Genetic and Transcriptional Organization of the clpC Locus in Bifidobacterium breve UCC 2003

Abstract

A homolog of the clpC ATPase gene was identified in the genome of Bifidobacterium breve UCC 2003. Since this gene is very well conserved among eubacteria, we employed a PCR-based approach using primers based on highly conserved regions of ClpC proteins in order to identify homologous genes in other bifidobacterial species. Analysis by slot blot, Northern blot, and primer extension experiments showed that transcription of clpC is induced in response to moderate heat shock regimes. Moreover, we identified in the genome sequence of B. breve UCC 2003 a gene, designated clgR, which is predicted to encode a transcriptional regulator involved in regulation of the bifidobacterial clpC gene. The role of this protein in the regulation of B. breve UCC 2003 clpC gene expression was investigated by performing gel retardation experiments. We show that a biologically active ClgR molecule requires one or more proteinaceous coactivators to assist in the specific binding of ClgR to the clpC promoter region.

Bifidobacteria are gram-positive anaerobic bacteria which are found in the gastrointestinal tracts of humans and animals. They are increasingly used as probiotic components in a wide range of functional foods (24). Successful application of bifidobacteria in this role requires that they are able to withstand substantial environmental insults, including pH fluctuations and heat and osmotic stresses. However, relatively little is known about the genetic basis of the stress response in bifidobacteria (18, 25, 26, 27). In order to survive adverse environmental conditions, bacteria undergo a complex program of differential gene expression that leads to transient induction of a subset of proteins that prevent the accumulation of unfolded and misfolded proteins (32). Most of these proteins are molecular chaperones or proteases like the HSP100/Clp ATPases (ClpA, ClpB, ClpC, and ClpX). The latter proteins constitute a closely related family that perform crucial housekeeping functions, including protein reactivation, remodeling activities, and specific targeting of proteins for degradation by the ClpP protease (2, 20, 31). Members of the Clp ATPase family of proteins have a modular structure, and their classification is based on the presence of either one or two ATP-binding domains.

ClpC belongs to the stress response-related Clp ATPase family, whose members act as chaperones and regulators of proteolysis (15, 17, 19). The cellular concentration of ClpC increases in response to environmental conditions that lead to the accumulation of nonfunctional, misfolded, or aggregated proteins. The underlying regulatory mechanism of ClpC induction has been extensively investigated in low-G+C-content gram-positive bacteria, in which the transcriptional repressor CtsR controls the expression of the clpC, clpE, and clpP genes (4, 8). In Bacillus subtilis transcription of the clpC gene is under the control of a vegetative σ^A promoter and the CtsR repressor, which binds a heptanucleotide repeat that overlaps the −10 or −35 box (4). In Streptococcus salivarius, clp genes are negatively controlled by both CtsR and the HrcA repressor (3). In this bacterium the stability of CtsR and consequently its repressive action appear in turn to be modulated via heat-triggered posttranslation modification by a putative protein kinase, an event that stimulates ClpP proteolytic action (8). In high-G+C-content gram-positive bacteria like Corynebacterium glutamicum and Streptomyces lividans the expression of clpC is under dual control by a transcriptional activator designated ClgR and sigma factor σ^H (1, 5). In the latter bacteria, it was shown that under stressful conditions ClgR binds to a palindromic sequence motif (CGC-N₅-GCG) upstream of both the clpC and clpP1P2 genes, which activates their transcription (5).

Since the ClpC chaperone has been considered a cornerstone in the biology of stress responses in many members of the Actinobacteridae group (5), we decided to characterize the clpC gene in bifidobacteria, whose genetic organization and possible stress-induced expression have not been explored to date. Moreover, we were specifically interested in the role of the ClpC chaperone in heat and osmotic stress responses. Identification of the genetic basis of heat and osmotic resistance for industrially applicable bifidobacteria is highly desirable for selecting probiotic strains that resist stressful conditions encountered during food manufacture (e.g., freeze-drying, freezing, or spray-drying) and in their natural environments, where the ability to respond rapidly to stress is essential for their survival (e.g., in the digestive tract).

In this report, identification and characterization of the clpC locus of B. breve UCC 2003 are described. Transcriptional induction of the B. breve clpC gene upon exposure to stressful conditions was investigated by slot blot, Northern blot hybridization, and primer extension analyses, while the first insight into regulation of the clpC gene in B. breve UCC 2003 was obtained by gel retardation assays.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

All Bifidobacterium strains were grown anaerobically in MRS (Difco, Detroit, MI) supplemented with 0.05% (wt/vol) l-cysteine HCl and were incubated at 37°C for 16 h. Escherichia coli was grown aerobically on a rotary shaker (150 rpm) at 37°C in LB medium (Difco) or was plated onto LB agar (Difco) plates when appropriate. Antibiotics were used at the following concentrations: ampicillin, 100 μg/ml; and kanamycin, 25 μg/ml. Isopropyl-β-d-1-thiogalactopyranoside (IPTG) (Fluka, Germany) was added to a final concentration of 1 mM when E. coli carrying pQE-ClgR was cultivated.

