OUP user menu

Effect of heat-shock and bile salts on protein synthesis of Bifidobacterium longum revealed by [35S]methionine labelling and two-dimensional gel electrophoresis

Kirsi Savijoki , Aki Suokko , Airi Palva , Leena Valmu , Nisse Kalkkinen , Pekka Varmanen
DOI: http://dx.doi.org/10.1016/j.femsle.2005.05.032 207-215 First published online: 1 July 2005


Experimental conditions for efficient protein radiolabelling and two-dimensional gel electrophoresis were developed for Bifidobacterium longum. Using these tools, protein synthesis in cells before and after heat-shock and bile salts treatment was investigated. Following heat-stress, 13 proteins were upregulated, of which HtrA, DnaK and GroEL were also moderately induced by bile salts, indicating close relationship between the heat and bile salts responses in bifidobacteria. Our work indicated that, as a consequence of prolonged heat-stress, HtrA undergoes sequential modification and proteolysis, and that this mechanism could be employed by bifidobacteria to respond to heat-stress.

  • Bifidobacterium longum
  • Stress
  • 2D-PAGE
  • [35S]labelling

1 Introduction

Bifidobacterium genus includes Gram-positive, strictly anaerobic bacteria, which naturally colonize the gastrointestinal (GI) tracts of humans and animals. They are increasingly incorporated into dairy foods as health-promoting bacteria, i.e., probiotics and have thus become an important objective in a number of studies aiming to obtain scientific evidence explaining the health-beneficial effects [[[]. Despite the increasing use of bifidobacteria as probiotics, the mechanisms by which these bacteria exert health benefits are still poorly understood.

As a consequence of stress, bifidobacteria are likely to activate synthesis of stress proteins, including molecular chaperones and proteases, to maintain viability that enables these bacteria to promote health on the host. Although the annotated genome sequence established on B. longum NCC2705 suggests a number of physiological traits, which may partially explain the successful adaptation and survival of bifidobacteria in gut [[], functional information about the proteins encoded by the genome is still required to achieve a comprehensive understanding of mechanisms involved in maintenance of viability. To date, knowledge regarding the individual stress proteins is exclusively based on transcriptional level experiments indicating the importance of chaperone proteins DnaK, GroEL, GroES and GrpE for stress tolerance of some bifidobacterial strains [[[]. Proteomic studies including modern mass spectrometric approaches such as multi-dimensional chromatography (MudPIT) coupled with tandem mass spectrometry (MS/MS) and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) have been used to identify most abundantly expressed proteins in Bifidobacterium infantis[[] during physiological growth and in generation of stress-related fingerprints of B. lactis[[0], respectively. Two-dimensional gel electrophoresis (2D-PAGE) coupled with protein identification using mass spectrometry and analytical software is a widely exploited proteomic approach that allows high resolution protein separation and accurate protein identification. Despite of drawbacks, such as limited throughput capability, requirements for large sample volumes, gel-to-gel variability, and the inability to measure low abundance proteins, 2D-PAGE is still the method of choice to investigate synthesis and modification of individual proteins under various environmental conditions. To date, studies involving 2D-PAGE analyses have not been reported for bifidobacteria.

In this study, proteomic tools including protein radiolabelling, 2D-PAGE and mass spectrometric analyses were applied to elucidate proteins with potential importance for stress tolerance of a B. longum 3A.

