Characterisation of glutamine fructose-6-phosphate amidotransferase (EC 126.96.36.199) and N-acetylglucosamine metabolism in Bifidobacterium
- First Online:
- Cite this article as:
- Foley, S., Stolarczyk, E., Mouni, F. et al. Arch Microbiol (2008) 189: 157. doi:10.1007/s00203-007-0307-9
- 179 Views
Bifidobacterium bifidum, in contrast to other bifidobacterial species, is auxotrophic for N-acetylglucosamine. Growth experiments revealed assimilation of radiolabelled N-acetylglucosamine in bacterial cell walls and in acetate, an end-product of central metabolism via the bifidobacterial d-fructose-6-phosphate shunt. While supplementation with fructose led to reduced N-acetylglucosamine assimilation via the d-fructose-6-phosphate shunt, no significant difference was observed in levels of radiolabelled N-acetylglucosamine incorporated into cell walls. Considering the central role played by glutamine fructose-6-phosphate transaminase (GlmS) in linking the biosynthetic pathway for N-acetylglucosamine to hexose metabolism, the GlmS of Bifidobacterium was characterized. The genes encoding the putative GlmS of B. longum DSM20219 and B. bifidum DSM20082 were cloned and sequenced. Bioinformatic analyses of the predicted proteins revealed 43% amino acid identity with the Escherichia coli GlmS, with conservation of key amino acids in the catalytic domain. The B. longum GlmS was over-produced as a histidine-tagged fusion protein. The purified C-terminal His-tagged GlmS possessed glutamine fructose-6-phosphate amidotransferase activity as demonstrated by synthesis of glucosamine-6-phosphate from fructose-6-phosphate and glutamine. It also possesses an independent glutaminase activity, converting glutamine to glutamate in the absence of fructose-6-phosphate. This is of interest considering the apparently reduced coding potential in bifidobacteria for enzymes associated with glutamine metabolism.
KeywordsBifidobacteriumGlmSGlutamine fructose-6-phosphate amidotransferaseN-acetylglucosamine
Bifidobacterium spp., members of the gastrointestinal microflora, are believed to play an important role in human health being associated with a number of beneficial effects including the maintenance of a healthy gut microflora and immunomodulation (Chiang et al. 2000; Ouwehand et al. 2002). The diversity and numbers of bifidobacteria colonising the human gastrointestinal tract differ with age, diet and health status (Haarman and Knol 2005; Harmsen et al. 2000; Matsuki et al. 2004; Ouwehand et al. 2001). Due to the associated health benefits and the concomitant interest in the development of probiotic and prebiotic foods and food ingredients, there is considerable research interest in identifying and understanding the metabolism of factors promoting growth of bifidobacteria both in the large intestine and in the foods. Several so-called bifidogenic factors have been identified to date, including N-acetylglucosamine-containing oligosaccharides from humans which promote the growth of Bifidobacterium bifidum (Petschow and Talbott 1991).
Previous studies suggest that B. bifidum is auxotrophic for N-acetylglucosamine (Gygory et al. 1954). In addition, it is known that B. bifidum DSM 20082 possesses galactosyl-N-acetyl-hexosamine phosphorylases which act on β-d-galactosido-(1,3)-N-acetylglucosamine and β-d-galactosido-(1,3)-N-acetylgalactosamine, both intermediates of mucin hydrolysis, resulting in the release of phosphorylated galactose and N-acetylhexosamines (Derensy-Dron et al. 1999). Indeed similar enzyme activity has been identified in Bifidobacterium longum (Kitaoka et al. 2005). It is therefore hypothesized that in the gastrointestinal tract, which could be considered as the natural ecological niche for B. bifidum, this organism is capable of acquiring N-acetylglucosamine via its mucin-degrading activity (Derensy-Dron et al. 1999; Katayama et al. 2005).