B. breve UCC 2003 is a human commensal isolated from an infant nursling stool. The optimal growth conditions for this strain were determined previously (van Sinderen, unpublished results). Briefly, the optimal growth temperature is 37°C, the minimal temperature is 30°C, and the maximal growth temperature is 43°C.

DNA isolation.

Genomic DNA was extracted by using the protocol described in a previous study (27).

Plasmids and plasmid construction.

The E. coli pQE-30 vector (QIAGEN, West Sussex, United Kingdom) was used for overproduction and purification of the B. breve UCC 2003 ClgR protein. The clgR gene from B. breve UCC 2003 was amplified using primers 903-uni (5′-CGCGGATCCGGTCATCGTGTTGTACTAAG-3′) and 903-rev (5′-CCCAAGCTTACGTTCGACGGACTCGAG-3′) containing BamHI and HindIII restriction sites, respectively. The PCR conditions were as follows. Each PCR mixture (50 μl) contained 20 mM Tris-HCl, 50 mM KCl, each deoxynucleoside triphosphate at a concentration of 200 μM, 50 pmol of each primer, 1.5 mM MgCl₂, and 1 U of Taq DNA polymerase (Gibco BRL, Paisley, United Kingdom). Each PCR cycling profile consisted of an initial denaturation step of 5 min at 95°C, followed by 35 cycles of amplification as follows: denaturation for 30 s at 95°C, annealing for 30 s at 55°C, and extension for 1 min at 72°C. The PCR was completed by elongation for 10 min at 72°C. The resultant 566-bp PCR fragment was digested with BamHI (Roche, Sussex, United Kingdom) and HindIII (Roche, Sussex, United Kingdom) and ligated into similarly restricted pQE30 using the T4 DNA ligase enzyme (Roche) to generate plasmid pQE-ClgR, which was introduced by electrotransformation into E. coli M15 (QIAGEN, United Kingdom) as described by Sambrook and Russell (14).

DNA amplification of the region flanking the clpC gene in Bifidobacterium animalis subsp. animalis.

Regions surrounding the clpC homologue in B. animalis subsp. animalis ATCC 25527 were determined by inverse PCR (9). One microgram of chromosomal DNA was digested with restriction endonuclease SspI (Roche) or DraI (Roche); the restriction fragments were self-ligated and amplified using primers CLP-A-inv (5′-CTCAATCGTCTCGGCAATCGAC-3′) and CLP-B-inv (5′-GTCTCTGCGCTGTCGCGCTC-3′) as described by Sambrook and Russell (14). The inverse PCR products obtained were used for direct sequencing using synthetic oligonucleotides.

DNA amplification and cloning of the uspA-clpC and clpC-codA gene constellations.

A DNA fragment corresponding to the uspA-clpC gene inventory was amplified from different bifidobacterial species using oligonucleotides clp-1 (5′-GTCGGTAAACCGTTCGAAC-3′) and clp-2 (5′-CAACTGCCAGCAGAACATC-3′), while the clpC-codA gene constellation was verified using oligonucleotides clp-3 (5′-CAACGATCGTCTGTTCGA-3′) and clp-4 (5′-GCAGGAAGGCATGTACTC-3′). Each PCR mixture (50 μl) contained 20 mM Tris-HCl, 50 mM KCl, each deoxynucleoside triphosphate at a concentration of 200 μM, 50 pmol of each primer, 1.5 mM MgCl₂ and 1 U of Taq DNA polymerase (Gibco BRL). Each PCR cycling profile consisted of an initial denaturation step of 5 min at 95°C, followed by 35 cycles of amplification as follows: denaturation for 30 s at 95°C, annealing for 30 s at 51°C, and extension for 1 min at 72°C. The PCR was completed by elongation for 10 min at 72°C.

The resulting amplicons were separated on a 1.5% agarose gel, which was followed by ethidium bromide staining. PCR fragments were purified using a PCR purification spin kit (QIAGEN, United Kingdom) and were subsequently sequenced.

Reference sequences.

The clpC-encoded protein sequences from the following bacteria (GenBank accession numbers are indicated in parentheses) were used for clpC-encoded protein comparison: Bifidobacterium longum NCC 2705 (NC_004307M), Mycobacterium bovis AF 2122/97 (NV002945), Corynebacterium efficiens YS314 (NC004369), Streptomyces coelicolor (NC003903), Thermobifida fusca YX (NZ_AAAQ00000000), and B. subtilis (D26185).

Homologous ClgR protein sequences from the following bacteria (GenBank accession numbers are indicated in parentheses) were used for ClgR comparative analysis: B. longum NCC 2705 (AE014771-9), Mycobacterium tuberculosis CDC1551 (CAA15541), C. glutamicum ATCC 13005 (BAB99355), S. coelicolor A3 (CAB42938), and T. fusca YX (ZP_00058963).

