| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular & Cellular Proteomics 5:1105-1118, 2006.
© 2006 by The American Society for Biochemistry and Molecular Biology, Inc.
,
,¶,¶,¶
,
From the Beijing Institute of Biotechnology, State Key Laboratory of Pathogen and Biosecurity, 100071 Beijing, China, College of Food Science and Engineering, Northwest Sci-Tech University of Agriculture and Forestry, 712100 Yangling, China, ** National Center of Biomedical Analysis, 100850 Beijing, China, and || National Institute for the Control of Pharmaceutical and Biological Products, 100050 Beijing, China
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXPERIMENTAL PROCEDURESRESULTS AND DISCUSSIONREFERENCES |
|---|
Bifidobacterium longum is a strict fermentative anaerobe, deriving its principal source of energy from substrate level phosphorylation during glycolysis. In the human GIT, bifidobacteria function as scavengers in the large intestine and possess a wide range of catabolic pathways that confer a growth advantage where readily fermentable carbohydrates are in short supply (2, 5). The ability of Bifidobacterium spp. to survive and persist in this competitive environment is made possible by use of negative transcriptional regulation as a flexible control mechanism in response to nutrient availability and diversity as well as predicted genetic features such as exo- and endo-glycosyl hydrolases and high affinity oligosaccharide transporters. These features likely help B. longum compete for uptake of structurally diverse oligosaccharides released from digestion of plant fibers. Several studies conducted on fructose-containing polymers as potential selective substrates for colonic bacteria have provided evidence that bifidobacteria are able to ferment these carbohydrates, particularly the short chains of ß-(2
1)-linked fructosyl units (6–10). Recently bifidobacteria were shown to possess only one pathway for the metabolism of glucose, the F6P phosphoketolase (F6PPK) pathway, otherwise known as the bifid shunt (11, 12). Analyses of fructose fermentation patterns suggested that B. longum could use D-fructose as the sole carbon source (11, 12); however, the uptake mechanism of fructose into the cell and proteins regulated by fructose remained to be defined.
The 2.26-Mb genome of B. longum strain NCC2705 was sequenced in 2002, predicting 1729 ORFs (2). The sequences were revised to gi:23464628 (containing 1727 ORFs) according to the National Center for Biotechnology Information (NCBI) website in 2005. Unlike genome studies, investigations at the proteomic level provide insights into protein abundance and/or post-translational modifications. It is also one of the best methods of investigating basic biological processes such as pathogenesis, physiology, and metabolic mechanisms (13, 14). In recent years 2-D electrophoresis has been used in proteomic studies of many bacteria, including Haemophilus influenzae (15–18), Mycobacterium tuberculosis (19–21), Escherichia coli (13, 22–24), Helicobacter pylori (25, 26), Bacillus subtilis (27, 28), Mycoplasma pneumoniae (29), Lactococcus lactis (30–33), Shigella flexneri (34), Agrobacterium tumefaciens (35), Vibrio cholerae (36), and Brucella melitensis (37).
Vitali et al. (38) reported the identification of 136 proteins in Bifidobacterium infantis B107 by using multidimensional chromatography and tandem mass spectrometry. To date, a two-dimensional reference map to correlate protein expression to physiological changes in B. longum has not been generated. In this study we focused on generating a base-line signature or cartographic reference of the B. longum strain NCC2705 proteome by identifying ORF expression and investigating the mechanism of fructose translocation, uptake, and utilization.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTS AND DISCUSSIONREFERENCES |
|---|
Preparation of Whole Cell Protein Extract—
B. longum NCC2705 cell pellets (about 0.30 g) were resuspended in 5 ml of lysis buffer (7 M urea, 2 M thiourea, 4% (w/v) CHAPS, and 50 mM DTT) containing complete protease inhibitors (Roche Applied Science). The cells were sonicated for 10 min on ice using a Sonifier 750 (Branson Ultrasonics Corp., Danbury, CT) with the following conditions: 2 s of sonication with a 2-s interval, set at 35% duty cycle. After adding 2.5 mg of RNase (Promega, Madison, WI) and 100 units of RQ1 DNase (Promega), the cell lysate was incubated for 1 h at 15 °C to solubilize proteins and centrifuged for 20 min at 20,000 x g to pellet the insoluble components. The supernatant was collected, and protein concentration was measured using the PlusOne 2-D Quant kit (Amersham Biosciences), and 1-mg aliquots were stored at –70 °C.
