Regulation of β-Galactosidase Expression in Bacillus megaterium DSM319 by a XylS/AraC-Type Transcriptional Activator

Cloning and characterization of the β-galactosidase and its regulator.Bacterial strains, plasmids, and phage vectors used in this study are listed in Table 1. A λ-EMBL4-based genomic library of B. megateriumDSM319 (18) was used for cloning the β-galactosidase-encoding region. The lac-negative E. coli strain JM107 was infected and plated on 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal)-containing medium. Positive clones were obtained as blue plaques. Subcloning a 6.8-kb internal XbaI fragment into plasmid vector pUCBM20 (pUCGal3 [Table 1]) gave rise to dark blue colonies. Sequence analysis revealed four complete and two incomplete open reading frames (ORFs) (Fig. 1).

• Open in new tab
• Download powerpoint

Fig. 1.

Schematic representation of the β-galactosidase-encoding genomic region of B. megateriumDSM319 wild-type and mutant strains. ORFs are indicated as arrows, the direction of which corresponds to transcription. Truncated ORFs are represented by 4′ and 5′. The same designation was chosen for genes inactivated by insertional mutagenesis. Potential Rho-independent terminators are marked by hairpins. bgaM, β-galactosidase gene; bgaR, XylS/AraC-type regulator gene; bglM, β-1,3-1,4-glucanase gene of P. macerans; nprM, neutral protease gene. ΔbgaR lacks 263 nt of the coding region. Restriction enzymes are abbreviated as follows: A,AflII; E, EcoRV; H, HincII; S,SalI; X, XbaI. The positions of probes used in Northern blot analyses are depicted from the top. Designations of corresponding strains are indicated on the right side.

The predicted polypeptide encoded by ORF bgaM (1,034 amino acids [aa]) exhibits striking similarities to β-galactosidases of several bacteria, such as Clostridium acetobutylicum(45% identity [11]), Streptococcus thermophilus (40% [24]), E. coli (36% [15]), and Klebsiella pneumoniae (36% [6]). The recently published sequence of the β-galactosidase of B. megaterium ATCC 14581 is almost identical (97% [25]). Accordingly, as for strain ATCC 14581, disruption of the chromosomal copy by insertional mutagenesis resulted in a β-galactosidase-deficient strain (MS981 [Fig. 1]). The gene was disrupted by applying a previously established single-copy replacement system (29). For insertional mutagenesis of bgaM, the following steps were performed. A 3.14-kb HindIII fragment of plasmid pUCGal3 was cloned intoHindIII-digested pUCTV2, resulting in pUCGalD. As the insert, a 2.38-kb chromosomal fragment containing the entirenprM gene of B. megaterium (18) was amplified with oligonucleotides 5′-GTTTTTATCAGTCGACTTGAAGAATTATT-3′ and 5′-GTAATGAAGTCGACCCAGTTTCCCTTTTATTAC-3′, cut withSalI, and ligated to SalI-digested pUCGalD; the resulting disruption vector was designated pDGal2. The genetic organization of the bgaM mutant strain MS981 is outlined in Fig. 1.

The polypeptide (275 aa) predicted for the divergently orientated locusbgaR exhibits obvious similarities to members of the XylS/AraC family of regulators. Within the highly conserved C terminus is situated the family profile, spanning aa 173 to 272 (entry PS01124 in PROSITE database). As for other XylS/AraC proteins, which are in most cases approximately 250 to 300 aa in size, the polypeptide has two helix-turn-helix DNA-binding motifs located within the family profile. Furthermore, the gene has a polarity opposite to that of the trait that is regulated as frequently observed for members of this family, the vast majority of which are positive transcriptional regulators (reviewed in references 9 and10).

Gene products of ORF2, ORF3 (158 and 160 aa), and ORF5 resemble hypothetical proteins of B. subtilis (16). The deduced amino acid sequence of truncated ORF4 was found to contain an ATP-binding cassette transporter family signature (for a review, see reference 23).

Two potential Rho-independent transcriptional terminators were identified downstream of bgaM and ORF2, respectively (Fig.1), and two overlapping CRE were found within the short intergenic region (129 nucleotides [nt]) of the β-galactosidase gene andbgaR (see Fig. 3).

Construction of bgaR and ORF2 mutant strains.β-Galactosidase expression in B. megaterium is subject to catabolite repression mediated by cis-acting CRE and the CcpA protein (25) but is in addition inducible by β-galactosides (Fig. 2A) (17). Because of the similarities of BgaR to XylS/AraC transcriptional regulators, it was tempting to disrupt the gene and thereby determine its function in β-galactosidase expression. We applied the already-mentioned gene replacement system (29) to generate a set of B. megateriummutants, the genetic structures of which are schematically presented in Fig. 1. A bgaR mutant strain, MS970, was generated by using the pUCTV2 derivative pDbgaR containing a 1.76-kb chromosomal fragment amplified with oligonucleotides 5′-CCATGACGGAATTCCCCAGCTG-3′ and 5′-GAAGTTATATCGAATTCAGCTGGAG-3′; in addition, it carries the β-1,3-1,4-glucanase gene (bglM) of Paenibacillus macerans (5), which was amplified from plasmid pBCmac with oligonucleotides 5′-ACCCAGGTAAAAGCTTCCAACACCG-3′ and 5′-GGACTAAAAAGCTTATGCCCAGCTGC-3′. The glucanase gene was subsequently ligated into the EcoRV site of thebgaR locus. Since we could not exclude a polar effect of the integrated bglM gene on expression of ORF2, which appeared to be part of the bgaR transcriptional unit (Fig. 2B), we also generated a bgaR deletion mutant (MS971 in Fig. 1). Deletion was achieved with vector pD2bgaR, which carries the above-mentioned bgaR locus, from which 263 nt betweenHincII and EcoRV was removed.