2 Materials and methods

2.1 Bacterial strains and media

The B. longum 3A (Bioferme Ltd., Kaarina, Finland) was propagated on MRS agar or in MRS broth (Difco Laboratories, Detroit, MI) supplemented with 0.005% cysteine (MRSc) to maintain low redox potential at 37°C in an anaerobic chamber (Concept Plus Anaerobe Work Station, Ruskinn Technology). To investigate de novo protein synthesis in B. longum the Methionine Assay Medium (MAM) purchased from Difco Laboratories (Detroit, MI) was modified to obtain a semi-defined medium (SDM) by the addition of 0.075 g l−1 valine, 0.035 g l−1 leucine, 0.032 g l−1 isoleucine, 0.1 g l−1 glutamine, 0.05 g l−1 glutamate, 0.043 g l−1 phenylalanine, 0.068 g l−1 asparagine, 0.047 g l−1 histidine, 0.06 g l−1 proline, 0.07 g l−1 lysine, 0.042 g l−1 alanine, 0.051 g l−1 threonine, 0.05 g l−1 serine, 0.05 g l−1 glycine, 1.5 g l−1 cysteine, 4 mg l−1 thiamine, panthothenate and riboflavin, 3 mg l−1 pyridoxine and nicotinic acid, 5 mg l−1d-biotin, B12, folic acid and myo-inositol and β-aminobentzoate, 0.5% fructose, 0.1% Tween 80 (vol/vol) and 1% MRS (vol/vol). SDM was sterilized using a 0.45 μm filter prior to inoculations. Growth of B. longum in SDM was measured with a U-2000 spectrophotometer (Hitachi) at 450 nm.

2.2 Preparation of protein extracts and 2D-PAGE

The growth experiments including stress treatments were carried out in an anaerobic chamber. Prior to inoculation, the media used for cultivations were incubated in anaerobic chamber for ?2 h. The heat-stress on B. longum cells was applied by culturing the cells in SDM to an OD450= 0.5, where aliquots of cells (1 ml) were incubated at 37°C or shifted to 47°C for 30 min. Cells were harvested by centrifugation, washed once with ice-cold H2O and homogenized with glass-beads (Ø 0.1 mm) 3 × 5 min at +4°C. Proteins were solubilized in 300 μl of isoelectric focusing (IEF) solution containing 8.7 M Urea, 3.5% CHAPS (3-[(3-cholamidopropyl)dimethylammonio-]-1-propanesulfonate)(w/v), 50 mM DTT, 0.5% Bio-Lyte 4/6 and Bio-Lyte 5/7 Ampholytes (Bio-Rad)(v/v) followed by centrifugation (13,200g, 15–30 min, RT) to remove the cell debris and glass-beads. Duplicate 11 cm ReadyStrip™ IPG (immobilized pH gradient)-strips for each sample (Bio-Rad) were rehydrated overnight with 125 μl of IEF solution including sample and IEF was carried out using a Protean IEF Cell according to instructions provided by the supplier (Bio-Rad). IEF was carried out at 20°C with an initial voltage of 250 V for 15 min followed by ramping to 8000 V until 35,000 V h was reached. Strips were then equilibrated sequentially in a buffer (Tris–HCl containing 6 M urea, 30% [vol/vol] glycerol, and 2% SDS) containing 1% DTT (w/v) or 2.5% iodoacetamide (w/v) for 10 min each and applied onto SDS – 8–16% gradient Criterion Precast gels (Bio-Rad) for electrophoresis in a Criterion Dodeca Cell (BioRad) at 200 V for ca. 1 h with running buffer containing 250 mM glycine, 25 mM Tris–HCl and 0.1% SDS (w/v). Following electrophoresis the gels were subjected to silver staining using the Silver Stain Plus kit (Bio-Rad) to visualize separated proteins. The 2D-gels were calibrated using 2D SDS–PAGE Standards (Bio-Rad).

2.3 Pulse-chase labelling of polypeptides

Several B. longum colonies grown on MRSc agar were inoculated in SDM (10 ml) and cultured under conditions indicated above. The overnight culture was diluted 100-fold in fresh SDM and allowed to grow to an OD450= 0.5 where 50–200 μl of cells were mixed with 30–60 μCi (1–2 × 107 Bq) of [35S]methionine and incubated at 37°C for 20 min. Protein labelling was completed by the addition of unlabelled methionine to a final concentration of 0.8 mg/l, and after 2 min chloramphenicol in a final concentration of 1 mg ml−1 was added to block the protein synthesis. Cells were harvested by centrifugation (13,200g, 3 min, +4°C), washed once with ice-cold H2O and subjected to liquid scintillation counting to calculate the incorporation efficiency of [35S]methionine. Briefly, cell pellets were suspended in 50 μl of H2O and then mixed with 3 ml of liquid scintillation cocktail (OptiPhase HiSafe 3, Wallac Oy) in a scintillation vial. Radioactivity in each sample was determined using a Wallac 1415 Liquid Scintillation counter according to the instructions provided (Wallac Oy). Incorporation efficiencies were calculated from two independent experiments using three parallel samples in each.

