. The disrupted dapE gene was transferred to the mobilizable vector pSUP301, and the nal plasmid, pDapCmSup, was mobilized from E. coli S17-1 into C. glutamicum 21253. After conjugation and DNA exchange between the dapE fragments on pDapCmSup and the chromosomal copy of the dapE gene, strain 21253 gained chloramphenicol resistance in conjunction with inactivation of the dapE gene. Desirable double-cross-over mutants were isolated from single-cross-over mutants by verifying chloramphenicol resistance and kanamycin sensitivity on complex medium plates. Disruption of the dapE gene was further conrmed by Southern blot analysis comparing the genomic DNA of the mutant strain 212DAPEM to the parent strain 21253, as shown in Fig. 3. Two independent probes were used for verication. The dapE gene probe indicated that the gene was internally disrupted by a 1.1-kb fragment of DNA, and the cat gene probe veried that the 1.1-kb internal fragment was the chloramphenicol resistance marker DNA. To further verify that the dapE gene was disrupted in strain 212DAPEM, in vitro enzymatic assays were performed on the crude extract of 21253, 212DAPEM, and two control strains. B. sphaericus 13A5 served as a negative control because this strain does not utilize the four-step succinylase pathway and therefore does not
A Lanes: 1 2 3 Hind III (wt) 23.1 kb

contain a dapE gene. Conversely, E. coli DH5a served as a positive control because this strain does not utilize the meso-Dap dehydrogenase pathway, so it contains all four genes in the four-step succinylase pathway. The N-succinyl-Dap desuccinylase assay results shown in Table 2 conrm that strain 212DAPEM does not have the desuccinylase activity. Assay controls indicate that the cofactor requirement is strict. The activity of this enzyme drops by 50% without the divalent cobalt ion. Amplication of meso-Dap dehydrogenase in the dapE-disruption strain, 212DAPEM Overexpression of the ddh gene obtained from PCR amplication was achieved by transforming C. glutamicum with plasmid pCSR1. Table 2 shows that strains
Fig. 3A, B Southern blot analysis validating mutant construction. Two Southern blots provide evidence for the disruption of the dapE gene in the chromosomal DNA of C. glutamicum 21253. Genomic DNA from both the parent and mutant strains was digested with PstI, subjected to agarose gel electrophoresis, blotted onto a nitrocellulose lter, and hybridized with a digoxigenin-11-dUTP probe. A, B Lanes: 1 the size marker fragments of labeled k DNA cleaved with HindIII, 2 the PstI-digested DNA of parent-type 21253 (wild type), 310 the PstI-digested DNA of dapE ) mutant candidates. Strain 212DAPEM containing the DNA (lane 9) was selected (on the basis of results of the N-succinyl-LL-dipimelate desuccinylase enzymatic assay) for further experimentation. The two panels show the results of two independently labeled DNA probes. A The probe specic for the dapE gene was isolated from the polymerase chain reaction product. When genomic DNA was probed with the dapE gene, the fragment of DNA in strain 21253 containing the gene is highlighted at 6.8 kb. The 1.1-kb displacement of the DNA fragments shown in lanes 310 corresponds to integration of a piece of DNA at the dapE gene location in the chromosome of the mutant strain candidates. B The probe specic for the chloramphenicol gene (cat) was isolated from pMG110 to conrm antibiotic marker integration into the Corynebacterium chromosome. When genomic DNA was probed with the cat gene, the restriction pattern reveals that this 1.1-kb increase in the mutant DNA fragment size (310) is indeed due to cat integration, which is noticeably absent from the wild-type DNA shown in 2

B 4 5 6 7 8 9 212DAPEM (dapE ) 10 1 2 3 Hind III (wt) 4 5 6 7 8 9 212DAPEM (dapE ) 10

