Gluconate as suitable potential reduction supplier in Corynebacterium glutamicum . Cloning and expression of gnt P and gnt K in Escherichia coli

Corynebacterium glutamicum is widely used in the industrial production of amino acids. We have found that this bacterium grows exponentially on a mineral medium supplemented with gluconate. Gluconate permease and Gluconokinase are expressed in an inducible form and, 6-phosphogluconate dehydrogenase, although constituvely expressed, shows a 3-fold higher specific level in gluconate grown cells than those grown in fructose under similar conditions. Interestingly, these activities are lower than those detected in the strain Escherichia coli M1-8, cultivated under similar conditions. Additionally, here we also confirmed that this bacterium lacks 6-phosphogluconate dehydratase activity. Thus, gluconate must be metabolized through the pentose phosphate pathway. Genes encoding gluconate transport and its phosphorylation were cloned from C. glutamicum, and expressed in suitable E. coli mutants. Sequence analysis revealed that the amino acid sequences obtained from these genes, denoted as gntP and gntK, were similar to those found in other bacteria. Analysis of both genes by RT-PCR suggested constitutive expression, in disagreement with the inducible character of their corresponding activities. The results suggest that gluconate might be a suitable source of reduction potential for improving the efficiency in cultures engaged in amino acids production. This is the first time that gluconate specific enzymatic activities are reported in C. glutamicum. Key terms: Corynebacterium glutamicum, gluconate metabolism, gnt Tomás Istúriz. Instituto de Biología Experimental, Facultad de Ciencias, Universidad Central de Venezuela, Apartado Postal 47557, Caracas 1041-A, Venezuela. toisturiz@hotmail.com Tel: 582127510111 Fax: 582127535897 Received: March 28, 2008. In Revised form: September 4, 2008. Accepted: October 28,2008 INTRODUCTION Corynebacterium glutamicum is an aerobic, gram-positive, non-sporulating and nonpathogenic bacillus that lives in the soil. It is a microorganism widely used in the industrial production of primary metabolites, such as L-glutamate, L-lysine and nucleotides, because of which there is ongoing interest in developing more efficient strains of corynebacterium. Research has resulted in significant advances in the biochemistry, physiology and molecular genetics of this organism, with special attention to its aminoacids biosynthetic pathways (Hermann, 2003; Jetten and Sinskey, 1995, Kaliwonski et al., 2003). Better conditions and more efficient C. glutamicum strains for amino acid production are obtained through cloning and the characterization of the genes involved in their biosynthesis, as well as metabolic studies with emphasis on the carbon flux distribution between glycolysis and the pentose phosphate pathway (PPP) under particular conditions of growth (Eggeling, 1994, Kirchner and Tauch, 2003, Sahm et al. , 2000, Vallino and Stephanopoulos 1993). Our interest in C. glutamicum is associated with the gluconate metabolism, which is one of our areas of study in E. coli (De Rekarte et al., 1994; Istúriz et al., 1986; Porco et al., 1998). It was known (Vallino and Stephanopoulos, PORCO ET AL. Biol Res 41, 2008, 349-358 35