2.4 2D-PAGE of pulse-chase labelled proteins

Protein synthesis under stress conditions was studied using the pulse-chase labelling protocol described above. Protein pulse-chase labelling for control cells was conducted at 37°C for 20 or 30 min in a reaction containing 200 μl of B. longum cells (OD450= 0.5) mixed with 45 μCi of [35S]methionine. The heat-shock was applied by transferring 200 μl of cells to an eppendorf tube in a heat-block at 47°C inside the anaerobic chamber. The bile salts stress was applied by adding bile salts to a final concentration of 0.1%. For heat-shock experiments, the radiolabel was added at time point 0 min. In case of bile salts treatment, the protein radiolabelling was initiated 20 min after the addition of bile salts, since the addition of this compound simultaneously with the radiolabel was found to result in poor protein labelling efficiency. Following a stress treatment, proteins were chased with the unlabelled methionine and chloramphenicol as described above. The samples were prepared and subjected to 2D-PAGE as described above. After the second dimension, the gels were fixed in 10% acetic acid (v/v)–40% ethanol (v/v) for 10 min and equilibrated in 30% ethanol (v/v)–2% glycerol (v/v) and dried between two sheets of cellophane. The dried 2D-gels were exposed to Molecular Imager screens (Bio-Rad) for approx. 16 h followed by scanning with a GS-525 Molecular Imager System (Bio-Rad). Images representing proteomes of B. longum were analyzed using PDQuest image analysis software (PDQuest 6.2)(Bio-Rad) to quantify the difference in intensity of spots between gels after normalization the radioactivity of each spot with the total radioactivity of the gel. Relative induction factors for protein of interests were calculated from three independent experiments. Significant differences in protein expression levels were determined by Student's t-test with a set value of P≤ 0.05.

2.5 In-gel digestion of protein

2D-gels of proteins extracted from cells grown in SDM were silver stained using the Silver Stain Plus kit according to the instructions provided by Bio-Rad. Protein spots of interest were cut out of gels and in-gel digestion was performed using previously described methods [[1]. Proteins were reduced and alkylated with iodoacetamide before digestion with trypsin (Sequencing Grade Modified Trypsin, Promega) overnight at 37°C. The peptides were extracted once with 25 mM ammonium bicarbonate and twice with 5% formic acid and the extracts were pooled. Before MALDI-TOF/TOF mass spectrometric analysis, the peptide mixture was desalted using Millipore μ-C18 ZipTip™.

2.6 Mass spectrometry

Mass mapping of the peptides generated was performed with an Ultraflex™ MALDI-TOF/TOF mass spectrometer (Bruker-Daltonics, Bremen, Germany) equipped with a nitrogen laser in a positive ion reflector mode using α-cyano-4-hydroxycinnamic acid as the matrix. The MALDI-TOF spectra were externally calibrated with the standard peptide mixture from Bruker-Daltonics (Bremen, Germany). In case of a peptide fragmentation analysis a peptide from above mass mapping analysis was selected as a precursor ion and subjected for further MS/MS fragment analysis in the MALDI-TOF/TOF lift-mode. Database searches with the aid of the genome data base established on B. longum NCC2705 [[] were carried out by either Mascot peptide map fingerprint or Mascot MS/MS ions search (http://www.matrixscience.com/).