9.4 kb 6.6 kb

4.4kb

2.3 kb 2.0 kb

331 Table 2 Specic enzymatic activities. Specic activities of mesodiaminopimelate (Dap) and succinyl-Dap desuccinylase in various C. glutamicum crude extracts are compared to the activity in E. coli and B. sphaericus control extracts. These activity results show the average of three independently prepared crude extracts, and the background activity (minus crude extract, substrates, or cofactors) has been subtracted from the nal value. Dehydrogenase error 5%; desuccinylase error 40% Strain Specic enzyme activities (nmol min)1 mg protein)1) Dehydrogenase C. glutamicum 21253 21253 (pCSR1) 212DAPEM 212DAPEM (pCSR1) B. sphaericus 13A5 E. coli DH5a 123 985 115 970 200 <55 Desuccinylase 2.5 1.9 <0.5 <0.5 <0.5 10.0

Fig. 5 Comparison of glucose uptake proles in C. glutamicum 21253 (j) and 212DAPEM (m). h, n Strains containing the plasmid PCSR1 (amplied ddh)

transformed with pCSR1 have an eightfold higher mesoDap dehydrogenase activity than the corresponding strains without the plasmid. B. sphaericus positive controls exhibited 60%70% higher meso-Dap dehydrogenase activity than C. glutamicum strains, consistent with literature data (Misono et al. 1979; Schrumpf et al. 1991). Impact of dapE disruption on cell growth, glucose uptake, and L-lysine production Wild-type and mutant strains were grown in 50 ml minimal medium shake-ask cultures to determine the eect of the dapE gene disruption on growth, glucose uptake, and lysine production rates in C. glutamicum. Growth proles were similar for all four strains, and data obtained during the exponential phase yielded equivalent estimates for the maximum specic growth rate (lmax 0.20/h), as shown in Fig. 4. The glucose uptake rates shown in Fig. 5 are another indication that

the phenotype of 212DAPEM (dapE) is indistinguishable from the parent-type 21253. Although there was an initial lag in the glucose consumption prole of strain 21253 (pCSR1), the maximum specic glucose uptake rate during exponential phase was comparable for all strains (qglu 0.30.4 g glucose g dry cell weight)1 h)1), indicating that the absence of the dapE gene does not aect glucose-intensive cellular processes required for cell maintenance or survival. Figure 6 shows the lysine production proles for the four strains. Although the time proles dier somewhat, lysine secretion rates, nal titers (1.2 g/l), and yields (0.06 mol lysine/mol glucose) are statistically equivalent, indicating that the lysine-producing capability of C. glutamicum is not impaired by the absence of the desuccinylase. This result was not aected by ddh overexpression. Our results thus demonstrate that amplifying enzymatic activity does not guarantee higher pathway ux, even in the absence of competing pathways.

Discussion
In order to probe the role of the dual pathway in lysine synthesis in C. glutamicum, a dapE )-disruption mutant

Fig. 4 Comparison of growth in C. glutamicum 21253 (j) and 212DAPEM (m). h, n Strains containing the plasmid pCSR1 (amplied ddh). Increases in cell density were measured by absorbance (A) at 660 nm and converted to dry cell weight (DCW) on the basis of the correlation DCW 0.28 A (Gubler et al. 1994)

Fig. 6 Comparison of lysine production rates in C. glutamicum 21253 (j) and 212DAPEM (m). h, n Strains containing the plasmid pCSR1 (amplied ddh)