• Open in new tab
• Download powerpoint

Fig. 2.

(A) β-Galactosidase activities for wild-type and mutant B. megaterium strains. Strain genotypes are depicted in Fig. 1. Hatched columns correspond to enzyme activities measured without additional sugars (LB medium), black columns represent activities of cultures grown with lactose, and white columns refer to cultures to which lactose and glucose were added simultaneously. Values given in Miller units (19) are the means of two experiments. Compared to LB medium without sugar, glucose addition resulted in lowered β-galactosidase activity of all strains (catabolite repression). Lactose as the sole carbon source caused a clear increase in enzyme activity in wild-type strain DSM319 and the ORF2 mutant MS972, whereas in mutant strains MS970 and MS971, in whichbgaR was inactivated, inducibility is lacking. (B) Transcript analyses of the bgaM gene and the bgaRoperon. Total RNA isolated from cultures of the wild-type strain DSM319 and bgaR deletion mutant MS971 grown in LB broth without sugars (lanes 1), with lactose (lanes 2), and with glucose and lactose (lanes 3) was subjected to Northern blot analysis. On the left, sizes of hybridizing transcripts (in kilobases) and positions of 16S and 23S rRNAs are indicated. Sizes of the bgaM and bgaRtranscripts, 3.1 and 1.5 kb, respectively, correspond to the length of the putative transcriptional units outlined in Fig. 1. Transcripts ofbgaR mutant MS971 are shortened as expected from the degree of deletion. The occurrence of minor transcripts may be due to degradation or displacement caused by rRNAs.

For checking a potential influence of ORF2 on bgaMexpression, mutant strain MS972 (Fig. 1) was constructed by employing disruption vector pDORF2, which carries a 1.72-kb chromosomal fragment generated with oligonucleotides 5′-TTTTTTCATGCGAATTCTTTCAGCCAC-3′ and 5′-CAATTTATGAATTCGGGAAAAATCGTCC-3′. ThebglM gene of P. macerans was integrated into the internal AflII site. Amplification ofbglM was done from plasmid pBCmac with primers 5′-GGGGGGCTTAAGTAAAATATTCCAACACCGTGG C-3′ and 5′-CCTGGAGTGGACTTAAGAGCTCATGCCCAGCTG-3′.

Comparative analysis of wild-type and mutant strains.The role of bgaR and ORF2 in regulation of β-galactosidase expression was examined by measuring β-galactosidase activities in wild-type and mutant strains. Strains were grown in Luria-Bertani (LB) medium to which sugars were added during mid-log phase. Samples were collected after cultivation for an additional hour and assayed for β-galactosidase activity by measuring hydrolysis of the chromogenic substrate o -nitrophenyl-β-d-galactopyranoside (ONPG) at 31°C (21).

In all strains examined, similar levels of enzyme activity were obtained when no sugar was added to LB broth irrespective of the genetic background. Likewise, in glucose-lactose-grown cultures enzyme activities were lowered to comparable levels in all strains, reflecting catabolite repression of the β-galactosidase. As for the wild-type B. megaterium DSM319, lactose caused a clear increase in β-galactosidase activity in the ORF2::bglM mutant MS972. Thus, the predicted ORF2 polypeptide does not play a crucial role in β-galactosidase induction. In contrast, both bgaR mutants, MS970 and MS971, completely lack inducibility. However, inducibility was restored when the amplified bgaR gene (5′-TATGACCTAGATAAAACCCG-3′ and 5′-CTCATTTCTAGAGTATTAGGTTATTAG-3′) was integrated into the SmaI site of pSVB2 (30) and introduced into mutants (data not shown).

To confirm the influence of the bgaR gene product on transcription of bgaM, both the deletion mutant MS971 and the wild-type strain DSM319 were subjected to comparative Northern blot analyses. An internal 1.75-kbHindIII/SalI fragment served as abgaM-specific probe. Additionally, for gathering information on regulation of bgaR expression, we included abgaR-specific probe, synthesized by PCR (5′-TATGACCTAGATAAAACCCG-3′ and 5′-CTCATTTCTAGAGTATTAGGTTATTAG-3′). Both probes, positions of which are indicated in Fig. 1, were digoxigenin labeled by in vitro transcription with the digoxigenin RNA labeling kit (Boehringer Mannheim, Mannheim, Germany). Total RNA from cultures grown under the conditions described for enzyme assays was isolated (30). Following gel electrophoresis in 1.5% (wt/vol) formaldehyde agarose gels, Northern blot hybridization was carried out at 55°C and chemiluminescence detection was performed (3). The results of these experiments are shown in Fig. 2B.