2.7 Immunoblotting analyses

HtrA specific antibodies utilized in this study were prepared as follows. The htrA coding region excluding the region encoding for putative signal peptide was amplified from Lactobacillus helveticus 53/7 [[2] using primers 5′-GGTGGATCCTCTTATTACGCAATGGACC-3′ and 5′-ATACGTCGACTAAATCGCGTGATATAGCTC-3′ digested with Bam HI and Sal I and cloned in the respective sites in pQE30 (Qiagen) in Escherichia coli M15[pREP4]. His6-HtrA was purified from E. coli carrying the pQE-6His-htrA by the standard procedure recommended by Qiagen. The purified His6-HtrA was used for custom antibody production in rabbits. For immunoblotting analyses B. longum cells were cultured to OD450= 0.5, where aliquots of cells (1 ml) were incubated at 37°C or exposed to a heat-shock at 47°C or bile salts (0.1%) for 30 min. Preparation of protein and 2D-gel electrophoresis with equal amount of protein (200 μg) extracted from the control and stressed cells were carried out as described above. Following 2D-PAGE proteins were transferred from the gel to a PVDF membrane (0.45 μm; Millipore) with the use of a Bio-Rad Trans-Blot SD according to the instructions provided (Bio-Rad). Membranes were treated as immunoblots with HtrA antibodies (1:5000) as the primary antibody and horse radish peroxidase-conjugated goat antibodies to rabbit immunoglobulin G (HRP; 1:50000) (Bio-Rad) as the secondary antibody. After the chemiluminescence detection the membranes were stained with Coomassie Blue to corroborate that the amount of protein separated in 2D-gels was comparatively same.

3 Results and discussion

3.1 Formulation of semi-defined medium for B. longum 3A

Investigation of synthesis rates of proteins is, if possible, accomplished by radiolabelling of proteins in a chemically defined medium (CDM) in order to obtain efficiently labelled protein sample. To date, a semi-defined medium (SDM) has been formulated for a B. infantis[[3]. However, in this medium amino acids are provided as casein hydrolyzate that results in medium rich with unlabelled methionine/cysteine, which would reduce the protein labelling efficiency with [35S]methionine/cysteine. To facilitate proteomic studies we exploited the commercially available MAM in order to obtain CDM for the strain under study. Since MAM (composed of 42 constituents) supplemented with methionine (200 μg ml−1) proved to be a poor medium for B. longum, this medium was optimized by increasing the concentration of certain constituents present in MAM and supplementing MAM with some new factors. For each growth experiment B. longum was first cultured overnight in MRSc broth and then diluted 1:100-fold into the MAM-derived CDM. Each supplement was tested by adding the desired constituent one at a time and measuring the optical density at 450 nm. This resulted in a medium in which the concentration of 13 amino acids (valine, leucine, isoleucine, phenylalanine, asparagine, histidine, proline, lysine, alanine, threonine, serine, glutamate, glycine) and 8 vitamins (thiamine, pyridoxine, panthothenic acid, riboflavin, nicotinic acid, β-aminobentzoic acid, d-biotin, folic acid) was increased. While this composition allowed moderate growth of B. longum 3A, the addition of certain new amino acids (glutamine, cysteine) and vitamins (myo-inositol and B12) into the medium improved the growth of this strain (data not shown). The ability of B. longum 3A to grow in the MAM-derived CDM without methionine was also tested. Since the omission of methionine was found not to affect the growth rate of the strain (data not shown), this amino acid was excluded in the following growth experiments. In addition to glucose (2.5%), which is the carbon source supplied by MAM, we found that the addition of 0.5% fructose stimulated growth of B. longum 3A resulting in increased growth rate and higher cell density (data not shown). There was also a noticeable increase in the growth rate when Tween 80 (0.5%), a detergent that is known to function as a putative lipid source when culturing certain Lactobacillus species [[4,[5] was included in the growth medium (data not shown). However, this medium did not support growth for several growth cycles, since B. longum grown overnight in CDM was not able to initiate growth in fresh CDM unless 1% of MRS broth was included in the medium (data not shown). The resulting semi-defined medium (SDM) was found to support rapid growth of B. longum 3A with a specific growth rate of 67 min (SD 2.7). The other growth parameters such as the final cell density and pH reached by the overnight cultures were OD450= 2.58 ± 0.1 and pH at 4.95 ± 0.04, calculated from two independent experiments.