332

was constructed and compared with the parent-type C. glutamicum 21253 phenotype. The disruption was conrmed by antibiotic selection, Southern blots, and enzymatic assays. The absence of N-succinyl-Dap desuccinylase activity had no eect on the growth rate, which is consistent with the results of Wehrmann et al. (1998), and no eect on the glucose-uptake rate. In addition, strain 212DAPEM (dapE )) produced lysine at a rate and yield similar to those of the parental strain 21253. Although a dapE ) strain grown in limiting ammonium may have a dierent morphology and increased susceptibility to lysis in comparison to the parent type (Wehrmann et al. 1998), these characteristics do not aect the lysine productivity of the mutant when grown in minimal medium with sucient ammonium to sustain normal growth. The one-step meso-Dap dehydrogenase pathway apparently compensates for the lack of the succinylase pathway ux to lysine. Although these results show that the dapE gene is dispensable for lysine production, it is important to note that the absence of the dapE gene does not imply that the entire four-step succinylase pathway is dispensable. Currently it is not clear if there are other cellular functions for the intermediates upstream of LL-Dap in the four-step pathway, and it is possible that, even in the dapE ) strain, the biosynthesis of these upstream intermediates is sucient to meet cellular demands. It is not clear if a larger-scale fed-batch or continuous culture of the dapE ) strain would elicit dierent physiological eects due to variations in intracellular uxes. The results shown in this study simply demonstrate that ux through the four-step pathway is not necessary for mesoDap or lysine production in small-scale batch shake asks. An eightfold amplication of meso-Dap dehydrogenase activity in strains 21253 and 212DAPEM also had no eect on growth, glucose uptake or lysine production. As shown in Table 2, the specic enzymatic activity of meso-Dap dehydrogenase under saturating in vitro conditions in parental strain 21253 is 123 nmol NADPH min)1 mg total protein)1. Assuming an equimolar conversion of NADPH to lysine in the reversible dehydrogenase reaction, and assuming that the dry cell weight (DCW) is 50% protein, this enzymatic rate corresponds to a specic lysine productivity of 0.54 g lysine g DCW)1 h)1 which is 50-fold higher than the activity needed for the production rates (approx. 0.012 g lysine/g DCW)1 h)1) observed in this study. In fact, this wildtype meso-Dap dehydrogenase activity is fourfold higher than the maximum production rates (approx. 0.15 g lysine g DCW)1 h)1) observed in larger-scale fed-batch cultures of C. glutamicum 21253 (Kiss and Stephanopoulos 1991). Since the wild-type meso-Dap dehydrogenase activity is already sucient to meet the cellular demand for lysine, it is not surprising that amplication of the ddh gene has no eect on lysine production rates. It should be noted that the correlation of enzymatic activity and lysine production is valid even when the succinylase pathway is not used, as shown by the proles

of strain 212DAPEM. The results of this genetic study support the hypothesis that meso-Dap dehydrogenase is not a rate-limiting step in lysine biosynthesis in smallscale cultures. The correlation between the in vitro measurement and calculated ux capacity of the mesoDap dehydrogenase enzyme suggests that the rate of formation of the tetrahydrodipicolinate precursor may limit lysine production in this system. Wehrmann et al. (1994) were successful in achieving a tenfold overexpression of the N-succinyl-Dap desuccinylase activity, but they do not report the eects of this amplication on lysine production. Since all enzymes in the four-step succinylase pathway have low activity and are produced in relatively low abundance (Chatterjee and White 1982; Lin et al. 1988; Schrumpf et al. 1991), they would probably require simultaneous amplication to eect higher ux through the pathway.
Acknowledgements We are grateful to Divya Janardhana for technical assistance. This research was supported by the National Science Foundation (grant BES-9634569) and the Department of Energy BES (grant DE-FG02-94ER14487). C. Shaw-Reid was the recipient of a National Science Foundation Graduate Fellowship and a Ford Foundation Dissertation Fellowship.

References
Bouvier J, Richaud C, Higgins W, Bogler O, Stragier P (1992) Cloning, characterization, and expression of the dapE gene of Escherichia coli. J Bacteriol 174: 52655271 Boyen A, Charlier D, Charlier J, Sakanyan V, Mett I, Glansdor N (1992) Acetylornithine deacetylase, succinyldiaminopimelate desuccinylase, and carboxypeptidase G2 are evolutionarily related. Gene 116: 16 Bradford M (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein using the principle of protein-dye binding. Anal Biochem 72: 248254 Chatterjee SP, White PJ (1982) Activities and regulation of the enzymes of lysine biosynthesis in a lysine-excreting strain of Bacillus megaterium. J Gen Microbiol 128: 10731081 Cremer J, Eggeling L, Sahm H (1991) Control of the lysine biosynthesis sequence in Corynebacterium glutamicum as analyzed by overexpression of the individual corresponding genes. Appl Environ Microbiol 57: 17461752 Follettie MT (1989) Development of recombinant DNA technology for Corynebacterium glutamicum and its inuence on the carbon ux in dierent strains. PhD thesis. Massachusetts Institute of Technology, Cambridge, Mass Follettie MT, Peoples O, Agoropoulou C, Sinskey AJ (1993) Gene structure and expression of the Corynebacterium avum N13 ask-asd operon. J Bacteriol 175: 40964103 Gilvarg C (1959) N-Succinyl-L-diaminopimelic acid. J Biol Chem 234: 29552959 Gubler M, Park SM, Jetten M, Stephanopoulos G, Sinskey AJ (1994) Eects of phosphoenol pyruvate carboxylase deciency on metabolism and lysine production in Corynebacterium glutamicum. Appl Environ Microbiol 40: 857863 Hanahan D (1983) Studies of transformation of Escherichia coli with plasmids. J Mol Biol 166: 557580 Ishino S, Mizukami T, Yamaguchi K, Katsumata R, Araki K (1987) Nucleotide sequence of the meso-diaminopimelate D-dehydrogenase gene from Corynebacterium glutamicum. Nucleic Acids Res 15: 3917 Jetten MSM, Sinskey AJ (1993) Characterization of phosphoenolpyruvate carboxykinase from Corynebacterium glutamicum. FEMS Microbiol Lett 111: 183188