DNA sequencing.

Nucleotide sequencing of both strands of PCR amplicons was performed by MWG-Biotech AG (Ebersberg, Germany).

Southern hybridization.

Ten micrograms of bacterial DNA was digested to completion with restriction endonuclease EcoRI or EcoRV used as recommended by the supplier (Roche, United Kingdom). These restriction enzymes were chosen because no restriction sites were observed within the amplified clpC gene fragment. Southern blotting of agarose gels was performed with Hybond N+ membranes (Amersham, Little Chalfont, United Kingdom) by following protocols outlined by Sambrook and Russell (14). Filters were hybridized with a B. breve-derived clpC probe which was labeled with α-³²P using the Random Primed DNA labeling system (Roche, United Kingdom). Subsequent prehybridization, hybridization, and autoradiography were carried out as described by Sambrook and Russell (14).

Overproduction of ClgR in E. coli.

A 300-ml culture of an E. coli M15 strain containing the pQE-ClgR plasmid was grown to an optical density at 600 nm of 0.6 prior to induction by addition of 1 mM IPTG. Three hours after induction, cells were harvested by centrifugation at 10,000 rpm for 10 min. Cell pellets were resuspended in lysis buffer (100 mM NaH₂PO₄, 10 mM Tris-HCl, 6 M guanidine hydrochloride; pH 8.0) as recommended by the supplier (QIAGEN, United Kingdom) and were lysed with gentle shaking at 27°C for 2 h. The extract was clarified by centrifugation at 13,000 rpm for 10 min, and the supernatant was used directly as a crude extract in gel mobility shift DNA binding assays.

Protein concentrations were determined using the Bio-Rad protein assay in conjunction with a bovine serum albumin standard curve.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed as described by Laemmli (10) using a 4% stacking gel and a 12% separating gel. Protein sizes were compared to a prestained protein marker (New England Biolabs, Wilbury Way Hitchin, Herts, United Kingdom).

RNA isolation and Northern blot analysis.

B. breve UCC 2003 cells were grown at 37°C in a modified MRS medium with no added NaCl to an optical density at 600 nm of 0.6, at which point the culture was held at 37°C (control culture for osmotic and temperature shock experiments) or shifted to 20°C, 43°C, 47°C, or 50°C, while osmotic stress was applied by addition of 5 M NaCl-containing prewarmed medium to obtain a final NaCl concentration of either 0.5 M or 0.7 M. At various times (15, 50, and 150 min with a temperature shift or 15 and 50 min with an osmotic shift), 30-ml aliquots of cultures were collected and briefly centrifuged to harvest cells. The temperature conditions and the sampling time were chosen based on the optimal growth conditions for B. breve UCC 2003 and on previously reports concerning stress responses in bifidobacteria (18, 25, 26). Total RNA was isolated by using macaloid acid (25) and was treated with DNase (Roche, United Kingdom). The quantity of RNA was evaluated as described by Sambrook and Russell (14). In slot blot experiments, 25 μg RNA was spotted onto a Zeta-Probe blotting membrane (Bio-Rad, United Kingdom) using a Bio-Dot SF microfiltration apparatus (Bio-Rad, United Kingdom) as specified by the manufacturer and was treated with one UV auto-cross-linking cycle by using a UV Stratalinker 1800 (Stratagene, La Jolla, CA). In Northern blot experiments, 15 μg RNA was subjected to electrophoresis on a 1.5% agarose-formaldehyde denaturing gel, transferred to a Zeta-Probe blotting membrane as described by Sambrook and Russell (14), and fixed by UV cross-linking using a Stratalinker 1800. PCR amplicons obtained by using primer combinations that were designed to target the B. breve UCC 2003 clpC, uspA, and codA genes were radiolabeled. Moreover, the probe used in the control slot blot hybridization experiment was obtained by PCR using primers 16SRT2 and 16SRT3 and PCR conditions that have been described previously (25). Prehybridization and hybridization were carried out at 65°C in 0.5 M NaHPO₄ (pH 7.2), 1.0 mM EDTA, and 7.0% (wt/vol) SDS. Following 18 h of hybridization, the membrane was rinsed twice for 30 min at 65°C in 0.1 M NaHPO₄ (pH 7.2), 1.0 mM EDTA, 1% (wt/vol) SDS and twice for 30 min at 65°C in 0.1 mM NaHPO₄ (pH 7.2), 1.0 mM EDTA, 0.1% (wt/vol) SDS and exposed to X-OMAT autoradiography film (Eastman Kodak, United States). The slot blot X-ray autoradiograms were quantitated by densitometric scanning. The hybridizing signals corresponding to clpC mRNA synthesized at 20°C, 43°C, 47°C, and 50°C and with 0.5 M NaCl and 0.7 M NaCl were normalized to the signal obtained at 37°C. All slot blot and Northern hybridization experiments were performed at least twice.