3.2 Heat-shock proteome of B. longum assessed by 2D-PAGE and MALDI-TOF

The standard heat-shock condition for B. longum was obtained as follows. A series of growth experiments were carried out by shifting cells grown in SDM at 37°C (OD450= 0.5) to 45, 47 and 49°C for 30, 60 and 120 min. Of the heat-shock conditions tested, shifting of cells from 37 to 47°C was found to result in growth rate corresponding to ?20–50% of the growth rate obtained at 37°C (data not shown). Cell viability was further examined by plating appropriate dilutions of cultures (OD450= 0.5) grown in SDM at 37°C and treated for 60 min at 47°C to MRSc plates, which revealed that the heat-shock treatment decreased the number of colony forming units less than 5% (data not shown). Since the temperature shift from 37 to 47°C caused stress on B. longum cells, as indicated by the reduced growth rate, and, on the other hand, it did not have a major effect on cell viability, this temperature was used to induce heat-stress response on B. longum. Experimental conditions for 2D-PAGE using 11 cm ReadyStrip™ IPG strips (pH 4–7) for IEF and SDS 8–16% gradient gels were developed to investigate the proteome of B. longum with the aim to identify proteins that are expressed in response to heat-stress. To this end, B. longum cells were cultured in SDM to appropriate cell density where heat-stress was applied and aliquots of cells were withdrawn for 2D-PAGE. Several protein spots cut out from the silver-stained 2D-gels were subjected to MALDI-TOF mass spectrometric analyses for protein identification. Additional MS/MS fragment analyses of selected peptides derived from mass mapping were conducted to confirm the final identification. Since the genome sequence for the strain under study was not available the protein identification was obtained with aid of the recently released sequence database established for B. longum NCC2705 [[]. Fig. 1 represents B. longum proteomes in the pH range between 4 and 7 during growth at 37°C and 30 min after shifting the cells (OD450= 0.5) to 47°C. Three proteins exhibiting constant expression levels before and after applying of heat-stress were identified as 50S ribosomal protein (L25), elongation factor TS (EfTS) and transketolase (Tkt) (Fig. 1(a) and Table 1), whereas 11 proteins involved in stress response, translation, transport and binding and/or pyrimidine biosynthesis were found to be upregulated in response to heat-stress (Fig. 1(b) and Table 1). In case of GrpE, the mass mapping analyses revealed only three peptides matching with the putative GrpE protein from B. longum NCC2705 and further attempts to obtain fragmentation data of these peptides failed. Therefore, proteins like the putative GrpE and the other small molecular weight proteins presented in Table 2 (protein spots 15–19 and 21), with insufficient trypsin cleavage sites, require other approaches for identification. One of the identified proteins, the putative HtrA was found to be synthesized as a ?67 kDa protein (Fig. 1(b)). Most of the HtrA/DegP proteins in the SWISSPROT protein data base have molecular weights ranging between 42–51 kDa (data not shown), while e.g. Staphylococcus aureus has two genes encoding cell-wall associated HtrA proteins of 42 and 90 kDa [[6]. The molecular weight of the B. longum 3A HtrA is in good agreement with the predicted HtrA (67 kDa) of B. longum NCC2705 [[]. The typical bacterial HtrA contains a putative membrane anchor, one or two PDZ domains and an amino-terminal sequence encoding for a leader peptide, which undergoes post-translational modification by cleavage to release a mature cell-wall protein [[7]. Based on the genome sequence, B. longum carries a single htrA gene coding for a protein containing a transmembrane segment, a trypsin-like serine protease domain and a PDZ domain (data not shown). However, no apparent features indicating the presence of signal peptide or signal peptide cleavage site in the N-terminal region of HtrA could be identified using the sequence analysis tools provided at the Expasy Molecular Biology Server (http://au.expasy.org/). Interestingly, another protein with an apparent molecular weight of 40 kDa and pI value of 4.3 was found to appear in 2D-gel representing the B. longum proteome during heat-stress (Fig. 1(b)). This protein was identified as a putative degradation product of the cell-wall associated HtrA (Table 1). It has recently been demonstrated that Bacillus subtilis exhibits two HtrA proteins of 62 and 45.7 kDa, the latter possibly representing the cleaved HtrA before its secretion into the medium [[8]. The exact N-terminal amino acid sequence of the 40 kDa HtrA protein of B. longum 3A and the putative cleavage site in the 62 kDa HtrA remain to be studied, as well as the cellular location of the different forms of HtrA.