Two-dimensional Polyacrylamide Gel Electrophoresis—
IEF was performed by using IPG strips (18 cm; Amersham Biosciences). The first dimensional isoelectric focusing was carried out as described previously (42). For the second dimension vertical slab SDS-PAGE (12.5%) was performed for about 4 h at 30 mA/gel using a Bio-Rad Protean II Xi apparatus (Bio-Rad). The gels were stained with Coomassie Brilliant Blue G-250 (Amresco, Solon, OH) and were scanned with ImageScanner (Amersham Biosciences). Image analysis was carried out using ImageMaster 2D Platinum software (Amersham Biosciences). Images from two independent cultures were compared. The relative volume of each spot was determined from the spot intensities in pixel units and normalized to the sum of the intensities of all the spots of the gel. Proteins displaying at least 3-fold volume variations in response to glucose in all experiments were considered in this work.
In-gel Protein Digestion and MALDI-TOF-MS—
The Coomassie-stained protein spots of interest were cut out, and in-gel protein digestion was performed as described previously (43). Peptides from digested proteins were resolubilized in 2 µl of 0.5% TFA. Peptide mass fingerprinting (PMF) measurements were performed on a Bruker ReflexTM III MALDI-TOF mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany) working in reflectron mode with 20 kV of accelerating voltage and 23 kV of reflecting voltage. A saturated solution of 
-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 0.1% TFA was used for the matrix. A total of 2 µl of the matrix solution and sample solution were mixed in a 1:1 (v/v) ratio and applied onto the Score 384 target well. Mass accuracy for PMF analysis was 0.1–0.2 Da with external calibration, and internal calibration was carried out using enzyme autolysis peaks; resolution was 12,000. PeakClean (www.proteomics.com.cn/tools/PkClean/) was used to remove contaminant peaks including matrix peaks, solvent peaks, and enzyme autolysis peaks.
Nanospray ESI-MS/MS—
The peptide solution after in-gel protein digestion was collected, lyophilized, reconstituted in 30 µl of 30% acetonitrile containing 0.1% TFA, and then desalted using ZipTip C18 pipette tips (Millipore, Bedford, MA). Electrospray ionization (ESI-MS/MS) was carried out with a hybrid quadrupole orthogonal acceleration tandem mass spectrometer (Q-TOF2, Micromass Ltd., Manchester, UK). The capillary voltage in MS and MS/MS experiments was set to an average of 900 V, and the sample cone voltage was 30 V. A microchannel plate detector was applied with 2200 V. The collision gas was argon with a pressure of 0.1 megapascal (Mpa), and collision energy was 50 V. Glu-Fibrinopeptide was used to calibrate the instrument in the MS/MS mode. MS/MS spectra were transformed using MaxEnt3 (MassLynx, Micromass Ltd.), and amino acid sequences were interpreted manually using PepSeq (BioLynx, Micromass Ltd.).
Protein Identification—
Database searches were performed by using the software Mascot (Matrix Science Ltd.) licensed in-house against the database of B. longum NCC2705 (Version 050126, 1727 sequences) and NCBInr (Version 050623, 2,564,994 sequences); the research results were checked using Mascot with free access on the Internet (www.matrixscience.com). Monoisotopic masses were used to search the databases, allowing a peptide mass accuracy of 100 ppm (300 ppm in some cases) and one partial cleavage. Oxidation of methionine and carbamidomethyl modification of cysteine were considered. For unambiguous identification of proteins, more than five peptides must be matched, and the sequence coverage must be greater than 15%. Fourteen spots representing 11 protein entries presenting a low covering percentage (15–20%) and corresponding to bifid shunt were reconfirmed by nanospray ESI-MS/MS.
The Codon Adaptation Index and Grand Average Hydropathicity—
The codon adaptation index (CAI) of all ORFs of B. longum NCC2705 was generated by CodonW software (44) in two steps. First a correspondence analysis of the codon usage of all ORFs was performed, and 1% of the total genes were used for calculating the 
values. Then the CAI values were calculated by software CodonW using those values. The same software was used for the calculation of the GRAVY of each protein as described previously (31).