INTRODUCTION
Corynebacterium glutamicum is an aerobic, gram-positive, non-sporulating and nonpathogenic bacillus that lives in the soil.It is a microorganism widely used in the industrial production of primary metabolites, such as L-glutamate, L-lysine and nucleotides, because of which there is ongoing interest in developing more efficient strains of corynebacterium.Research has resulted in significant advances in the biochemistry, physiology and molecular genetics of this organism, with special attention to its aminoacids biosynthetic pathways (Hermann, 2003;Jetten andSinskey, 1995, Kaliwonski et al., 2003).Better conditions and more efficient C. glutamicum strains for amino acid production are obtained through cloning and the characterization of the genes involved in their biosynthesis, as well as metabolic studies with emphasis on the carbon flux distribution between glycolysis and the pentose phosphate pathway (PPP) under particular conditions of growth (Eggeling, 1994, Kirchner and Tauch, 2003, Sahm et al., 2000, Vallino and Stephanopoulos 1993).Our interest in C. glutamicum is associated with the gluconate metabolism, which is one of our areas of study in E. coli (De Rekarte et al., 1994;Istúriz et al., 1986;Porco et al., 1998).It was known (Vallino and Stephanopoulos, 1994b) that C. glutamicum grows in mineral media with glucose as the sole carbon source.However, if supplemented only with gluconate, the growth is linear, produces cell lysis and the Entner-Doudoroff pathway (EDP) activities are not detected (Lee et al., 1998).The addition of glucose to the gluconate containing cultures alleviates this problem.Moreover, the specific production of L-lysine in this microorganism was enhanced when gluconate was used as a secondary carbon source with glucose, presumably by relieving the limiting factors in the lysine synthesis rate as the NADPH supply (Lee et al., 1998).Diverse metabolic flux studies have revealed a correlation between lysine production and the NADPH supplied by carbon flux through the pentose phosphate pathway (Wittmann andHeinzle, 2002, Beker et al., 2007).Recently, metabolic flux engineering has addressed the over expression of the zwf gene, which encodes glucose 6-phosphate dehydrogenase, resulting in increased lysine production, probably due to an overall NADPH excess (Becker et al., 2007).
In regard to the organization of the initial gluconate metabolism genes in C. glutamicum, a recent study identified and analyzed gntP and gntK as the genes involved, which are responsible of the gluconate transport and the gluconokinase activity respectively (Letek et al., 2006).Although the expression analysis revealed monocistronic transcripts and constitutive expression for both genes (Letek et al., 2006), a subsequent study (Frunzke et al., 2008) identified two regulators (GntR1 and GntR2), which repress the expression of genes involved in gluconate metabolism (e.g.gntK, gntP and gnd) in the absence of the substrate.
Here we report physiological and genetic studies on the initial steps of gluconate utilization by C. glutamicum, i.e., substrate transport and its phosphorylation.While in E. coli, this acid sugar once phosphorylated is both, oxidatively descarboxilated by the 6-phosphogluconate dehydrogenase (Gnd), the third enzyme of the pentose phosphate pathway (PPP) and dehydrated by the 6phosphogluconate dehydratase (Edd), the first enzyme of the EDP (Fraenkel, 1996), in C. glutamicum, gluconate seems to feed exclusively in the PPP (Vallino and Stephanopoulos, 1994b).We have confirmed that C. glutamicum grows on modified mineral medium supplemented with gluconate as a sole carbon source.In this condition, while the Edd activity is not detected, low gluconokinase and gluconate transport activities are expressed in an inducible form.The growth of cells in mineral medium supplemented with glucuronate indicated the presence of the second enzyme of the EDP, 2-keto-3deoxy-6-phosphogluconate aldolase (KDPG aldolase), which was confirmed through enzymatic assays.Using oligonucleotide primers designed from the C. glutamicum genome sequences reported in the GenBank (NC 003450), the gluconokinase and gluconate permease genes were amplified from this bacterium, cloned and expressed in appropriate E. coli mutants.RT-PCR experiments indicated that these genes are expressed constitutively in C. glutamicum, which is not in agreement with the inducible character of the corresponding enzymatic activities.Because the results suggest that gluconate catabolism in C. glutamicum is a suitable reduction potential supplier, the complementation of culture media with this substrate might be used to improve efficiency in the amino acids production state of the bacteria.

Organisms
The strains and plasmids used in this study are listed in Table I.E. coli strains are K12 derivatives.

Growth of bacteria
The cells were routinely grown aerobically at 37 °C in volumes of 10 ml for growth curves and 20 ml for enzyme assays in 125 ml flasks fitted with side arms, on a gyratory water bath (model G76, New Brunswick) at about 200 cycles min -1 .In each case, the growth was monitored by reading the optical density in a Klett colorimeter with a N° 42 filter (one Klett unit is approximately 2 x 10 6 cell ml -1 ).