As for DSM319, low-level transcription of the bgaM gene occurs in MS971 in LB medium without any sugar added. Due to catabolite repression, bgaM transcripts were not detectable in cultures supplied with glucose and lactose. However, addition of lactose caused a clear increase of bgaM transcription only on the wild-type background. In the bgaR deletion mutant MS971, no accumulation of bgaM transcripts occurred. These findings correspond to those of the enzymatic assays summarized in Fig. 2A and are in accordance with positive transcriptional regulation ofbgaM by BgaR.

In medium lacking glucose, bgaR transcription could readily be observed in both the wild-type strain and mutant MS971 irrespective of lactose addition. As depicted in Fig. 2B, transcripts of thebgaR operon in mutant MS971 are shortened by approximately 300 nt, which accords well with the 263-bp deletion of bgaR. Remarkably, transcription of bgaR is—as forbgaM—also subject to catabolite repression; no transcript was detected when glucose was added along with lactose.

Based on our results, we propose a model for regulation of β-galactosidase expression in B. megaterium as outlined in Fig. 3. The transcriptional start point (data not shown) as well as the structure of the bgaM gene and its promoter region (Fig. 3) is almost identical to that published for strain ATCC 14581. A 27-bp inverted repeat overlaps thebgaM promoter containing two partially overlapping CRE. Base substitutions within both overlapping CRE caused partial relief of catabolite repression, and that is why it was suggested that the inverted repeat may serve as the binding site for the catabolite repressor proteins (25). The trans-acting catabolite control protein CcpA is effective by binding to CRE, which need not necessarily overlap the respective promoter but can be located at a considerable distance upstream or downstream (14). Thus, though we did not determine the transcriptional start of thebgaR operon, the localization of the CRE within the intergenic region spanning only 129 nt probably facilitates negative regulation by CcpA of both the bgaM gene and the oppositebgaR operon.

• Open in new tab
• Download powerpoint

Fig. 3.

Regulation of β-galactosidase expression in B. megaterium. The bgaM gene and loci of thebgaR operon are represented by arrows indicating transcriptional orientation (gene signature as in Fig. 1). Transcriptional terminators are sketched as hairpins. The sequence of the intergenic region between bgaM and bgaR is written on top with start codons marked by small arrows. Potential ribosomal binding sites (S/D) were found at appropriate distances. −35 and −10 regions of the bgaM promoter are boxed. The corresponding start point of the bgaM transcript (TS) is indicated by a dot. Two overlapping CRE are located within the promoter region. The arrows underneath indicate direct imperfect repeats (DR) constituting a putative regulator binding site for BgaR. In the presence of glucose, phosphorylated HPr causes CcpA repressor binding to CRE, which facilitates catabolite repression of both divergently orientated transcriptional units. Absence of glucose leads to weak transcription of bgaM but clear transcription of thebgaR operon. When lactose is added exclusively, it acts as an effector of BgaR, eventually resulting in transcriptional activation of bgaM.

For B. megaterium ATCC 14581, the bgaR gene apparently corresponds to a locus that was incompletely sequenced and designated ORF1 by Shaw et al. (25). The predicted polypeptide of the latter comprising 33 aa is almost identical to the corresponding N terminus of the bgaR gene product of DSM319; only two residues are different.

Concerning positive regulation of β-galactosidase expression, we searched for a putative BgaR regulator binding site within the intergenic region of bgaM and bgaR. We identified two imperfect direct repeats again in both DSM319 and ATCC 14581 (DR in Fig. 3). As for the bgaM promoter, such sites for XylS/AraC regulators proximal to the RNA polymerase binding region have been found in most cases to overlap or abut the −35 region of the gene or operon that is regulated (10). Gel mobility shift assays using purified bgaR gene products should help to elucidate the role of these imperfect direct repeats in binding the regulator.

Regulation of β-galactosidase expression appears to be different inB. subtilis. Besides catabolite repression by CcpA, induction of enzyme expression was found to be negatively controlled by another repressor protein belonging to the LacI/GalR regulator family (7).

To our knowledge, positive gene regulation of β-galactosidase expression by a member of the XylS/AraC family has been reported only for Staphylococcus xylosus, in which, however, formation of the regulator is not subject to catabolite repression but is obviously constitutive (1).

Catabolite repression of the regulator BgaR makes the B. megaterium genetic system controlling β-galactosidase expression extraordinarily economic and elegant likewise since it links synthesis of the transcriptional regulator to glucose consumption.

Nucleotide sequence accession number.The sequence of the 6.8-kb chromosomal XbaI fragment of B. megateriumDSM319 containing the genes bgaR and bgaM has been submitted to the EMBL database under accession no. AJ000733 .