Silver-stained 2D-gels of cellular proteins extracted from non-stressed (a) and heat-shock treated (b) B. longum 3A cells. The numbered spots represent proteins, which were identified by MALDI-TOF/(TOF) analyses (Table 1). The horizontal and vertical axes represent pH (pI) range and molecular weight (kDa), respectively. 2D-gels were calibrated with the Bio-Rad 2D SDS–PAGE standards.

View this table:

Identity of B. longum 3A proteins separated by 2D-PAGE

Spot no.aPutative function (Acc. no.)bpI/Mw (kDa)MALDI-TOF/(TOF)e
Mass mappingMS/MS fragment analysis
ObservedcTheoreticaldNo. peptides matchedfSequence coverage (%)f
150S ribosomal protein L25 (AAN24666)4.7/224.7/21.8n.i.gn.i.SEFGKGVAR, ATTIKLEGEAR
2Elongation factor TS (AAN25299)5.1/285.1/29.91643.5n.i.
3Transketolase Tkt (AAN24535)5.2/755.0/75.91012.8AGELPEGFDK
4Serine protease HtrA (AAN24379)4.2/674.3/67.557.1ADEFNPQGVDQTPR
5Serine protease HtrA (AAN24379)4.3/40h710.8SDTVEADBVTR
6Chaperone protein DnaK (AAN24348)4.5/664.7/66.92242.3n.i.
7Chaperone protein GroEL (AAN23869)4.6/574.7/56.81333.6n.i.
8Lactaldehyde reductase FucO (AAN25460)4.7/404.7/40.6616.2GAIKEIPAVAK, ATEEDILAIYK
9Transaldolase Tal (AAN24534)4.8/404.8/39.71539.8n.i.
10Argininosuccinate synthase ArgG (AAN24866)5.0/445.0/45.8618.7RSDSSLYDYK, PYSIDQNVWGR, LATYDSGDTFDQK
11Chaperone protein GrpE (AAN24347)i4.4/204.5/23.6311.4n.i.
12Orotate phosphoribosyltransferase PyrE (AAN24603)6.0/225.7/24.9627.3NIDTVFGPAYK, VLLVDDVMTAGTAVR
13ATP binding protein of ABC sugar transporter MsiK (AAN24180)5.8/455.8/40.8514.1TQIAALQR, IAGTPKDEIR
14Clp ATPase ClpA/ClpB (AAN25051)4.8/974.9/96.3710.6NALVALPSASGSSTSQPQASR
  • aSpot numbering refers to proteins indicated in Fig. 1.

  • bPutative functions and accession numbers were obtained from the GenBank database established for B.longum NCC2705 [[].

  • cpI/Mw values were determined on the basis of the data presented in Fig. 1.

  • dDetermined with Compute pI/Mw tool at http://us.expasy.org/tools/pi_tool.html.

  • eMass mapping of peptides by MALDI-TOF coupled with an additional MS/MS fragment analysis in the MALDI-TOF/TOF lift- mode. Final identification was obtained with the aid of the database searches using Mascot peptide map fingerprint or.

  • fSequence coverage (%) refers to the total protein sequence represented by the peptide fragments matched in the search result.

  • gn.i., no identification.

  • hThe exact proteolytic cleavage site in HtrA not identified.

  • iIdentification of the protein is suggestive due to a low number of matched peptides obtained by mass mapping and the lack of identification in the subsequent MS/MS fragment analysis.

View this table:

Stress-induced proteins of B. longum

Protein spotaIdentified proteinbEstimatedInductionc
pIMw (kDa)HeatBile
  • aSpot numbering refers to proteins indicated in Fig. 1.

  • bProtein identified by immunoblotting and MALDI-TOF/(TOF) analyses.

  • cInduction fold: -, not induced; +, below 2-fold; ++, between 2- and 5-fold, +++, between 5- and 10-fold; ++++, between 10- and 15-fold; +++++, above 70-fold.

  • dProteins were not detected in silver-stained 2D-gels.

  • en.d., Identity not known.

  • fIdentification of the protein is suggestive due to a low number of matched peptides obtained by mass mapping and the lack of identification in the subsequent MS/MS fragment analysis.