Preparation of crude extracts
Cells were harvested by centrifugation, resuspended in 50 mM Tris-HCl 10 mM MgCl 2 (pH 7.6) and disrupted by 30s sonication pulses (16 and 2 pulses for C. glutamicum and E. coli respectively) in a Braun Sonic 2000 (12T probe, 45 wattage level) with cooling periods between pulses.Cell debris was removed by centrifugation at 2700xg for 15 min.

Enzyme assays
The gluconokinase, gluconate 6phosphate dehydrogenase, 6phosphogluconate dehydratase (Edd) and KDPG aldolase (Eda) activities were assayed as previously described (Fraenkel and Horecker, 1964).KDPG for Eda assays was obtained as described (Conway et al., 1991): E. coli DF214 carrying the pT280 plasmid, which contains the E. coli edd gene, generated by PCR and inserted immediately downstream of the lac promoter (Egan et al., 1992), was grown overnight in Lb containing ampicillin (100 μg ml -1 ) and IPTG (0,5 mM) at 37 °C.Cultures were centrifuged for 5 min.(3000xg) and the cells re-suspended in buffer MES-MgCl 2 [50mM 2(N-morpholino) ethanesulfonic acid, 10 mM MgCl 2 ] to a final A 550 of 1.Cells from 2 ml of this suspension, once centrifuged as before, were re-suspended in 500 μl of the same buffer, and disrupted by two 30s sonication pulses.The disrupted cells were centrifuged (27000xg, 15 min.)and the supernatant was added to 12.5 ml of 6-phosphogluconate 5 mM.This mixture was incubated at 37 ºC, for 30 min, heat inactivated by incubation at 90 ºC for 5 min., centrifuged (27000xg, 15 min.)and the supernatant was used as a substrate for KDPG aldolase assays.Activities are reported as nmol min -1 (mg protein) -1 .

DNA isolation and manipulation
Plasmids and total DNA were isolated using standard DNA manipulation protocols (Ausubel et al., 1999;Birboim and Doly, 1982).In order to obtain the total DNA from C. glutamicum, cells were pretreated by re-suspension in lysis buffer containing lysozyme (15 mg ml -1 ) and incubated in a shaking bath for 3h, at 37 ºC.

PCR amplification, cloning and sequencing
For amplification of genes encoding gluconate transport and phosphorylation activities from C. glutamicum, primers were designed from the GenBank DNA sequences of the corresponding putative genes (accession number NC003450).Primers P1 (5'-AGCCGGATACAATCCCA ATACAGC-3') and P2 (5'-CGATTTCAGT CGGATTATCACCCG-3'), for gluconate permease gene, and primers P3, (5'-AAACTTACGCCAGGAAGTATCCGC-3') and P4, (5'-GTGTTCTTGCCATCCATTG TGCC-3') for gluconokinase gene.Taq polymerase from Invitrogen was used for PCR.Samples of 50 μl, were prepared according to the manufacturer's instructions.The mixture was heated at 94 °C, for 5 min, followed by 30 cycles of the following program: 1 min at 94 °C, 1 min at 60 °C and 1 min at 72 °C.PCR products were analyzed by agarose (1.5%) gel electrophoresis and sequenced by using an automated ABI 377 instrument (CeSAAN, IVIC).PCR products were cloned into the pCR ® 2.1-TOPO vector (Promega) according to the instructions of the manufacturer.They were then used to transform the E. coli DH5α(mcr) strain.