Cloning the xfp Gene from B. longum NCC2705 and ATCC15707 and Phosphoketolase Assay—
Total cellular DNA was prepared from B. longum NCC2705 or ATCC15707 as described by Meile et al. (45). The oligonucleotide primers used for cloning xfp gene, xfp P1, and xfp P2 are listed in Table I. The PCR was carried out as follows: initial PCR activation step, 95 °C for 5 min; amplification, 35 cycles of 95 °C for 1 min, 55 °C for 1 min, and 72 °C for 3 min; final extension step, 72 °C for 10 min. The resulting PCR products were sequenced, and sequences were compared with sequences in the GenEMBL database by the algorithm of Pearson and Lipman (FastA and TFastA) (46). Phosphoketolase activity was measured spectrophotometrically as ferric acetyl hydroxamate produced from the enzymatically generated acetyl phosphate by the previously described procedures of Racker (47) and Meile et al. (45).
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTS AND DISCUSSIONREFERENCES |
|---|
|
The 899 matched spots on the corresponding micropreparative electrophoresis gels were excised and destained. After in-gel trypsin digestion, MALDI-TOF-MS and/or ESI-MS/MS were performed to identify proteins. Among the 899 protein spots processed, 708 spots were successfully identified, representing 369 protein entries by PMF; 14 spots representing 11 protein entries were reconfirmed by ESI-MS/MS (Supplemental Table 3). The identified spots are labeled on the integrated map of pH 4–7 (Fig. 2), and identified spots are listed in Supplemental Tables 1–3. Maps and information of all identified proteins are available at www.mpiib-berlin.mpg.de/2D-PAGE and www.proteomics.com.cn.
|
|
22% of the predicted proteins. The possible explanations are that 1) many proteins are inherently difficult to be identified by these methods (low abundance proteins, hydrophobic proteins, and proteins with high or low molecular weight), or 2) B. longum NCC2705 proteins occurred at several locations of the 2-DE gels. The cellular localizations of all 369 identified proteins predicted by PSORT Version 2.0 (www.psort.org) are listed in Supplemental Table 1. 320 proteins identified are cytoplasmic; 17 proteins were predicted in the cytoplasmic membrane, two proteins were predicted in the cell wall, and 30 spots had unknown cellular localization. Moreover eight detected proteins (BL0033, BL0077, BL0141, BL0352, BL1345, BL1386, BL1638, and BL1714) were predicted to have a signal peptide. Surprisingly four proteins were annotated as extracellular proteins suggesting that they might be interact tightly with cell surface components.
Extracytoplasmic proteins play critical roles in establishing and maintaining interactions between a microbe and its environment. Many cell wall surface proteins of Gram-positive bacteria are covalently anchored to the cell wall by a mechanism requiring a C-terminal anchoring motif, consisting of a conserved amino acid sequence Leu-Pro-X-Thr-Gly (LPXTG, where X is any amino acid) (51). In bifidobacteria, extracytoplasmic proteins might mediate important host interactions, such as adhesion, nutrient availability, immune system modulation, and pathogen inhibition. Interestingly two proteins (spot 88, BL1132; and spot 747, BL1064) with unknown cellular location and one protein (spot 403, BL0603) predicted to localize to cytoplasmic membrane displayed a clear Gram-positive cell surface anchor motif. This evidence suggests that these proteins may be involved in host interactions.
CAI, GRAVY Value of Each Protein, and High Abundance Proteins—
Synonymous codons are not generally used at equal frequencies. The degree of codon bias is related to both the content of the isoacceptor tRNAs and the level of gene expression as the result of selection to increase translational efficiency. Thus, highly expressed genes tend to predominantly use synonymous codons with the most abundant tRNA ("major" or preferred codons), whereas weakly expressed genes show a more frequent use of the "minor" or unpreferred synonymous codons (44, 52). Accordingly highly expressed genes show a base composition that departs more strongly from that expected by a mutational equilibrium.
