Analysis of the ß-glucoside utilization (bgl) genes of Shigella sonnei: evolutionary implications for their maintenance in a cryptic state

Arun S. Kharat1 and S. Mahadevan1

Department of Molecular Reproduction, Development and Genetics, Indian Institute of Science, Bangalore 560012, India1

Author for correspondence: S. Mahadevan. Tel: +91 80 309 2607. Fax: +91 80 360 0683. e-mail: mahi{at}serc.iisc.ernet.in


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The pattern of expression of the genes involved in the utilization of aryl ß-glucosides such as arbutin and salicin is different in the genus Shigella compared to Escherichia coli. The results presented here indicate that the homologue of the cryptic bgl operon of E. coli is conserved in Shigella sonnei and is the primary system involved in ß-glucoside utilization in the organism. The organization of the bgl genes in S. sonnei is similar to that of E. coli; however there are three major differences in terms of their pattern of expression. (i) The bglB gene, encoding phospho-ß-glucosidase B, is insertionally inactivated in S. sonnei. As a result, mutational activation of the silent bgl promoter confers an Arbutin-positive (Arb+) phenotype to the cells in a single step; however, acquiring a Salicin-positive (Sal+) phenotype requires the reversion or suppression of the bglB mutation in addition. (ii) Unlike in E. coli, a majority of the activating mutations (conferring the Arb+ phenotype) map within the unlinked hns locus, whereas activation of the E. coli bgl operon under the same conditions is predominantly due to insertions within the bglR locus. (iii) Although the bgl promoter is silent in the wild-type strain of S. sonnei (as in the case of E. coli), transcriptional and functional analyses indicated a higher basal level of transcription of the downstream genes. This was correlated with a 1 bp deletion within the putative Rho-independent terminator present in the leader sequence preceding the homologue of the bglG gene. The possible evolutionary implications of these differences for the maintenance of the genes in the cryptic state are discussed.

Keywords: cryptic genes, transcriptional activation, insertion elements, antitermination

The GenBank accession number for the bglR–bglG sequence reported in this paper is AF183894.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Members of the family Enterobacteriaceae exhibit substantial differences in their ability to utilize ß-glucosides such as salicin, arbutin and cellobiose. Though earlier studies have confirmed this at the biochemical level (Schaefler & Malamy, 1969 ), the genetic basis for this difference is not known.

Among the four different genetic systems involved in ß-glucoside utilization in Escherichia coli (Prasad & Schaefler, 1974 ; Kricker & Hall, 1987 ; Parker & Hall, 1988 ; Hall & Xu, 1992 ), the bgl operon, encoding the functions necessary for the transport and metabolism of arbutin and salicin, is best characterized. An unusual feature of the E. coli bgl operon is its crypticity. In wild-type cells, the structural genes of the operon are maintained in a functional form, but they are neither expressed nor are they inducible. Though wild-type cells are phenotypically Bgl-, a single mutational event can lead to the simultaneous acquisition of an Arb+ Sal+ (arbutin and salicin-positive) phenotype (Schaefler & Malamy, 1969 ). Mutations that activate the operon map predominantly within the regulatory locus bglR and in most cases are caused by insertion of IS elements upstream of the promoter, leading to enhancement of transcription from the bgl promoter (Reynolds et al., 1981 , 1986 ; Schnetz & Rak, 1992 ). Transcriptional activation is mediated by the disruption of negative elements located near the promoter (Lopilato & Wright, 1990 ; Schnetz, 1995 ; Singh et al., 1995 ; Mukerji & Mahadevan, 1997 ; Caramel & Schnetz, 1998 ). Activation results in the expression of a phospho-enol-pyruvate-dependent phosphotransferase (encoded by the bglF gene) and a phospho-ß-glucosidase (encoded by bglB) which can cleave salicin and arbutin (Fox & Wilson, 1967 ; Prasad & Schaefler, 1974 ). In addition, E. coli also constitutively expresses the enzyme phospho-ß-glucosidase A, specific for arbutin, encoded by the unlinked bglA locus (Prasad et al., 1973 ). In the presence of the phosphotransferase encoded by bglF, cells can exhibit an Arb+ phenotype even when the bglB gene of the operon is non-functional.

Regulation of the operon in response to the availability of ß-glucosides is brought about by a mechanism involving antitermination of transcription, mediated by the bglG and bglF gene products. BglG and BglF represent a novel class of regulatory systems, with the BglF permease acting as the sensor and the BglG antiterminator functioning as the response regulator (Amster-Choder & Wright, 1993 ; Mahadevan, 1997 ; Rutberg, 1997 for a review).