RT PCR analysis
To examine the gntK and gntP gene expression, reverse transcription-PCR (RT-PCR) analysis was performed with the total RNA isolated by the Trizol reagent (GIBCO/BRL) according to the instructions of the manufacturer.The cells were precultured at 37 °C, in MCGC medium with fructose and collected during the log phase.Aliquots of the pre-culture were diluted 10fold in the same medium, and cultures (20 ml) in the presence of fructose and glucose were carried out at 37 ºC, until reaching an absorbance of 0.6 at 600nm.In each case, the RT reaction was carried out with 1 μg of the respective total RNA with M-MLV Reverse Transcriptase, using random primers, following the manufacturer's instructions.PCR (35 cycles) performed with the primers P5 (5'-ACCCCAGCTAAC GCAGTGTC-3') and P6 (5'-CGGTTGCCT AGGAAGAACAG-3'), and primers P7 (5'-AGCAGCCGAAGGCTTACATA-3') and P8 (5'-CAACCTGGACTAGCCACCAT-3') for gntP and gntK ORFs, respectively, consisted of denaturation at 95 °C for 1 min, annealing at 58 °C for 1 min, and extension at 72 °C for 1 min.The PCR products were analyzed by 1.5% agarose gel electrophoresis.The relative amounts of RT-PCR products on the gel were compared by measuring the band density after the color of the image obtained was reversed by using a model GS-700 imaging densitometer (Bio-Rad).This experiment was repeated at least twice.As controls, PCR were carried out in RNA samples without RT.

Gluconate catabolism in C. glutamicum
Initially, we confirmed previous results (Lee et al., 1998;Vallino and Stephanopoulos, 1994b) about the capability of C. glutamicum (ATCC 13032) to utilize gluconate as a sole energy and carbon source.This strain, pre-cultivated in Lb, grew aerobically at 37 o C in MCGC medium supplemented with glucose (0.5%) or gluconate (0.5%) with generation times of 90 and 130 minutes, respectively.The lag period was about an hour in the former condition and two hours in the latter.Fructose, galactose, maltose and glucuronic acid were also used as carbon sources for C. glutamicum growth (data not shown).
We investigated the presence of gluconate activities in C. glutamicum grown in MCGC medium, supplemented with fructose, gluconate, or glucuronate.Table II shows that while the activities for transport and phosphorylation of this substrate are induced in the gluconate culture [32 pmol x 10 7 cells min -1 and 47 nmol min -1 (mg prot) -1 , respectively], 6-phosphogluconate dehydratase is not detected in any of the conditions assayed.However, KDPG aldolase is induced when the cells are grown in the presence of glucuronate [38 nmol min -1 (mg prot) -1 ].Likewise, the specific activity of 6-phosphogluconate deshydrogenase, expressed in a semiconstitutive form, shows a 3-fold higher level in gluconate than in fructose.Interestingly, the transport and phosphorylation activities in C. glutamicum were lower than those detected in the E. coli strain M1-8 grown under similar experimental conditions.Although gluconate is not a good carbon source for C. glutamicum growth (Vallino and Stephanopolous, 1994a), the capture and phosphorylation of this substrate might improve bacterial amino acid production by the generation of reducing power via the Gnd enzyme.

Cloning of genes for transport and phosphorylation of gluconate in C. glutamicum
Recently, two genes, gntK and gntP, involved in gluconate catabolism of C. glutamicum were reported (Letek et al., 2006).Based on that information, and having at our disposal E. coli gntK and gntT mutants, we proceeded to demonstrate through cloning, complementation and enzymatic assays in the above E. coli suitable mutants, if the C. glutamicum ORFs reported by GenBank and investigated by Letek et al. (2006) certainly encode activities of transport and phophorylation of gluconate.It is known that C. glutamicum genes are expressed in E. coli (Eikmanns, B. 1992).
Two sets of primers were prepared, which were designed from sequences reported in the GenBank as responsible ORFs for the transport (GenBank ID 1020851) and phosphorylation (GenBank ID 1020432) of gluconate.The two PCR products obtained from the C. glutamicum (ATCC 13032) genome were of approximately 1723 bp and 861 bp, In agreement with previous reports (Patek et al., 2003;Letek et al., 2006), presumptive promoter regions with a -10 (TATAGT) for gntP and -10 (TATGAT) for the gntK ORF were identified.Sequences resembling -35 regions, which is not conserved in C. glutamicum (Patek et al., 2003), were not identified.
The deduced amino acid sequences of the ORFs from pTAEP y pTAEK resemble those of the corresponding proteins in E. coli.The former product has 28%, 27% and 30% identity with GntT, GntU and IdnT, respectively, and the latter has 42% identity with GntK.According to the data bank, no other C. glutamicum ORF has been identified as a presumptive protein encoding for a gluconate transporter or gluconate phosphorylation activity.