CAI estimates the degree of synonymous codon adaptation in a coding region compared with the optimal usage. A value of 1.0 indicates the maximum codon use fit, and values <1.0 indicate use of less preferred codons. Fig. 4 compares the CAI distributions of genes coding for the proteins identified on the pH 4–7 gel with those of all proteins. The genes with the highest CAI encode glycolytic enzymes or proteins of the translational apparatus. The proteins encoded by genes with a CAI value <0.5 account for 77.0% of the total proteins and only 55.2% of the currently identified proteins. All genes having a CAI value >0.75 were identified on the 2-D maps. These results demonstrated that proteins encoded by genes with a high CAI were abundant and easily identified. Interestingly three hypothetical proteins with a CAI value below 0.1 were also identified (spot 217, BL1039; spot 329, BL0667; and spot 7, BL0256). The abundance of all spots in the 2-D gel is listed in Supplemental Table 1, and the 50 most abundant proteins are labeled as b1–b50. Proteins with a CAI value above 0.5 are involved in energy metabolism, fatty acid and phospholipid metabolism, protein synthesis, and cellular processes. They represented a large part of the most abundant proteins. Translation elongation factor Tu and universal stress protein were the top two abundant proteins. Our results were similar to the results from E. coli (13), L. lactis (33), and B. subtilis (28).
|
|
Preliminary studies of the biochemistry, genetics, and regulation of these essential enzymes in the nutritionally important bifidobacteria showed that phosphoketolases are key thiamine diphosphate (ThDP)-dependent enzymes of bifid shunt. Phosphoketolase links several metabolic routes to the bifid shunt, including N-acetylhexose fermentation, galactose catabolism (the Leloir pathway), and peptidoglycan biosynthesis. Similar results have been reported in other microorganisms (10, 11, 53). In bifidobacteria, there is evidence for the existence of two distinct F6P phosphoketolase enzymes (34, 35): a specific enzyme (F6PPK) solely for F6P found in humans, such as Bifidobacterium dentium, and a dual substrate xylulose-5-phosphate/F6P phosphoketolase (Xfp; pI, 5.0; 92,469 Da) found in animals, such as Bifidobacterium globosum. The dual specificity xylulose 5-phosphate/F6P phosphoketolase is encoded by the gene xfp, which was first described in Bifidobacterium animalis subsp. Lactis by Meile et al. (45). Native Xfp was predicted to be a homohexamer with a molecular mass of 550,000 Da. The subunit size upon SDS-PAGE (90,000 Da) matches the predicted size (92,469 Da) calculated from the amino acid sequence of the isolated gene (named xfp), which encodes 825 amino acids (11). In the genome of infant-derived NCC2705 strain, a single copy of xfp is identified at locus BL0959.
Interestingly 21 spots of Xfp (molecular weight, 92,697; pI, 5.06) differing in charge and mass were identified in the present work by MALDI-TOF and ESI-MS/MS (see Supplemental Table 3). The proteins were distributed into pH ranges 4.2–4.4 and 6.0–6.2 in the gel. To study the regulation and expression of the gene coding for the enzyme, we cloned and sequenced the ORF (2475 bp; 825 amino acids) from B. longum NCC2705 and ATCC15707. After analyzing homology to sequences in GenBankTM using the Basic Local Alignment Search Tool (BLAST) available on the National Center for Biotechnology Information server, we found that the xfp sequences from B. longum NCC2705 and ATCC15707 are very similar to the 10 xfp coding sequences from other bifidobacteria. The Xfp amino acid sequence from B. longum NCC2705 and ATCC15707 contains a ThDP-dependent enzyme signature sequence YGX5PX3VX2IXGDGE (amino acids 165–184), which closely matches the consensus [LIVMF]-GX5PX4VXIXGDG-[GSAC] (PROSITE PS00187). (In the PROSITE database, ambiguities are indicated by listing the acceptable amino acids for a given position between square brackets.) Another motif (GDGX24–27NN), which is common to ThDP-binding enzymes like acethydroxy acid synthases, transketolases, E1 (decarboxylase) components of 2-ketoacid and acetoin dehydrogenases, and others, is also present in B. longum NCC2705 Xfp in a modified form, GDGEX30D (amino acids 181–215). In B. longum ATCC15707, the motif is present in Xfp in conserved form, GDGEX30N. The peptide sequence of the Xfp protein from NCC2705 derived by ESI-MS/MS (Fig. 6) indicates that Xfp has a mutation in amino acid position 215 resulting in Asp instead of Asn. This modification corresponds to an increase in the global negative charge due to the acidic residue (Asp215). Modified Xfp showed activity similar to that of the conserved Xfp from ATCC15707. The enzymatic activities in cell extracts were 22 and 24 nmol min–1 mg of protein–1, respectively. However the cDNA sequence of xfp obtained by RT-PCR was identical to the original gene (data not shown). We speculate that the natural variant might likely be formed by post-translational modification, which was confirmed with further investigation. We suggest that the structural variants of Xfp could be used to provide information for the structure, property, and function of the enzyme.