Even though recent studies on the transcriptional activation of the bgl operon have provided a wealth of information on the molecular mechanisms involved in its silencing, from an evolutionary point of view, its crypticity remains a puzzle. This is made more enigmatic by the differential capabilities of closely related members of the Enterobacteriaceae to utilize ß-glucosides. Earlier biochemical studies on five different isolates of Shigella sonnei showed that they fall into two classes, class I strains that are capable of mutating to Sal+ in two steps and class II strains that are unable to mutate to Sal+ even after prolonged incubation (Schaefler & Malamy, 1969 ). To date, no information at the genetic or molecular level is available that can offer a satisfactory explanation for these differences. The results of our studies reported here provide information about the differences in the organization and expression of the bgl genes in S. sonnei in comparison to those of E. coli. Understanding the divergence of the bgl genes at the functional and molecular level in these two closely related organisms is of significance in gaining insights on the reasons for the maintenance of the operon in a cryptic state against evolutionary pressure.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Isolation of S. sonnei strain AK1.
Strain AK1 was isolated from the town sewer line, and was identified and characterized using standard microbiological procedures (Falkow & Mekalanos 1990 ; Watanabe & Okamura, 1991 ). The sewage sample was collected and allowed to sediment for 6 h at room temperature. The clear supernatant was streaked on MacConkey lactose medium and incubated at 37 °C for 18 h. Lac- colonies were checked for growth on deoxycholate/citrate medium and confirmed by growth on Shigella/Salmonella agar. A single isolate obtained was checked for biochemical, morphological and culture characteristics and was confirmed as Shigella sonnei.

Bacterial strains and plasmids.
The strains and plasmids used in this study are listed in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 
Isolation of Bgl+ mutants of AK1.
Strain AK1 (Arb- Sal-) was streaked on MacConkey arbutin plates, incubated at 37 °C for 96 h and allowed to papillate. One of the papillae showing a strong Arb+ phenotype was purified on a MacConkey arbutin plate. Strain AK101, derived from the papilla, is Arb+ but remained Sal-. In the next step, strain AK101 was streaked on a MacConkey salicin plate. The plates were incubated at 37 °C for 120 h to allow papillation. The Arb+ Sal+ strain AK102 was derived from one such papilla. Approximate mutation frequencies were calculated from the papillation frequencies. In an independent experiment, 16 Arb+ mutants were isolated from papillae on MacConkey arbutin plates for detailed analysis of the spectrum of activating mutations.

DNA manipulations.
Plasmid isolation, DNA manipulations and Southern analyses were carried out as described by Sambrook et al. (1989) . Isolation of genomic DNA was carried out with minor modifications of a published protocol (Owen & Borman, 1987 ).

PCR analysis.
Oligonucleotide primers used to amplify genomic DNA from Shigella sonnei and E. coli strains were: SM1 (forward, -250), 5'-GTGGATCCATCTTCTACTACGTGAAG-3'; SM2 (reverse, +340), 5'-AGGAATTCGACTTAAGAGTTCGCTTA-3'. PCR products were cloned into pUC19 at the BamHI and EcoRI restriction sites. Plasmids used in in vivo transcription assays were obtained by subcloning BamHI–EcoRI fragments from the above clones in pBR322.

DNA sequencing.
Sequencing of both strands was carried out by the dideoxy chain-termination method using Sequenase, version 2 (US Biochemicals) as described in the manufacturer’s protocol. Sequence was analysed manually as well as using the ScanJet 4C sequence-analysis and gel-scanning software(Kodak).

Analysis of the insertion element within bglB.
Oligonucleotide primers used to amplify the insertion element along with overlapping bglB region from S. sonnei were: SM49 (forward, +3634), 5'-CACCGTTACCCGGAAGAT-3' and SM50 (reverse, +3810), 5'-CAGCGGCTTGATCCCCGC-3'.

PCR products were purified and used for sequencing the flanking regions and part of the insertion element using Taq DNA polymerase (Gibco BRL) as described in the manufacturer’s protocol.

Transcriptional analysis.
S1 nuclease protection assay.
Analysis of transcription in vivo was carried out using oligonucleotides complementary to bgl and bla transcripts, as described previously (Singh et al., 1995 ; Mukerji & Mahadevan, 1997 ). RNA extraction and S1 nuclease assays were carried out using the same procedure. The probes used in the assay are described below. The protected fragments corresponding to the bgl and bla transcripts were visualized after autoradiography. The transcripts were quantitated by measuring the radioactivity corresponding to the bands. The data shown are representative of at least three independent experiments in each case.