333 Jetten MSM, Sinskey AJ (1995) Recent advances in the physiology and genetics of amino acid-producing bacteria. Crit Rev Biotechnol 15: 73103 Jetten MSM, Follettie MT, Sinskey AJ (1995) Eect of dierent levels of aspartokinase on lysine production by Corynebacterium lactofermentum. Appl Microbiol Biotechnol 43: 7682 Kindler SH, Gilvarg C (1960) N-Succinyl-L-a,e-diaminopimelic acid deacylase. J Biol Chem 235: 35323535 Kiss RD, Stephanopoulos G (1991) Metabolic activity control of the L-lysine fermentation by restrained growth fed-batch strategies. Biotechnol Prog 7: 501509 Lin Y, Myhrman R, Schrag MI, Gelb M (1988) Bacterial N-succinyl-L-diaminopimelic acid desuccinylase: purification, partial characterization, and substrate specificity. J Biol Chem 263: 16221627 Misono H, Togawa H, Yamamoto T, Soda K (1979) meso-a-eDiaminopimelate D-dehydrogenase: distribution and the reaction product. J Bacteriol 137: 2227 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning laboratory manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Schafer A, Kalinowski J, Simon R, Seep-Feldhaus A-H, Puhler A (1990) High-frequency conjugal plasmid transfer from gramnegative Escherichia coli to various gram-positive coryneform bacteria. J Bacteriol 172: 16631666 Schrumpf B, Schwarzer A, Kalinowski J, Puhler A, Eggeling L, Sahm H (1991) A functionally split pathway for lysine synthesis in Corynebacterium glutamicum. J Bacteriol 173: 4510 4516 Schwarzer A, Puhler A (1991) Manipulation of Corynebacterium glutamicum by gene disruption and replacement. Biotechnology 9: 8487 Sonntag K, Eggeling L, De Graaf AA Sahm H (1993) Flux partitioning in the split pathway of lysine synthesis in Corynebacterium glutamicum. Eur J Biochem 213: 13251331 Sundharadas G, Gilvarg C (1967) Biosynthesis of a-e-diaminopimelic acid in Bacillus megaterium. J Biol Chem 242: 3983 3988 Wehrmann A, Eggeling L, Sahm H (1994) Analysis of dierent DNA fragments of Corynebacterium glutamicum complementing dapE of Escherichia coli. Microbiology 140: 33493356 Wehrmann A, Morakkabati S, Kramer R, Sahm H, Eggeling L (1995) Functional analysis of sequences adjacent to dapE of Corynebacterium glutamicum reveals the presence of aroP, which encodes the aromatic amino acid transporter. J Bacteriol 177: 59915993 Wehrmann A, Phillipp B, Sahm H, Eggeling L (1998) Dierent modes of diaminopimelate synthesis and their role in cell wall integrity: a study with Corynebacterium glutamicum. J Bacteriol 180: 31593165 Wu B, Georgopoulos C, Ang D (1992) The essential Escherichia coli msgB gene, a multicopy suppressor of a temperature-sensitive allele of the heat shock gene grpE, is identical to dapE. J Bacteriol 174: 52585264 Yeh P, Sicard AM, Sinskey AJ (1988) General organization of the genes specically involved in the diaminopimelate-lysine biosynthethic pathway of Corynebacterium glutamicum. Mol Gen Genet 212: 105111