Primer extension analysis.

The 5′ end of the clpC RNA transcript was determined as described in a previous study (21). The following synthetic oligonucleotide was used: clpC-prom (5′-CATCTCTTCGACCTGCTTG-3′). The primer extension experiment was also performed using the following oligonucleotide: clpC-prom2 (5′-GAGCAGCAGGTGTTCGGTG-3′).

Gel mobility shift DNA binding assays.

A 200-bp DNA fragment corresponding to the clpC promoter region (from position 77 to position −123 with respect to the putative start codon) was amplified by PCR with primers c-uni (5′-GTTGTGCTGGAGAGTCCTC-3′) and c-rev (5′-CTTCCATCGATCACAGGATG-3′). The resultant amplicon was purified using a G50 Spincolumn (Amersham) and then labeled using [γ-³²P]dATP and T4 polynucleotide kinase (New England Biolabs). The level of radioactive labeling was measured using a Beckman LS multipurpose scintillation counter. Crude extract from E. coli overproducing ClgR protein was employed in gel mobility shift DNA binding assays. Furthermore, an extract derived from cells containing the empty vector without the clgR gene was used as a control to rule out the possibility that other proteins present in the protein fraction might bind to the labeled clpC promoter fragment.

Binding reactions were performed in 20-μl (final volume) mixtures containing the labeled clpC promoter fragment (3,000 cpm) and various amounts of ClgR-containing crude extract in the presence of 1 μg of calf thymus DNA and binding buffer (50 mM Tris-HCl, pH 7.5; 50 mM NaCl; 10 mM MgCl₂). Following incubation at 37°C for 30 min, samples were then loaded on a 4% polyacrylamide gel, and electrophoresis was performed at 28 V · cm⁻¹ for 1 h. Bands were visualized by autoradiography at −70°C using Kodak Biomax MR film.

All gel retardation assays were performed at least twice.

Protease treatment of crude cell extract from B. breve UCC 2003.

Twelve micrograms of crude cell extract from B. breve UCC 2003 was incubated with 30 U of pronase (Roche) for 4 h at 37°C as recommended by the supplier. The pronase was subsequently chemically inactivated by incubating the mixture with 4 μl protease inhibitor cocktail (1× solution; Roche) as recommended by the supplier. The control sample was treated in the same way except that no pronase was added.

Nucleotide sequence accession numbers.

The nucleotide sequence data for the clpC locus of B. breve UCC 2003 and B. animalis subsp. animalis ATCC 25527 have been deposited in the GenBank database under accession numbers AY722390 and AY837844, respectively. The nucleotide sequence data for the clgR gene of B. breve UCC 2003 have been deposited in the GenBank database under accession number AY837843.

RESULTS AND DISCUSSION

Identification of clpC genes in B. breve.

The highly conserved amino acid sequences of assumed Clp proteins in high-G+C-content gram-positive bacteria (1, 5) were used to identify the putative clpC gene in the sequenced B. breve UCC 2003 genome (S. Leahy, J. A. Moreno Muñoz, M. O'Connell-Motherway, G. F. Fitzgerald, D. G. Higgins, and D. van Sinderen, unpublished data). The deduced amino acid sequence for a single open reading frame (ORF) from the UCC 2003 strain, designated clpC, displayed highly significant similarity over its entire length to ClpC subunits defined for B. longum (89% identity and 90% similarity), S. coelicolor (64% identity and 76% similarity), and M. tuberculosis (63% identity and 75% similarity) (Fig. 1a). The amino acid alignment with other prokaryotic ClpC ATPases showed that the presumptive ClpC of B. breve UCC 2003 possesses all recognized structural motifs which are typical for this protein family (18) (Fig. 1b), including two nucleotide-binding regions, ATP-binding region 1 (amino acids 190 to 417) and ATP-binding region 2 (amino acids 483 to 673), harboring the characteristic ATPase A and B boxes. The latter two domains are separated by a 66-amino-acid spacer region (amino acids 484 to 672) and are between a 189-amino-acid leader sequence at the N terminus and a 196-amino-acid trailer sequence at the C terminus.

FIG. 1.

(a) Comparison of the clpC locus in B. breve UCC 2003 with the corresponding loci in various bacteria. Each arrow indicates an ORF. Corresponding genes are indicated the same type of arrow. Open arrows indicate the clpC gene; grey arrows indicate genes with putative functions which are indicated above the arrows; and solid arrows indicate genes encoding hypothetical proteins. Open boxes link genes exhibiting ≥63% amino acid similarity, and genes exhibiting ≤62% amino acid similarity are linked by shaded boxes. The levels of amino acid identity to the B. breve UCC 2003 sequence, expressed as percentages, are also indicated. (b) Alignment of the amino acid sequences of ClpC from high-G+C-content gram-positive bacteria and B. subtilis. The regions containing the two ATP-binding domains (ATP-1 and ATP-2) are enclosed in boxes. The amino acid signatures are indicated. Bl, B. longum NCC 2705; Bb, B. breve UCC 2003; Mt, M. bovis AF 2122/97; Cg, C. efficiens YS314; Sc, S. coelicolor A3; Tf, T. fusca YX; Bs, B. subtilis.