3.3 Proteins in B. longum are efficiently radiolabelled during growth in SDM

To investigate synthesis rates of proteins, experimental conditions for protein radiolabelling using [35S]methionine in SDM were developed. The labelling efficiency of proteins in SDM was evaluated by the use of liquid scintillation counting. Optimization of the reaction conditions revealed the highest incorporation efficiency of 4.33 × 10−3μmol min−1 (SD 0.5) with 45 μCi of [35S]methionine (1.5 × 107 Bq) requiring 200 μl of cells and incubation of 20 min at 37°C. The labelling efficiency was also assessed by analyzing the protein extracts using the classical SDS–PAGE that permitted clear detection of proteins in less than 4 h of exposure to the imaging screen. These results indicated the applicability of the developed approaches to investigate de novo protein synthesis in the B. longum 3A.

3.4 Stress response proteomics using protein radiolabelling and 2D-PAGE

Pulse-chase labelling coupled with 2D-PAGE was exploited to investigate global changes in protein synthesis after applying of heat and/or bile salts stress on B. longum. Cells subjected to pulse-chase labelling during physiological growth (at 37°C) resulted in detection of approx. of 400 proteins following 2D-PAGE (Fig. 2(a)). Cells exposed to a heat-stress at 47°C indicated that the synthesis of 13 polypeptides was increased (Fig. 2(b) and Table 2). Four proteins showing increased synthesis rates during heat-shock (Fig. 2(b), protein spots 15, 17, 18 and 20) were found to be below detection limit in silver-stained gels (Fig. 1(b)) and could not be identified. Six heat-induced proteins were identified by MALDI-TOF/(TOF) analyses (Table 2). Of these heat-shock proteins, GroEL, ClpA/B and the putative GrpE were found to be upregulated to a 2–14-fold higher level, whereas HtrA and DnaK exhibited over 70-fold higher expression levels (Fig. 2(b) and Table 2), suggesting the importance of these stress proteins under heat-shock conditions. One of the heat-induced proteins was identified as an argininosuccinate synthetase (ArgG) that was found to be increased 8.5-fold upon heat-stress (Fig. 2(b), spot 10). A similar finding has recently been obtained by the transcriptome analysis on B. subtilis, which has revealed that the expression of the operon genes coding for proteins involved in arginine metabolism, including also that of argG, are strongly induced upon heat-stress [[9]. The heat-induction of the arginine metabolism genes was suggested to have resulted from the inactivation of the repressor protein involved in regulation of this operon [[9].


Assessment of protein synthesis of B.longum 3A by [35S]methionine labelling in SDM followed by 2D-PAGE. (a) 2D-image of proteins extracted from cells pulse-chase labelled (a) at 37°C and (b) 47°C, or (c) from cells pulse-chase labelled at 37°C in the presence of 0.1% bile salt. The spot numbers 1–13 (a) represent proteins exhibiting constant expression level (Table 2). The spot numbers 4–21 (b) and 4–7 (c) represent proteins displaying increased synthesis (Table 2). Proteins marked with arrows were identified by MALDI-TOF/(TOF) analyses. Triangles refer to upregulated proteins that were not identified by MALDI-TOF(TOF). Horizontal and vertical axes represent pH (pI) range and molecular weight (kDa), respectively. 2D-gels were calibrated with the Bio-Rad 2D SDS–PAGE standards.

First experiments to investigate the effect of bile salts on protein synthesis in B. longum revealed that the addition of this compound simultaneously with the radiolabel resulted in low protein labeling efficiency (data not shown). Therefore, the cells were first allowed to adapt to 0.1% bile for 20 min (37°C) prior to initiation of the pulse-chase labelling with [35S]methionine (20 min at 37°C). Evaluation of 2D-gels of the labelled proteins revealed that HtrA, DnaK and GroEL were synthesised approx. 1.5–2-times (P≤ 0.05) more during bile salts stress compared to that in the control cells (Fig. 2(a)–(c) and Table 2). Although a number of studies have demonstrated that HtrA exhibits both the chaperone and protease activity [[7], and that it plays an important role under various stress conditions and virulence in certain bacteria [[0[5], no studies suggesting the importance of this protein for bile salts tolerance has yet been reported. In GIT, bifidobacteria must defend themselves against the action of bile salts in order to retain viability and to promote health. Bile salts are detergent-like compounds made by the liver and secreted into the bile in the small intestine, where they act on bacterial membranes by disrupting their lipid bilayer structure. Bile salts can also lead to protein denaturation that may account for the increased synthesis observed for HtrA during bile salts stress. In case of DnaK, studies of another bifidobacterial strain, B. adolescentis, have demonstrated that the expression of the dnaK gene can be induced by both the bile salts and heat-shock treatments [[]. In bacteria such as Enterococcus faecalis and Propionibacterium freudenreichii, the proteomic studies suggest that the two stress response pathways leading to heat and bile salt tolerance are closely related [[6[8].