|
Bifidobacteria colonize at the lower GIT, an environment poor in mono- and disaccharides because they are consumed by the host and by microflora in the upper GIT. As previous studies suggested, B. longum utilizes a variety of plant-derived dietary fibers, such as arabinogalactans and gums (2, 54). We identified 44 proteins assigned to the carbohydrate transport metabolism category, four of which were related to oligosaccharide hydrolysis: BL0682 (spot 293, xylan esterase), BL0978 (spot 758, ß-galactosidase, EC 3.2.1.23), and BL0982/BL1573 (spots 732/513, isoamylase, EC 3.2.1.68); two other proteins involved in the fermentation of nondigestible dietary carbohydrates were identified: BL0673 (spots 714, 716, and 727; ATP-binding protein of ATP-binding cassette (ABC) transporter for sugars) and BL1656 (spots 261, 313, and 812; phosphoglycerate mutase, EC 5.4.2.1).
Amino Acid Metabolism—
B. longum has genes involved in synthesis of at least 19 amino acid from NH4 and major biosynthetic precursors including phosphoenolpyruvate, oxaloacetate, oxoglutarate, and fumarate provided by a partial Krebs cycle lacking fumarase, oxoglutarate dehydrogenase, and malate dehydrogenase (2). These genes have been predicted in the genome of B. longum NCC2705. In the present map, 101 proteins (about 32% of all functional proteins identified) related to amino acid metabolism were identified according to the results of a search in the database of the Kyoto Encyclopedia of Genes and Genomes (www.kegg.com/kegg/pathway.html). We identified a majority of proteins corresponding to amino acid metabolism pathway, urea cycle, and metabolism of amino groups.
Transamination reactions have recently attracted attention because they are the first step for the synthesis of important flavor or aroma compounds in amino acid catabolism pathways (55). We found five aminotransferases: AspC (aspartate transaminase, EC 2.6.1.1, BL1286), HisC (histidinol-phosphate transaminase, EC 2.6.1.9, BL1296), SerC (probable phosphoserine transaminase, EC 2.6.1.52, BL1660), IlvE (probable Branched-chain-amino-acid transaminase, EC 2.6.1.42, BL0852), and probable aminotransferase (aspartate transaminase, EC 2.6.1.1, BL0783). These enzymes possess the catalytic potency to deaminate aspartate, histidinol phosphate, branched-chain amino acids, and other amino acids. But we did not detect the proteins related to lysine degradation.
Stress Proteins—
B. longum is moderately aerotolerant and encodes homologs of enzymes that repair oxidative damage. The enzymes responsible for minimizing the toxicity of active oxygen species were predicted in the genome of B. longum NCC2705; NADH peroxidase and superoxide dismutase were absent, but a NADH oxidase was present (2). Under our experimental conditions we found alkyl hydroperoxide reductase (ahpC, BL0615), a protein that can reverse oxidative damage to proteins and lipids. We hypothesize that AhpC might play an important role in reversing oxidative damage to proteins.
Heat shock proteins corresponding to DnaK (BL0520), GroEL (BL0002), and GroES (BL1558) can also be induced by salt stress, mild acid treatment, and UV irradiation in B. longum NCC2705. They were identified as highly abundant proteins in this study.
Bifidobacteria are a dominant population of the human intestinal microflora. One of the most important metabolic activities is deconjugation of bile salts, which occurs naturally in human intestines. The responsible enzyme, choloylglycine hydrolase (BL0796, EC 3.5.1.24), which catalyzes the hydrolysis of glycine- and/or taurine-conjugated bile salts into amino acid residues and deconjugated bile salts (bile acids), was identified in our study. It might play a role in B. longum resistance to toxic levels of bile salts in the gastrointestinal environment. This enzyme may be medically relevant because in recent years the possibility of using bile salt deconjugation by lactic acid bacteria to decrease serum cholesterol levels in hypercholesterolemic patients or to prevent hypercholesterolemia in individuals with normal cholesterol levels has received increasing attention (56).