In B. breve UCC 2003 the clpC gene is located directly downstream of an ORF with significant homology to the codA gene (encoding a cytosine deaminase) of Clostridium perfringens (55% similarity) and upstream of an ORF homologous to the uspA gene (encoding a presumed universal stress protein) of S. coelicolor A3 (25% similarity).

We also analyzed the gene inventory of the clpC locus of a phylogenetically distant taxon, B. animalis subsp. animalis ATCC 25527. The clpC DNA region of the ATCC 25527 strain was obtained by PCR using conserved bifidobacterial clpC primers (ClpC1 and ClpC2) combined with an inverse PCR strategy. In this distantly related bifidobacterial species the clpC gene is located directly downstream of the uspA gene and upstream of the aspRS gene encoding an aspartyl-tRNA synthetase (Fig. 1a).

Screening of publicly available bacterial genomes did not reveal a similar gene inventory except for other bifidobacteria (Fig. 1a shows some examples). PCR amplification using DNA extracted from bifidobacterial species representing the principal phylogenetic groups of the genus (24) and subsequent sequencing using primer pairs targeting conserved DNA sequences within the uspA and clpC genes (Fig. 2a) yielded the expected amplicons that ranged from 980 to 1,100 bp long for all bifidobacterial species used (Fig. 2b). Therefore, the genetic uspA-clpC constellation was highly conserved and typical for bifidobacterial species. In contrast, on the basis of PCR results it appeared that the clpC-codA gene order is restricted to the species B. breve and to B. longum biotype longum, B. longum biotype suis, and B. longum biotype infantis (Fig. 2c). Moreover, using the same sets of primers (clp-1/clp-2 and clp-3/clp-4) in PCR experiments with DNA extracted from strains that were not bifidobacteria (e.g., Lactobacillus, Lactococcus, and Streptococcus strains) did not yield any PCR products (data not shown).

FIG. 2.

(a) Schematic representation of the PCR strategy employed to investigate the gene inventory of the clpC locus in bifidobateria. (b) PCR products of various Bifidobacterium species obtained by using a PCR approach in which primers clp-1 and clp-2 for the uspA-clpC genetic constellation were used. (c) PCR products of various Bifidobacterium species obtained by using a PCR approach in which primers clp-3 and clp-4 for the clpC-codA genetic constellation were used. Lanes MK contained the molecular size marker X (Roche, United Kingdom); lanes N contained the negative control (no DNA in the PCR mixture).

Based on genome analysis, we found that similar to the situation in the B. longum NCC 2705 genome, homologues of the clpP1, clpP2, clpX, and clpB genes are located in the genome of B. breve UCC 2003, suggesting the presence of a multigene clp family in bifidobacteria (17; Leahy et al., unpublished). It is notable that these genes displayed a low level of DNA homology to the clpC gene (<27%), which eliminated the possibility of cross-hybridization during hybridization experiments.

Heat induction of the clpC gene in B. breve UCC 2003.

Expression of the clpC gene in high-G+C-content gram-positive bacteria is induced by a number of protein-denaturing stress treatments (15), among which heat and osmotic stresses are well documented (21, 30). In a previous study we showed that induction of dnaK and clpB chaperone genes in B. breve UCC 2003 is most effective at NaCl concentrations of 0.5 M and 0.7 M (20, 21; M. Ventura, J. G. Kenny, G. F. Fitzgerald, and D. van Sinderen, unpublished results). To determine if induction of clpC occurs upon exposure to stressful conditions in B. breve UCC 2003, slot blot hybridization was used to analyze total RNA isolated from B. breve cultures following exposure for up to 150 min to temperatures ranging from 20°C to 50°C and to NaCl concentrations of 0.5 M and 0.7 M (Fig. 3a).

FIG. 3.

Heat, cold, and osmotic shock induction of the B. breve UCC 2003 clpC gene. Total RNA was isolated from B. breve UCC 2003 following exposure to various extreme temperatures for specific times and was analyzed by slot blot hybridization. (a and c) Slot blot hybridization using RNA extracted from cells incubated for up to 150 min at a range of temperatures or with different NaCl concentrations using the clpC gene (a) or the 16S rRNA gene (c) as a probe. (b) Schematic representation of the mRNA level of induction. The different types of bars indicate the times for which heat and osmotic shocks were applied. The fold induction compared to the control culture (UCC 2003 cells grown at 37°C in modified MRS medium with no added NaCl) is indicated above each bar.