Amino sugars such as N-acetylglucosamine are major components of peptidoglycan and lipopolysaccharides in bacterial cell walls. Bacteria are both capable of acquiring these amino sugars from their own environment and synthesizing them within the cell as required. The key enzyme for de novo amino sugar synthesis is glutamine fructose-6-phosphate amidotransferase (GlmS), an enzyme that generates glucosamine-6-phosphate from fructose-6-phosphate and glutamine (Milewski 2002). This enzyme has been characterized in a number of bacteria, in particular Escherichia coli. It is organized as a homodimer composed of a 30 kDa N-terminal domain with glutaminase activity and a 40 kDa C-terminal domain with isomerase activity, with a channel permitting transfer of an ammonium ion from the glutaminase to the isomerase domain (Teplyakov et al. 2001).
Considering the growth-stimulating properties of N-acetylglucosamine, this study focuses on N-acetylglucosamine metabolism in Bifidobacterium. It includes characterization of the bifidobacterial GlmS, a key enzyme linking the hexose metabolic pathway to the biosynthetic pathway for cell wall peptidoglycan precursors.
Materials and methods
Bacterials strains, plasmids and growth conditions
Bacterial strains used in this study
Reference or source
Child feces, Skerman et al. (1980)
204 3 a
205 1 b
206 1 a
1603 2 c
Adult feces, Skerman et al. (1980)
1601 III B
Bottle fed infant feces
Bottle fed infant feces
Bottle fed infant feces
Infant feces, Skerman et al. (1980)
Adult feces, Skerman et al. (1980)
203 IV B
204 I H
Child feces, Skerman et al. (1980)
BL 21 (DE3)/pLysS
F−ompT hsdSB (rB− mB−) gal dcm (DE3)/pLys (Cmr)
Stratagene, Studier et al. (1990)
General DNA methods
Total DNA was obtained from Bifidobacterium strains as described by Rossi et al. (2000). DNA manipulations, including restriction enzyme analysis, ligation and transformation were done by standard methods (Sambrook et al. 1989), or according to the manufacturer’s instructions. PCR was routinely performed in a Master Thermal Cycler (Eppendorf, Hamburg, Germany). DNA sequencing was performed by Genoscreen (Lille, France).
Monitoring of N-acetylglucosamine metabolism
Bifidobacterium bifidum DSM 20082 was grown at 39°C in modified Garches medium supplemented with either fructose, N-acetylglucosamine, or N-acetylglucosamine and fructose (5%) (Krzewinski et al. 1997) under a CO2 atmosphere (GENbox anaer, Biomerieux) in anaerobic jars. At regular time intervals, the optical density (OD) was measured at 550 nm. The concentration of fructose and N-acetylglucosamine in the cell-free supernatant was determined using the methods described by Rimington (1931) and, Good and Bessman (1964), respectively.
To follow N-acetylglucosamine metabolism, radiolabelled N-acetylglucosamine (N-acetyl-d-[1-14C] glucosamine, 0.5 μCi/ml, Amersham) was added to the medium. At particular time intervals, culture samples were taken and were filtered through a 0.22 μm membrane filter (Millipore). The filter was washed three times using 5 ml of H2O, dried and transferred into 4 ml of aqueous counting scintillant (Lipoluma). The radioactivity was measured with a liquid scintillation counter (Beckman) in order to monitor incorporation of N-acetylglucosamine into cell wall fraction (peptidoglycan). The amounts of N-acetylglucosamine and acetate (endproduct of d-fructose-6-phosphate shunt; Fig. 5) in the filtrates were measured by HPLC using an HPX-87H ion exclusion column (BioRad). Analyses were carried out at room temperature using 5 mM H2SO4 as the eluent, at a flow rate of 0.6 ml min−1. The products were monitored by continuous-flow detection of radioactivity using a Flo-One β detector (Packard).
Results are given as means ± SD of triplicate samples. Differences observed between culture conditions were checked for significance by the paired t-test, assuming equal variances and considering both sides of distribution. Differences were significant at P < 0.05.