Peptidoglycan Precursor Assembly and Nucleotide Metabolism—
Nine key enzymes identified in the present map, glmU, nagA, ddlA, glnA1, glnA2, murA, murB, murD, and murF, are involved in the synthesis of a cytoplasmic peptidoglycan precursor (UDP-MurNAc-pentapeptide) (33, 57). MurA (BL1267, UDP-N-acetylglucosamine 1-carboxyvinyltransferase) and MurB (BL1561, UDP-N-acetylenolpyruvoylglucosamine reductase) are likely to catalyze the synthesis of UDP-MurNAc from UDP-GlcNAc. MurD (BL1321, UDP-N-acetylmuramoylalanine-D-glutamate ligase) and MurF (BL1319, UDP-N-acetylmuramoylalanyl-D-glutamyl-2,6-diaminopimelate-D-calanyl-D-alanine) can add D-glutamate and the dipeptide D-Ala-D-Ala, respectively, to form the UDP-MurNAc-pentapeptide. However, we did not detect MurC and MurE.
B. longum has all genes required for pyrimidine and purine nucleotides biosynthesis from glutamine. The present map contains 29 enzymes involved in the synthesis of purine (15 proteins) and pyrimidine (14 proteins) nucleotides. As a dairy fermentative bacterium, bifidobacteria need trace amounts of free purine bases in the medium (guanine, hypoxanthine, and guanosine (58)). These purines positively regulate the enzymes required for de novo synthesis of precursors for ATP and GTP, but we did not identify these proteins. Therefore, we hypothesize that De Man-Rogosa-Sharpe medium provided the cell with the necessary amount of purine bases and/or nucleotides, which stimulated the growth of Bifidobacterium.
Discrepancies of Theoretical Prediction and Experimental Results—
From the statistical results, proteins with pI 4–6 and of 10–70 kDa are more likely to be identified on the 2-D gel. The gel-estimated pI and molecular weight analyzed by using the ImageMaster 2D Platinum software matched well with those obtained from the theoretical predictions. Discrepancies between gel-estimated and theoretical masses might result from post-translational proteolytic processing and modification. These phenomena have been reported in similar studies (34, 59). Different from their theoretical pI values, several proteins appeared in the relatively acidic region (such as spots 60, 321, and 525) or in the alkaline region (such as spots 86, 484, and 792). This might result from the cleavage of acidic regions or alkaline regions and phosphorylation.
Proteomic analysis enables a global inspection of post-translational modifications as many of these events are manifested in changes of charge or molecular weight. Post-translational modifications are known to play an important role in eukaryotes, whereas less is known about their role in bacterial physiology. In this experiment, post-translational modifications appeared to be common in B. longum NCC2705. One hundred and fifteen proteins identified appeared in more than one spot on the gel. These proteins are listed in Supplemental Tables 1 and 2, and nine proteins were identified in more than 10 distinct spots. They include elongation factor Tu (BL1097), chaperone (BL0002), xylulose-5-phosphate/fructose-6-phosphate phosphoketolase (BL0959), transketolase (BL0716), ATP synthase ß chain (BL0357), LacZ (BL0978), pyruvate kinase (BL0988), transaldolase (BL0715), and elongation factor G (BL1098). Another 11 proteins were identified from more than five spots: choloylglycine hydrolase (BL0796), DppA2 (BL1386), DnaK protein (BL0520), phosphoglycerate kinase (BL0707), 6-phosphogluconate dehydrogenase, decarboxylating II (BL0444), ketol-acid reductoisomerase (BL0531), trigger factor chaperone (BL0947), acetate kinase (BL0969), enolase (BL1022), RNA polymerase 
-E factor (BL1357), and UDP-glucose 4-epimerase (BL1671). These 20 proteins, accounting for 30.8% of the total proteins identified, represent a large part of the most abundant proteins. We speculate that many of them should be biologically modified. The versatility of modification might reflect the need for its growth functions.
Hypothetical and Conserved Hypothetical Proteins Identified as Abundant Proteins—
Ninety-five proteins annotated as "hypothetical protein" including 38 conserved hypothetical protein and 57 proteins with unknown functions were identified that should be considered as real proteins. The presence of these proteins is interesting because up to now no function has been assigned to them. Proteome studies focusing on the presence of these proteins during growth phase, stress, or other processes may give insight into their cellular roles (35).
Proteomes of B. longum Strain NCC2705 Grown on Fructose and Glucose—
Bifidobacteriun spp. can survive and persist in a competitive environment of the human gastrointestinal tract. Several studies have proved that bifidobacteria are able to ferment fructose-containing polymers in particular the short chains of ß-21-linked fructosyl units (6–10). Although analyses of fructose fermentation patterns revealed that B. longum could use D-fructose as a sole carbon source, the events following the uptake of fructose into the cell and proteins regulated by fructose remained to be defined.