Based on the strength of the hybridization signal, the strongest expression of the clpC gene occurred at 43°C following 150 min of exposure, whereas exposure to higher temperatures, to high NaCl concentrations, or to a low temperature (20°C) did not appear to significantly increase the level of clpC transcription (Fig. 3a). Densitometric analysis of Northern slot blots revealed that the levels of clpC mRNA increased 15-fold when cells were subjected to heat shock at 43°C after 150 min (Fig. 3b). Moreover, an amount of total RNA identical to that used in the induction experiment was slot blotted and hybridized using the 16S rRNA gene as a probe in a control experiment. The strength of the hybridization signal was similar for all the slots, thus confirming the uniformity of the RNA samples employed in the induction experiments (Fig. 3c).

Characterization of clpC transcription activity by Northern blotting and primer extension analysis.

Transcription of clpC was further evaluated using cultures grown at 37°C or cells subjected to elevated temperatures (43°C and 47°C) or to osmotic shock following addition of NaCl by Northern blot analysis using DNA fragments corresponding to the clpC gene or surrounding genes as probes (Fig. 4a). Hybridization with a clpC-specific probe revealed a 2.6-kb transcript, as well as a smaller 1.6-kb transcript (Fig. 4a), which may have been derived from specific cleavage and/or degradation of the larger 2.6-kb transcript. In fact, analysis of the DNA sequence of clpC revealed a number of inverted repeats within the open reading frame which could form potential stem-loop structures and thereby serve as cleavage and/or processing sites (Fig. 4a), a situation previously described for the clpC gene of Lactococcus lactis (6, 23) and Streptococcus mutans (11). The significance of this phenomenon is unknown, but cleavage at processing sites may lead to a product with a changed half-life and may thus be involved in the regulation of clpC expression.

FIG.4.

(a and b) Northern hybridization analysis of B. breve UCC 2003 mRNA of the clpC locus (a) and primer extension analysis of the putative promoter sequences for the clpC gene (b). (a) Positions of the transcripts on the clpC genome map. The estimated sizes of the mRNA are indicated. Hairpins indicate possible rho-independent terminators. 23S rRNA and 16S rRNA indicate the positions of the 23S rRNA and 16S rRNA transcripts. (b) Primer extension results obtained by using an oligonucleotide targeting the 5′ ends of the clpC gene. The experiment was performed using cells grown at different temperatures. Underlining indicates the −10 putative hexamer; an asterisk indicates the transcription start point; IR indicates an inverted repeat; and G indicates the position of the guanine residue which corresponds to the transcription start site. (c) Partial comparison of the putative promoter sequences for clpC homologues from five bifidobacterial species. Bl, B. longum NCC 2705; Bs, B. longum type suis LMG 21814; Bb, B. breve UCC 2003; Ba, B. animalis subsp. lactis LMG 18906; Bd, B. dentium JCM 1195.

The shift to heat shock conditions (43°C) clearly increased the strength of expression of the transcripts (Fig. 4a). In agreement with the slot blot hybridization results, it was found that transcription of the clpC gene of B. breve UCC 2003 appeared to be only weakly activated following exposure to 47°C, whereas specific induction of clpC was not obvious following osmotic stress due to exposure to 0.5 M and 0.7 M NaCl for 50 min (Fig. 4a). Northern analysis of the DNA sequences surrounding the clpC gene using RNA extracted from unstressed and heat-shocked cells did not reveal any specific transcripts, indicating that the uspA and codA genes are not part of the clpC locus and do not appear to be expressed at a detectable level under the conditions used.

The transcriptional start site was identified upstream of the assumed start codon of the clpC gene by primer extension analysis using total RNA isolated from heat-shocked and unshocked cells (Fig. 4b). The primer extension results were confirmed using a second primer, clpC-prom2 (data not shown). Analysis of the putative promoter regions revealed a potential vegetative promoter-like sequence having a putative −10 hexamer (TAGAGT), while no obvious consensus −35 sequence was found (Fig. 4c). The regions upstream of the clpC genes of B. breve UCC 2003 and B. longum NCC 2705 were aligned in an attempt to identify putative regulatory elements. For completeness, we also determined the putative promoter regions of the clpC genes for the closely related organism B. longum type suis and for two more distantly related Bifidobacterium species (Bifidobacterium dentium and B. animalis subsp. lactis). As shown in Fig. 4c, a consensus promoter sequence could be deduced from the five sequences. The putative −10 region, the ribosome binding site region, and the transcriptional start site were conserved in all bifidobacterial sequences examined. Moreover, a number of other DNA motifs were shown to be conserved in all the strains used. Notably, a 4-bp (ATTT/C) inverted repeat was detected in the region upstream of the −10 box of all bifidobacterial clpC promoter regions, which might have a regulatory function (Fig. 4b and 4c).

clgR homologous gene in B. breve UCC 2003.