Sequence determination of glmS from B. longum DSM 20219 and B. bifidum DSM 20082
The putative glmS genes of B. longum DSM 20219 and B. bifidum DSM 20082 were amplified by PCR using chromosomal DNA as templates and primer pairs glmS ATG (5′-ATG TGT GGA ATC GTT GGA TAC GCG G-3′)/glmS TAA (5′-TTA CTC AAC GGT CAC GGA CTT GGC G-3′) and HS2 (5′-TAG GCT GGT GAA CCA TGT GTG GAA TCG TTG-3′)/R1 (5′-GCG GAT CCT TAC TCA ACG GTC ACG GAC TT-3′), for B. longum and B. bifidum, respectively. PCR was routinely performed using the Taq PCR Core kit (Qiagen) with the following amplification conditions: 94°C for 5 min followed by 30 cycles of 94°C (1 min), 60°C (1 min) and 72°C (2 min) and an additional extension time at 72°C (5 min). PCR amplification products were cloned in pCR2.1 DNA (TOPO cloning kit, Invitrogen).
The BLAST program was used for sequence homology searches in the National Center for Biotechnology Information GenBank database (http://www.ncbi.nih.nlm.gov; Altschul et al. 1997). The accession numbers of the two available B. longum genome sequences used in this study are NC_004307 (B. longum NCC2705 complete genome sequence) and NZ_AABM00000000 (B. longum DJO10A, unfinished sequence, whole genome shotgun sequencing project). Amino acid sequence similarity and sequence alignment were analysed with Antheprot (http://antheprot-pbil.ibcp.fr).
Expression and purification of the B. longum GlmS protein
The putative glmS gene was amplified by PCR from B. longum DSM 20219 genomic DNA with primers GlmS(PciI)ATG (5′-GGC TGG TGA CAC ATG TGT GGA ATC GTT GG-3′, PciI restriction site underlined) and GlmS(XhoI)TAA (5′-CAC ATA CCG ATG TCT CGA GCT CAA CGG TCA C-3′, XhoI restriction site underlined). The amplicon was cleaved and inserted between the NcoI and XhoI sites in the expression vector pET28b(+) (Novagen, Madison, WI, USA), creating an in-frame fusion between the 3′ end of the putative glmS gene and the six histidine codons. A 5′ in-frame fusion was also created using primers GlmS(NdeI)ATG (5′-GGC TGG TGA CAT ATG TGT GGA ATC GTT GG-3′, NdeI restriction site underlined) and GlmS(Bam HI)TAA (5′-CAC ATA CCG GAT CCG CTT ACT CAA CGG TCA CG-3′, Bam HI restriction site underlined). The resulting plasmids, pET28b-GlmSCtHis and pET28b-GlmSNtHis, were transformed into E. coli BL21(DE3). The presence of the correct insert and orientation were confirmed by restriction enzyme digestion and sequencing.
The recombinant His-tagged GlmS proteins were over-produced from cells growing exponentially at 37°C in LB medium containing kanamycin (50 μg ml−1). At an OD600 of 0.6, glmS gene expression was induced by the addition of IPTG to a final concentration of 1 mM. Incubation was continued overnight, after which time the cells were harvested by centrifugation. Cells were washed and resuspended in sonication buffer (0.3 M NaCl, 10 mM sodium phosphate buffer, pH 7). Cells were disrupted by sonication, and the cellular debris was removed by centrifugation. The supernatant was applied to a 4 ml cobalt-based affinity column (BD Biosciences, Palo Alto, CA, USA) equilibrated in sonication buffer, and the column was washed with 10 volumes of the same buffer. The His-tagged GlmS proteins were eluted with sonication buffer containing 150 mM imidazole. The purification process was followed by monitoring A280 and SDS-PAGE (Laemmli 1970). Protein concentrations of cell extracts were determined using the method of Bradford, with bovine serum albumin as a standard (Bradford 1976).
Recombinant GlmS enzyme assays and definition of enzymatic activity unit
The standard assay mixture (200 μl) contained 25 mM imidazole, 8 mM fructose-6-phosphate and 17 mM glutamine (pH 6.9, 37°C). After pre-incubation at 37°C, the reaction was initiated by the addition of an aliquot of purified recombinant His-tagged GlmS (13 μg) and incubated for 150 min. Enzyme activity was quantified by measuring glutamate production (Milewski, 2002) using the l-glutamic acid kit (R-Biopharm). This assay was used throughout the entire study. Unless otherwise specified, enzyme activity was expressed in micromoles of glutamate formed per minute and specific activity in micromoles of glutamate formed per minute per mg of protein.