Since mass spectrometric identification of protein spots indicated that HtrA is processed during heat-stress, we wished to examine the HtrA expression in more detail. First, B. longum cells (200 μl) were subjected to a prolonged pulse-labelling at elevated temperature (30 min at 47°C) prior to 2D-gel separation of proteins and autoradiography. Prolonged labelling under heat-stress conditions resulted in appearance of multiple heat induced protein spots migrating with varying molecular weights (45–67 kDa) and pI values (4.0–4.3) (Fig. 3(a)), possibly representing modified HtrA. In addition, immunoblotting of 2D-gels loaded with protein samples isolated from cells (1 ml) grown at 37°C and cells incubated at 47°C for 30 min was used to examine the HtrA expression. Only a single protein spot with pI and molecular weight in agreement with that calculated for HtrA ofB. longum NCC2705 was detected in B. longum 3A cells under non-stressed conditions (Fig. 3(b)), indicating specific cross-reaction of the antibodies with B. longum HtrA. As expected, the strongest signal with HtrA antibody was obtained from cells exposed to heat-shock (Fig. 3(b)). Several proteins (40–66 kDa) cross-reacting with HtrA antibodies were detected in sample from heat stressed cells (Fig. 3(b)). Thus, the pulse-labelling and the immunoblotting analyses both suggest that HtrA is expressed in multiple protein spots and that HtrA undergoes proteolytic processing during heat-shock. Since the exact cellular location of HtrA is presently unknown, we cannot exclude the possibility that the expression pattern observed reflects the maturation process of HtrA during translocation to the cell-wall of B. longum cells. Interestingly, a recent study of Gram-positive and Gram-negative bacteria has shown that a number of proteins, including DnaK, GroEL, and certain biosynthetic and metabolic enzymes, undergo specific tagging via phosphorylation resulting in highly acidic proteins under stress conditions, or conditions which overload the proteolytic system [[9]. This protein modification was suggested to be the signal that directs these proteins to the general degradation pathway in bacteria [[9]. Our results suggest that during heat-shock the B. longum HtrA is post-translationally modified both by a tagging process and proteolysis. However, further studies are needed to elucidate whether the putative isoforms of B. longum HtrA represent phosphorylated proteins tagged for degradation. Nevertheless, the strongly induced synthesis rate of HtrA during heat-shock in B. longum suggests importance of this protein under heat-stress conditions, possibly by assisting folding or degradation of heat-denatured proteins.


Expression of HtrA in response to heat and bile salts stress. (a) Portions of 2D-gels representing proteins of B. longum 3A in the pH range of 4.0–4.8. Cells were subjected to pulse-chase labelling for 30 min at 37 or 47°C. (b) Immunoblotting of 2D-gels with HtrA specific antibodies. Protein samples (200 μg) used for immunoblots were extracted from B. longum before and 30 min after exposing the cells to heat- or bile salt stress. Detection system based on chemiluminescence was used to visualize the cross-reacting proteins. DnaK, GroEL and HtrA are marked with arrows. Circled proteins refer to probable isoforms of HtrA. Horizontal and vertical axes represent pH (pI) range and the molecular weight (kDa), respectively. 2D-gels were calibrated with the Bio-Rad 2D SDS–PAGE standard.


This work was supported by TEKES (40079/01), BioFerme Ltd, Oy Fazer Ab, Oy Sinebrychoff Ab, Academy of Finland (78646) and ABS graduate school funding to A. Suokko.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
View Abstract