The oligonucleotide probes used in the bgl initiation studies are complementary to bgl sequence from -10 to +42. Since there is a transversion at position +25 in the S. sonnei sequence, the probes for E. coli (probe I) and S. sonnei (probe II) differ at this position, indicated by a bold letter (see below). These probes carry a 10 nt noncomplementary region to allow differentiation of the undigested probe from the signal from the protected fragment.

The oligonucleotide probe to quantitate the bglG-specific mRNA (probe III) is complementary to the bglG sequence from +120 to +159. As the S. sonnei and E. coli nucleotide sequences do not show any differences in this region, the same probe was used in the S1 nuclease protection assay in these two organisms. The 3' end of the probe carries an 8 nt random sequence that was deliberately introduced to differentiate between the undigested probe and the signal corresponding to the protected fragment. The probe concentration has been standardized previously (Singh et al., 1995 ) to ensure that it is in molar excess of the mRNA.

The probe I sequence is 5'-CCAGTCATTTATTAATGGTT T T T A T A A CGAACAT C C A G G T T C G G A A A T T A A T - 3'; the probe II sequence is 5'-CCAGTCATTTATTAATGTTT T T T A T A A C G A A C A T C C A G G T T C G G A A A T T A A T - 3'; probe III sequence 5'-GTTGAGAATTTTGGTGATTTGCATGTTCATAGCAAGGACCAACATCGT-3'.

Northern blot analysis.
Estimation of bglB transcription was carried out by Northern blot analysis as described by Sambrook et al. (1989) . Total RNA was extracted from the strains JF201 ({Delta}bgl), AK1 (Arb- Sal-), AK101 (Arb+ Sal-) and AK102 (Arb+ Sal+). RNA (20 µg) was used for blotting. The blots were probed using a probe specific for the E. coli bglB gene, obtained by PCR amplification. A DNA fragment carrying the E. coli ung gene was used to probe the same blot as an internal control. Transcripts were quantitated by scanning the bands obtained after autoradiography using a Hewlett Packard ScanJet 4C with gel-scanning software from Kodak. The bglB transcript was normalized by determining the bgl/ung ratio.

Assay for bglB activity (saliginine assay).
Measurement of phospho-ß-glucosidase B activity, specific for the bglB gene, was carried out using a procedure that was similar to the assay described by Prasad & Schaefler (1974) . Cultures were grown in LB at 37 °C with or without addition of 7 mM salicin as inducer. Cells were harvested, washed with 0·85% physiological saline solution and resuspended in 0·1 vol of the same solution. To 0·1 ml cells, an equal volume of 140 mM salicin was added and the tubes were incubated at 37 °C for 30 min. The reaction was stopped by the addition of 0·5 ml 2 M Na2CO3 solution. After addition of 0·5 ml 0·6% 4-aminoantipyrine solution, the tubes were incubated at room temperature for another 15 min, followed by the addition of 0·5 ml 4% K3[Fe(CN)6] solution. The red colour obtained within 10 min was measured spectrophotometrically at 509 nm. The activity, expressed in units, was calculated as described by Mahadevan & Wright (1987) .


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Isolation of Bgl+ mutants of S. sonnei strain AK1
The wild-type S. sonnei strain AK1, a sewage isolate used in this study, cannot utilize the aryl ß-glucosides arbutin and salicin. Colonies of the strain remain pale yellow on MacConkey arbutin and MacConkey salicin indicator medium even after 72 h. However, on MacConkey arbutin plates, reddish brown papillae were obtained after 96 h incubation at 37 °C. Strain AK101, isolated from one such papilla, showed an Arb+ phenotype, but remained Sal- on MacConkey salicin indicator plates. Upon prolonged incubation of the plates, pale red papillae were observed after 108–120 h. Strain AK102, derived from one such papilla, formed red colonies within 16 h on MacConkey salicin plates and is therefore a salicin-positive derivative of AK1, obtained in two steps. Strain AK1 is similar to the three S. sonnei strains out of five described by Schaefler & Malamy (1969) that can be activated to give a Sal+ phenotype in two steps. Approximate estimates of mutation frequencies are 10–8 (WT to Arb+) and 10–12 (Arb+ to Sal+). This property of S. sonnei is different from that of E. coli strains, which can papillate to Sal+ in a single step within 48 h, suggesting differences in the organization of the bgl genes in the two organisms. This is consistent with an earlier observation that three out of five S. sonnei strains tested showed a pattern of two-step activation (Schaefler & Malamy, 1969 ).