Analysis of the B. breve UCC 2003 genome sequence, as well as the completed B. longum NCC 2705 genome sequence (16), revealed a gene homologous to clgR, which could be responsible for regulation of clp genes. The encoded protein, which is 189 amino acids long and has a molecular mass of 20.89 kDa, exhibits 46% identity with ClgR of C. glutamicum and 42% identity with the homologous protein in S. lividans (5). Furthermore, the bifidobacterial ClgR-like protein possesses a helix-turn-helix motif which is typical of transcriptional regulators. The clgR-like genes in the UCC 2003 and NCC 2705 genomes are located downstream of the ftsK and pgsA3 genes, in a genetic constellation similar to those described for other members of the Actinobacteridae group (5). Alignment of the ClgR-like proteins of B. breve UCC 2003 and B. longum NCC 2705 with those of other high-G+C-content gram-positive bacteria indicated that the amino acid residues present in the helix-turn-helix region are highly conserved (Fig. 5a). Previous studies defined the three-dimensional structure of this category of transcriptional regulators (5, 33). Therefore, the amino acid residues within the helix-burn-helix which interact with DNA and thereby confer recognition sequence specificity could be identified. The amino acid residues in the ClgR protein of strain UCC 2003 that is homologous to the proteins shown to interact with DNA in the clpC gene of C. glutamicum are highlighted in Fig. 5a. All these amino acid residues are conserved in the B. breve UCC 2003 and B. longum NCC 2705 ClgR proteins, indicating that ClgR in these bacterial species interacts with very similar recognition sequences.

FIG. 5.

(a) Alignment of ClgR proteins from several Actinobacteridae. The amino acid sequence of ClgR of B. breve UCC 2003 (Bb) was aligned with the sequences of the ClgR proteins of S. coelicolor A3 (Sc), M. tuberculosis CDC1551 (Mt), T. fusca (Tf), C. glutamicum ATCC 13005 (Cg), and B. longum NCC 2705 (Bl). Shading indicates conservation at a position in at least 50% of the proteins in the alignment; black shading indicates identical residues, and gray shading indicates similar residues. Dotted and solid lines above the aligned sequences indicate α-helices and turns, respectively, which encompass the DNA-binding domain. The amino acid residues involved in DNA binding are indicated by dots below the aligned sequences. (b, c, d, and e) Gel retardation assays were performed by incubating 0 to 1.5 μg of crude lysate from IPTG-induced E. coli(pQE-clgR) (b), 1 μg of crude lysate from B. breve UCC 2003 cultures exposed to 43°C (c), 1 μg of UCC 2003c from cells grown at 37°C (d), or 1 μg of UCC 2003c from cells exposed to 43°C after pronase treatment (e) with radiolabeled clpCp as the probe. ClgRc, crude lysate from IPTG-induced E. coli(pQE-clgR). A plus sign indicates the presence of UCC 2003c. The arrow indicates the lane that contained 1 μg of crude lysate from E. coli(pQE). Panel e shows the gel retardation results obtained by using clpCp as the probe and ClgRc plus UCC 2003c which prior to the binding experiment had been treated with pronase (+) or incubated in the absence of pronase (−).

clgR binds to the clpC promoter region.

In other members of the Actinobacteridae group, like S. lividans and C. glutamicum, the ClgR protein was shown to bind the clpC promoter regions, indicating that it acts as a transcriptional activator of the clpC gene.

In order to determine whether ClgR of B. breve UCC 2003 binds to the promoter region of the clpC gene, gel retardation experiments were performed. The ClgR protein of strain UCC 2003 was overproduced in E. coli, and a total cell extract was used in gel mobility shift DNA binding assays with 200-bp radiolabeled DNA fragments corresponding to the promoter region of the clpC gene (clpCp). The data showed that the crude extract from E. coli overproducing ClgR (ClgRc) did not affect the mobility of the clpCp fragment (Fig. 5b and 5c). In contrast, when ClgRc was used together with 1 μg of crude extract from B. breve UCC 2003 cells, designated UCC 2003c, that had been subjected to heat stress (43°C for 150 min), the mobility of the clpCp fragment was clearly reduced (Fig. 5c). In contrast, no difference in migration was observed when the clpCp fragment was incubated with 1 μg UCC 2003c from heat-shocked B. breve cells without the presence of ClgRc (Fig. 5c). Moreover, when ClgRc was used together with 2 μg of crude extract from B. breve UCC 2003 cultures grown at 37°C, no DNA shift of the clpCp fragment was observed (Fig. 5d). This suggests that ClgR binds to the clpC promoter in B. breve UCC 2003. Of note, a coactivator(s) may be involved in such an interaction. Similar findings have been obtained for other bifidobacterial stress-induced genes (e.g., dnaK and clpB), where coactivators were shown to be indispensable for DNA binding (Ventura, unpublished results).