To identify the catabolic route allowing D-fructose fermentation, we first compared the proteomic profile of B. longum strain NCC2705 grown on fructose or glucose. Their 2-DE patterns are highly similar, and many landmark spots have counterparts. In comparison with glucose catabolism, the enzymes of bifid shunt in fructose catabolism were identified in this study, indicating that intracellular fructose and glucose are catabolized via the same degradation pathway. However, we observed a greater than 3-fold variation for the 18 identified proteins (Table II and Fig. 7A). These spots showed clear differences in intensity or position possibly caused by amino acid exchanges and/or small deletions. The possibility that some of the variants are due to different post-translational modifications cannot be excluded. To investigate the involvement of the carbohydrate substrate on expressed genes and the encoded proteins, we cultivated B. longum NCC2705 in modified Garches medium containing either D-glucose (2 g/liter) or fructose (2 g/liter) as the sole carbon source. Total RNAs were used as the template for expression analysis of interesting genes by RT-PCR (Fig. 7B). Comparison with an internal control (16 S rDNA) provided a semiquantitative measurement of the level of these genes expression. The intensities of the bands provided an approximation of their relative abundance. Overall the results of RT-PCR confirmed up- and down-regulation in the transcription level of interesting genes in cells grown on fructose and glucose, which were consistent with their protein expression.
|
|
As for proteins involved in phosphorylation of sugar, we identified a possible fructokinase (Frk, BL1339, protein in PfkB family of sugar kinases) responsible for the incorporation of intracellular fructose in two culture patterns. At the same time, the results of semiquantitative RT-PCR measurement demonstrated a higher level of frk transcription in fructose medium, suggesting that frk expression is subject to glucose-mediated repression and/or fructose induction. Our results agree with that of Caescu et al. (10), who suggest that fructokinase is the enzyme that is necessary and sufficient for the assimilation of fructose into this catabolic route in B. longum A10C.
Unfortunately we failed to identify glucokinase (glkA, EC 2.7.1.2) in cells grown on fructose or glucose. The RT-PCR results also showed a lower level of glkA transcription in cells grown on fructose or/and on glucose (data not shown). Proteins involved in phosphorylation of glucose are often regulators of catabolism of certain carbon (energy) sources. The catabolite repression of genes has already been reported for Bifidobacterium spp. (60). Thus, we thought GlkA was a low abundance regulator of sugar catabolism in B. longum NCC2705. Interestingly we identified a second gene putatively encoding a glucokinase (glk, EC 2.7.1.2) located 1.5 kb upstream of the frk locus in B. longum NCC2705 grown on fructose or glucose; the two genes (glk and frk) are simply separated by the gene encoding a putative regulator of the ROK (Repressor, ORF, Kinase) family. Thus, we deduced there were two glucokinases (GlkA and Glk) responsible for glucose phosphorylation, and an Frk was responsible for fructose phosphorylation in B. longum NCC2705.
Further Investigation of BL0033—
At present, no studies on BL0033 have been reported. To further investigate the expression of BL0033, we compared the proteomes and analyzed the transcription level of cells grown on fructose or glucose after 8 h (early exponential phase), 13 h (midexponential phase), and 16 h (end of exponential phase) according to the bacterial growth curves. The ratios of BL0033 to 16 S rDNA (RT-PCR band intensities, in arbitrary units) were determined as a function of the carbohydrate substrate and time of growth. As depicted in Fig. 8, A and B, the effect of fermentation time on expression of BL0033 was very clear. We observed a consistent up-regulation of intensity for these proteins in 2-D maps with time of growth; the volumes (%) of these spots were 13.9492% (8 h), 15.0830% (13 h), and 18.2231% (16 h). These experiments demonstrate a significantly higher level of BL0033 transcription and expression in cells grown on fructose than in those grown on glucose.
|
|
Surprisingly another ORF, which encodes an ATP-binding protein of an ABC transporter (BL0034) and flanks downstream of BL0033 in the NCC2705 genome, showed slight up-regulation in cells grown on fructose compared with glucose. We speculate that BL0033 and BL0034 are responsible for fructose uptake together in B. longum NCC2705. The mechanism of regulation of fructose translocation and the increase of fructose uptake still need to be characterized.