To detect glucosamine-6-phosphate, a High Performance Anion-Exchange Chromatography (HPAEC) technique was used. This analytical system consists of a Dionex Bio-LC GPM-II quaternary gradient module (Dionex Corporation, Sunnyvale, CA, USA) linked to an eluent degas (He) module, a Dionex Carbopac PA-100 column (4 × 250 mm) in combination with a Carbopac PA-100 guard column (3 × 25 mm). Aliquots of the assay, prepared as described above, were loaded onto the column pre-equilibrated with 0.1 M NaOH. Phosphorylated monosaccharides were eluted using the following gradient of sodium acetate in 0.1 M NaOH: 0–10 min, 0 M; 10–30 min, 0–0.5 M. All gradients were terminated with 0.5 M sodium acetate for 5 min, followed by reequilibration with 0.1 mM NaOH for 15 min. The effluent (flow-rate of 1 ml min−1) was monitored using a Dionex PED detector containing a gold electrode with an Ag–AgCl reference electrode, in the pulsed amperometric detection (PAD) mode. Enzyme-specific activity was expressed in micromoles of glucosamine-6-phosphate formed per minute per mg of protein.
To measure glutaminase activity the standard assay mixture described above was used but with the exclusion of fructose-6-phosphate. Glutaminase activity was quantified by measuring glutamate production at 37°C using the l-glutamic acid kit (R-Biopharm).
To measure isomerase activity the standard assay mixture described above was used but with the exclusion of glutamine. Isomerase activity was followed by the measurement of glucose-6-phosphate production at 37°C using the d-glucose kit (R-Biopharm).
pH and temperature optima of recombinant His-tagged GlmS
The optimal pH for recombinant GlmS activity was determined at 37°C in either a 100 mM sodium citrate or sodium phosphate buffer over a pH range of 3–5.5 and 5–9, respectively. The effect of temperature on GlmS activity was determined by measuring the enzyme activity at different temperatures in the range of 4 to 60°C. The standard reaction conditions described above were used.
Kinetic parameters of recombinant GlmS
Kinetic parameters were studied using a range of glutamine concentrations (0.1 to 20 mM). Activity was estimated at 55°C and pH 7 following procedures described above. The kinetic constants Km and Vmax were calculated from Hanes–Wolf plots.
Results and discussion
N-acetylglucosamine auxotrophy and metabolism
It has been suggested that N-acetylglucosamine is an essential growth factor for B. bifidum (Gygory et al. 1954). Furthermore, N-acetylglucosamine and indeed other N-substituted derivatives of d-glucosamine have been shown to have growth promoting activity for B. bifidum var. pennsylvanicus (Lambert and Zilliken 1965; Veerkamp 1969). In order to investigate this further the growth patterns of a number of Bifidobacterium spp. were monitored in the presence and absence of N-acetylglucosamine.
Bifidobacterium bifidum DSM 20082 and B. longum DSM 20219 were grown on modified Garches medium supplemented with fructose. While B. longum can grow on this medium, B. bifidum DSM 20082 was not able to grow with fructose as the sole carbon source. Supplementation of this medium with N-acetylglucosamine was required for growth of B. bifidum, thus leading to the conclusion that B. bifidum DSM 20082 is auxotrophic for N-acetylglucosamine. Indeed, N-acetylglucosamine auxotrophy was observed for all B. bifidum strains tested (strains tested are described in Table 1). In contrast, B. adolescentis, B. pseudocatenulatum, B. dentium, B animalis and B. breve strains (Table 1) did not require supplementation with N-acetylglucosamine for growth. It should be noted that, apart from an extended lag phase when grown on N-acetylglucosamine as the sole carbon source, there was no significant difference in the overall growth patterns of B. bifidum DSM 20082 when grown on medium supplemented with both fructose and N-acetylglucosamine compared with solely N-acetylglucosamine (data not shown). Furthermore, fructose and N-acetylglucosamine were metabolized simultaneously, although the presence of fructose led to a reduction in N-acetylglucosamine metabolism.