E. coli complementation by C. glutamicum cloned genes
In correspondence with the sequence analysis, the pTAEP clone complemented the E. coli mutant TMC297 on both mineral gluconate and GBTB plates, indicating that the genomic C. glutamicum DNA fragment carried by this clone certainly includes the gntP gene, specifying gluconate transport activity.On the contrary and unexpectedly, the pTAEK clone did not complement the E. coli mutant TGN282 on similar plates.
The failure of pTAEK to complement the E. coli mutant TGN282 was a deceptive presumption.We observed that transformed colonies did not arise on mineral gluconate plates.However, those that arose on GBTB plates were white (non-fermentative) and particularly smaller than E. coli DH5α(mcr) transformed colonies with the same plasmid, but selected on Lb Amp-plates.This observation suggested that the failure of growth could be the result of a 6phosphogluconate accumulation due to high levels of gluconokinase activity, when the transformed cells were selected on gluconate containing plates.The toxicity caused by the intracellular accumulation of phosphorylated compound has been reported (De Rekarte et al., 1994).

Activities of gluconate metabolism in E. coli transformed cells
In order to demonstrate the expression of the C. glutamicum cloned gluconate genes in E. coli, and also to support the above hypothesis, the specific activities for transport and phosphorylation of gluconate were estimated in transformed E. coli DH5α(mcr) cells, as well as in transformed E. coli mutants, cultivated in a CAA medium supplemented with fructose or gluconate (Table III).In the E. coli mutant TMC297 carrying the pTAEP plasmid, the gluconate transport was expressed in a partially constitutive form (691 pmol x 10 7 cells min -1 in fructose vs 1194 pmol x10 7 cells min -1 in gluconate).E. coli mutant TGN282, transformed with the pTAEK plasmid, did not grow in gluconatesupplemented medium.However, when these cells were cultivated in fructose, they registered high levels of gluconate kinase activity [2894 nmol min -1 (mg prot) -1 ].Similar results were observed in the E. coli DH5α(mcr) carrying the same plasmids; while E. coli DH5α(mcr)(pTAEP) grew on gluconate containing medium and showed high levels of gluconate uptake (425 pmol x 10 7 cells min -1 cultivated in mineral fructose), E. coli DH5α(mcr)(pTAEK) did not grow on gluconate supplemented medium, but showed high levels of gluconate kinase specific activity when cultivated in fructose [1429 nmol min -1 (mg prot) -1 ], confirming that the segment carried by the plasmid certainly encodes a gluconokinase

Expression of C. glutamicum gntP and gntK genes
In order to examine the possibility of identifying the inducible character of GntP and GntK in expression studies, RT PCR assays were made from RNA isolated from C. glutamicum, cultivated in gluconate or fructose as sole carbon sources.The results did not show significant differences between the two growth conditions investigated, suggesting a constitutive expression for these genes (data not shown).