Conclusions—
Bifidobacteria are predominant commensal bacteria in the intestinal microflora and exert various beneficial effects on human health. Hence they are widely used as active ingredients in functional dairy-based products. Although bifidobacteria have been studied for over a century, a lack of genetic tools and uniformity among studies have prevented a comprehensive and coherent view of their biosynthetic capabilities. Our proteome analysis of B. longum NCC2705 extends previous studies in regard to the physiological characteristics, supports the hypothesis formulated by Schell et al. (2) of the adaptation to the human gastrointestinal tract, and more importantly confirms the expression of a large number of proteins related to their habitat (2).
In this study, reference maps of B. longum with variable pH gradients (such as pH 3–10, pH 4–7, pH 4.5–5.5, and pH 4–5) were constructed. A total of 899 spots were processed, and 708 spots representing 369 protein entries were identified by MALDI-TOF-MS and/or ESI-MS/MS. We described a significant number of cellular pathways (a global view of glycolysis and energy metabolism, amino acid metabolism, peptidoglycan precursor assembly, and nucleotide metabolism) related to important physiological processes at the proteome level. At the same time, CAI, GRAVY value of each protein identified, and high abundance proteins also were analyzed. The most abundant proteins include energy metabolism components, periplasmic ABC transporter proteins, and fatty acid and phospholipid metabolism enzymes, and other proteins involved in protein synthesis and cellular processes represent a large proportion. Ninety-five hypothetical proteins were experimentally identified. Maps and information about all identified proteins, which will be helpful for subsequent physiological studies of B. longum, are available at www.mpiib-berlin.mpg.de/2D-PAGE and www.proteomics.cn.
We also compared the proteomic profile of B. longum strain NCC2705 grown on fructose or glucose; results indicated that intracellular fructose and glucose are catabolized via the same degradation pathway. However, sugar-binding proteins specific to fructose (BL0033) had a 10-fold higher expression level in cells grown on fructose than on glucose. Five BL0033 isoforms differing in charge were identified by MALDI-TOF and ESI-MS/MS. We can conclude that the protein encoded by BL0033 is induced by fructose and may play an important role in the transport pathway of fructose. Moreover an ATP-binding protein of an ABC transporter (BL0034) was slightly up-regulated in cells grown on fructose than on glucose. We imagine that BL0033 and BL0034 together play a role in fructose uptake in B. longum NCC2705. These results were exciting and significant; little was previously known about the mechanism of fructose transport in B. longum NCC2705. These findings may document that fructose was not carried over the cytoplasmic membrane by a PEP:fructose PTS. The uptake of fructose into the cells most likely occurs via a specific transport system. Furthermore the expressed quantities of frk and Frk had significantly increased under these conditions, suggesting that frk expression is subject to glucose-mediated repression and/or fructose induction.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Published, MCP Papers in Press, March 20, 2006, DOI 10.1074/mcp.M500410-MCP200
1 The abbreviations used are: GIT, gastrointestinal tract; 2-D, two-dimensional; CAI, codon adaptation index; ABC, ATP-binding cassette; F6P, fructose 6-phosphate; F6PPK, F6P phosphoketolase; PMF, peptide mass fingerprinting; rDNA, ribosomal DNA; 2-DE, two-dimensional electrophoresis; ThDP, thiamine diphosphate; MurNAc, N-acetylmuramic acid; PEP, phosphoenolpyruvate; PTS, phosphotransferase system. 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
S The on-line version of this article (available at http://www.mcponline.org) contains supplemental material.
¶ These authors contributed equally to this work.
To whom correspondence may be addressed: Beijing Institute of Biotechnology, State Key Laboratory of Pathogen and Biosecurity, 20 Dongdajie St., Fengtai District, 100071 Beijing, China. Tel.: 86-10-66948836; Fax: 86-10-63833521; E-mail: wanghl{at}nic.bmi.ac.cn
To whom correspondence may be addressed: National Inst. for the Control of Pharmaceutical and Biological Products, 2 Tiantanxili, 100050 Beijing, China. Tel.: 86-10-67058416; Fax: 86-10-67058402; E-mail: zengming{at}nicpbp.org.cn
|
REFERENCES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTS AND DISCUSSIONREFERENCES |
|---|
-L-arabinofuranosidase from Bifidobacterium longum B667.
Appl. Environ. Microbiol.
69,5096
–5103