In order to characterize the nature of this coactivator(s), protease treatment (with pronase) of UCC 2003c followed by pronase inactivation by chemical means was performed. The treated UCC 2003c was subsequently used in gel mobility shift assays (Fig. 5e). Interestingly, when an amount equivalent to 1 μg (2.5 μl) of pronase-treated UCC 2003c was incubated with 1.5 μg of ClgRc, no displacement of clpCp was detected (Fig. 5e). However, clear retardation of the clpCp fragment was observed in a control experiment employing the same amount of UCC 2003c which had been treated in the same way but without addition of pronase (Fig. 5e). Therefore, we concluded that a proteinaceous coactivator(s) may be involved in ClgR binding to the clpC gene from B. breve UCC 2003. Thus, we speculated that the bifidobacterial ClgR homologue is a transcriptional activator for regulation of the clpC gene, as has previously been demonstrated for other high-G+C-content gram-positive bacteria (5). In C. glutamicum and S. coelicolor, clpC gene expression is controlled by a regulatory network involving the transcriptional activator ClgR, as well as the ECF sigma factor σ^H (5). Interestingly, no stress-induced sigma factors like σ^H and σ³² have been identified so far in the B. breve UCC 2003 or B. longum NCC 2705 genome sequences (16, 25). This suggests that alternative molecules (e.g., a coactivator[s]) may be involved by playing a role similar to that of the stress-induced sigma factor in other members of the Actinobacteridae group.

Estimation of the number of clpC genes in Bifidobacterium genomes.

Members of the genus Streptomyces, which also belongs to the Actinobacteridae, contain several clpC homologues in their genomes (1). To determine whether the members of the genus Bifidobacterium also contain multiple copies of the clpC gene, Southern hybridization of EcoRI- and EcoRV-digested genomic DNA from 11 bifidobacterial species was performed with the clpC gene (data not shown). Each of the bifidobacteria investigated yielded a single band; the bands were different sizes in all of the strains used except one or two strains in which apparent incomplete digestion yielded additional but less intense bands. This clearly suggests that just a single copy of the clpC gene is present in these bifidobacterial genomes. This finding was confirmed by sequence analysis of the entire complete genomes of B. breve UCC 2003 and B. longum NCC 2705 (16) and the incomplete B. longum DJO10A (DOE Joint Genome Institute) genome, each of which harbors a unique copy of the clpC gene.

Conclusions.

In contrast to other high-G+C-content gram-positive bacteria, which contain multiple copies of ClpC chaperone-encoding genes (1), the bifidobacterial genomes investigated in this study were shown to contain a single clpC homologue, which is flanked by a homologue of the uspA gene (9). Moreover, in B. breve UCC 2003, B. longum NCC 2705, and B. longum DJO10A other stress-inducible genes (e.g., groEL, cspA, and cspB) are located in the same chromosomal region as the clpC gene (16, 25; DOE Joint Genome Institute), suggesting that there is deliberate clustering of stress response genes in these genomes.

Expression of the B. breve UCC 2003 clpC gene was found to be induced at the highest levels by moderate heat shock regimens (temperature difference between the optimal growth temperature and the applied stress temperature, 6 K) but not by low-level temperature exposure (20°C), by severe heat stress (temperature difference between the optimal growth temperature and the applied stress temperature, 10 to 13 K), or by osmotic stress. These data are in clear contrast to findings for M. tuberculosis (5), S. coelicolor, and C. glutamicum (5, 30), in which ClpC expression is induced only by severe heat stress.

We found that the B. breve ClgR homologue is a likely regulator for clpC expression, although the typical consensus ClgR binding motif (CGC-N₅-GCG) (5) is absent in the promoter region of bifidobacterial clpC genes. This finding, together with the fact that a proteinaceous coactivator molecule is probably needed for clpC regulation in the UCC 2003 strain, suggests that clpC regulation in bifidobacteria depends on a different regulatory mechanism than that found in other members of the Actinobacteridae group (1, 5). Future research will be focused on identification of the ClgR operator site in bifidobacteria by means of DNase footprinting experiments.

ClpC is very well conserved across the archaeal, eubacterial, and eukaryal domains of life, and consequently the clpC gene is included in the category of housekeeping genes. Due to this ubiquitous distribution, functional preservation, and sequence conservation, the clpC gene is considered a valuable molecular marker for bacterial phylogeny (12). With the introduction of polyphasic taxonomy (12, 22) many molecular markers that are alternatives to 16S rRNA gene sequences have been described for tracing bifidobacterial phylogeny, such as the tuf (28, 29), atpD (13, 27), groEL (7, 25) recA (28), dnaK (25), and grpE (25) genes, all of which offer the advantage of using amino acid sequences to infer bacterial phylogenies, thereby avoiding the problems associated with rRNA gene sequences (28). Use of the clpC gene as another molecular marker should be useful for corroborating and completing the evolutionary history of bifidobacterial species.

In summary, the data presented in this study offer the first insights into clpC expression, regulation, and genetic organization in bifidobacteria and could provide an additional molecular marker for taxonomic and bacterial tracing purposes.