Sequence analysis of glmS from B. longum DSM 20219
Analysis of the two available B. longum genome sequences revealed the presence of a putative glutamine fructose-6-phosphate amidotransferase-encoding gene [corresponding to genes BL1175 (NP_696344) and 2525 (ZP_00121014), respectively] on the B. longum NCC2705 (Schell et al. 2002) and DJO10A genomes. This 1,893 bp ORF encodes a predicted protein of 630 amino-acids with a calculated molecular mass of 68.6 kDa (Mwcalc, http://www.infobiogen.fr). This falls within the size range (589–716 amino acids or 64–80.6 kDa) for GlmS proteins reported for other organisms, both prokaryotic and eukaryotic (Milewski 2002).
The putative glmS sequence of B. bifidum DSM 20082 demonstrates 99.9 and 99.8% nucleotide and amino acid identity, respectively, with that of B. longum DSM 20219. An alignment of the B. longum DSM 20219 GlmS sequence with all the available bifidobacterial putative GlmS sequences revealed a high level of conservation across species, with one exception—the putative GlmS of B. adolescentis (NZ_BAAD00000000) demonstrates only 90% a.a. identity and lacks the 28 N-terminal amino acids (Fig. 3).
Due to the key role played by GlmS in linking the biosynthetic pathway for N-acetylglucosamine to the hexose metabolic pathway, it was decided to perform a biochemical characterization of the recombinant bifidobacterial GlmS and in particular to monitor glucosamine synthesis and GlmS enzyme activities.
Purification of B. longum GlmS and biochemical characterisation of recombinant GlmS activity
This study encompasses a preliminary characterization of N-acetylglucosamine metabolism in bifidobacteria including characterization of a bifidobacterial GlmS, a key enzyme linking the d-fructose-6-phosphate shunt of bifidobacteria to the biosynthetic pathway for cell wall peptidoglycan precursors. The GlmS of B. longum DSM 20219 was purified and shown to possess a glutamine fructose-6-phosphate amidotransferase activity, leading to the conversion of fructose-6-phosphate and glutamine to glucosamine-6-phosphate, a precursor of N-acetylglucosamine. The observation of an independent glutaminase activity (i.e. the ability to convert glutamate to glutamine in the absence of fructose-6-phosphate) is of particular interest because analysis of the available B. longum NCC2705 genome sequence revealed this organism may lack the coding potential for some of the enzymes associated with glutamine metabolism.
For all organisms in which this enzyme has been characterized, GlmS functions in partnership with glucosamine-6-phosphate deaminase (NagB), a catabolic enzyme converting glucosamine-6-phosphate to fructose-6-phosphate and ammonia (Fig. 2). The genes nagA (encoding an N-acetylglucosamine-6-phosphate deacetylase) and nagB can be found in the annotated genome of B. longum NCC2705 and are organized, as in other organisms, adjacent to each other on the genome. Although NagB can also catalyse the reverse biosynthetic pathway (Vogler et al. 1989), it is kinetically unfavourable (Bustos-Jaimes et al. 2005; Calcagno et al. 1984). The biosynthetic and degradative steps for glucosamine and N-acetylglucosamine are subjected to tight regulation (Alvarez-Anorve et al. 2005; Plumbridge et al. 1993; Vogler et al. 1989). Studies of E. coliglmS and nag mutants reveal the over-expression and/or deregulation of nagB leading to sufficient glucosamine biosynthesis for cell growth and survival (Plumbridge 1995; Plumbridge et al. 1993; Vogler et al. 1989). Growth experiments performed in this study indicate that, while fructose did not affect assimilation of N-acetylglucosamine into cell walls, it did lead to the reduced assimilation of N-acetylglucosamine via the fructose-6-phosphate shunt. Preliminary growth experiments monitoring the assimilation of fructose in B. bifidum (unpublished data) suggest ineffective conversion of fructose-6-phosphate to N-acetylglucosamine. It will therefore be of interest in future studies to investigate the interplay between GlmS and NagB in bifidobacteria, in order to determine whether differences amongst bifidobacterial species may account for the observed N-acetylglucosamine auxotrophy in B. bifidum.
This work was supported in part by the Centre National de la Recherche Scientifique (Unité de Glycobiologie Structurale et Fonctionnelle, UMR CNRS-USTL 8576; Director, Dr Jean Claude Michalski), by the Université des Sciences et Technologies de Lille and by the Region Nord-Pas de Calais (CPER 2000–2006).