DISCUSSION
The data reported here provides evidences for the presence, in C. glutamicum, of enzymes involved in the transport and phosphorylation of gluconate, which are codified by gntP and gntK genes, respectively.Basic and novel aspects of the gluconate catabolism in this bacterium have been investigated.The results indicate that C. glutamicum grows exponentially in MCGC medium with gluconate (0.5%) as the sole carbon source and, for the first time, data involving specific gluconate activities in these conditions are reported.Gluconate, once incorporated and phosphorylated, is decarboxilated oxidatively via PPP, because this bacterium lacks Edd, but not Gnd.These characteristics, and the enhancement of Gnd specific activity in gluconate containing cultures, which is mainly addressed to generate reducing power, might facilitate the search for strategies to improve the efficiencies of C. glutamicum in production conditions.It is known that the yield of Llysine by this bacterium is increased if cultivated in glucose plus gluconate (Lee et al., 1998;Coello et al., 1992).This result, as well as ours, can be explained by the coordinate and negative regulation of GntR1 and GntR2, two redundant repressors of gntP, gntK and gnd whose actions are interfered by gluconate (Frunzke et al., 2008).The C. glutamicum growth in a mineral gluconate medium results in the derepression of the mentioned genes with a significant increase in the levels of Gnd, previously expressed as constitutive basal activity.Because gluconate cometabolizes with glucose in this bacterium, the growth rate in mineral gluconate plus glucose medium not only increases (compared to the medium supplemented with either substrate {Frunzke et al., 2008}), but also participates in the generation of reduction potential, as gluconate is totally metabolized via PPP and glucose can still be partitioned at the glucose-6-phosphate level.The importance of the reducing power in the C. glutamicum lysine production was also observed through the over-expression of the zwf gene (Becker et al., 2007); so the utilization of gluconate by this bacterium with the corresponding increase of Gnd activity and reduction potential, might stimulate the optimization of culture conditions to improve production conditions without complex genetic manipulations.

Initial activities of gluconate metabolism in C. glutamicum
The specific GntP and GntK activities were expressed in an inducible form and showed lower levels than those of the E. coli mutant (gntR) M1-8, cultivated in MCGC medium with gluconate (Table 2).The low levels of specific activities detected for GntP and GntK might explain why C. glutamicum shows linear growth in a basic mineral medium supplemented with gluconate as sole carbon source (Vallino and Stephanopolous, 1994b), and why this growth is improved when glucose is added or when a special mineral media (MCGC), which contains citrate, is used.Perhaps gluconate is not able to support the growth of this bacteria and an additional source of energy is required.The fact that these low levels are detected, even in cells cultivated in MCGC supplemented with gluconate, explains the difficulty of detecting gluconate transport and gluconokinase activities in extracts from cells cultivated in the same media with fructose; due probably to catabolite repression caused by the presence of this substrate (Letek et al., 2006), or the effect of regulators recently revealed (GntR1 and GntR2; Frunzke et al., 2008).
The inducible character of GntP and GntK seems to contrast with the constitutivity of their respective genes as reported by Letek et al. (2006) and observed by us on the basis of nonquantitative expression studies.In this concern, it is clear that the negative regulatory circuit uncovered by Frunzke et al. (2008, see above) certainly supports the inducibility, not only of these genes, but also the semiconstitutivity of gnd.Consequently, the dissimilarity between our results might be due to different sensitivities among the techniques used, since contrarily to RT-PCR, the enzyme assay registers the final product, i.e., the protein; alternatively, the presence of a unknown regulatory circuit blocking the translation of messengers in conditions of non-induction, should not be discarded.It is not advantageous for the cell, in energetic terms, to synthesize GntP and GntT in absence of gluconate.
C. glutamicum gntP and gntK genes, cloned in the pCR ® 2.1-TOPO vector, were expressed constitutively in E. coli mutants lacking their own gluconate activities.Specific levels of GntP and GntK in the transformants were particularly high, probably as a consequence of the multicopy character of the vector used, so it was not possible to infer some effects of the intracellular medium.Notably, GntK activity could be registered only in extracts of cells cultivated in MCGC plus fructose, where the formation of gluconate 6phosphate is low and the toxicity of the cell is not compromised.Because the Gnd activity in C. glutamicum increases in the presence of its substrate and seems to be a signal of production conditions, it would be of interest to study this in mutants with a high capacity to form gluconate 6phosphate from gluconate.
Lb and collected during the exponential phase; then cultivated in CAA medium with the indicated carbon sources at 0.2%, from approximately 10 KU up to 120 KU.Transformed cells were grown in the presence of ampicillin (100 μg/ml).NI, not investigated.ND, not detected.NG, no growth.For units, see Materials and methods.The values for activities represent means ± standard deviations from two independent experiments.