Identification of the bgl genes of AK1
E. coli has at least three different cryptic systems for the utilization of salicin. Since no reports regarding the nature of the genes involved in ß-glucoside utilization in S. sonnei are available, we attempted to characterize the genes responsible for conferring a Bgl+ phenotype to strain AK1. The time involved in mutating to Bgl+ suggested the involvement of the homologue of the E. coli bgl operon since the activation of the cel and asc operons requires incubation for longer time periods in E. coli strains (Kricker & Hall, 1984 ; Hall & Xu, 1992 ). This was confirmed by Southern analysis of genomic DNA from AK1 and its activated derivative AK101 (data not shown). The signal obtained from S. sonnei DNA was as strong as that from E. coli, suggesting a high degree of homology between the bgl sequences in the two organisms.

PCR amplification and nucleotide sequence analysis of the bglR region
Since the Southern analysis suggested the presence of a bgl homologue in S. sonnei, we attempted to further confirm this by amplifying the regulatory region using PCR and characterizing the clones from the wild-type and mutant strains. The bgl regulatory region was amplified using primers designed on the basis of the E. coli bglR sequence. Analysis of the PCR fragments on agarose gels showed that the sizes of the fragments amplified from genomic DNA of S. sonnei are similar to those obtained from E. coli controls, confirming the earlier results of the Southern analysis. The size of the fragments obtained from the Arb+ mutant AK101 showed an increase of about 800 bp, consistent with the possibility that the mutant carries an insertion of DNA within the regulatory region.

Clones containing the PCR fragments derived from the wild-type and Bgl+ mutant strains of S. sonnei were subjected to nucleotide sequence analysis to determine the extent of homology at the nucleotide level, and the site and nature of the insertion. The sequence of the bglR region and a part of the bglG gene are presented in Fig. 1. The sequence of the bglR region in the case of the wild-type strain AK1 differs only in two positions from the E. coli sequence (Schnetz et al., 1987 ). Interestingly, one of the alterations is within the stem of the putative Rho-independent terminator in the leader sequence preceding the first structural gene. A similar mutation, deletion of a G residue at position +100 generated after mutagenesis, was earlier shown to result in partially constitutive expression of a reporter gene and increased readthrough of transcription in vitro in E. coli (Mahadevan & Wright, 1987 ). The functional significance of this change is discussed below. The second difference is a C to A transversion at position +25 within the leader region. The partial sequence of the first structural gene is identical to that of the E. coli bglG gene (with the exception a transversion and two transitions), suggesting that the S. sonnei bgl operon may also be regulated by a mechanism involving antitermination of transcription.



View larger version (49K):
[in this window]
[in a new window]
 
Fig. 1. Nucleotide sequence of bglR region from S. sonnei strain AK1. Bases marked with arrows indicate differences seen from the E. coli sequence. The inverted repeat within bglR, one of the silencer elements characterized in E. coli, is indicated by broken arrows. The Rho-independent terminator preceding bglG is indicated by solid arrows. The position of the IS1 insertion is shown by an open inverted triangle and the 9 bp duplication flanking the insertion is underlined. The filled triangle indicates the position of the mutation at +100. The putative promoter elements are boxed.

 
Sequence analysis of the PCR fragment from the Arb+ strain AK101 showed the presence of an extra sequence of DNA within bglR. The partial sequence of this element suggests that it is IS1, reported previously in S. sonnei (Lawrence et al., 1992 ) (data not shown). The site of insertion (Fig. 1) also suggests that the S. sonnei bgl operon may also be activated by a mechanism similar to that of E. coli. The 9 bp target duplication brought about by IS1 is underlined in Fig. 1.

Analysis of the pattern of activation of the bgl operon in S. sonnei
In E. coli, 95–98% of the mutations that activate the bgl operon are linked to the operon and only 2–5% of the activating mutations are unlinked, involving loci such as hns, gyr, bglJ and leuO (Defez & DeFelice, 1981 ; DiNardo et al., 1982 ; Giel et al., 1996 ; Ueguchi et al., 1998 ). To check whether the pattern of activation is identical in S. sonnei, the 16 independently isolated Arb+ mutants of AK1 (see Methods) were analysed to determine the nature of the activating mutation. Genomic DNA from the mutants was used for PCR amplification and Southern analysis of the bglR region. One mutant showed a pattern similar to strain AK101 that carries an IS1 insertion in bglR. The remaining 15 mutant strains showed a pattern similar to the parent strain AK1, indicating the absence of insertions within bglR. To examine whether the activating mutations in these mutants are located within any of the known loci, plasmid pHMG409 (Goransson et al., 1990 ), which carries the wild-type copy of the hns gene, was introduced into the mutants by transformation. In 14 mutants, the presence of the wild-type copy of the hns gene resulted in an Arb- phenotype, suggesting that the activating mutations map within the hns gene. The location of the mutation in one strain remains unknown. Therefore, although the bgl genes share a high degree of homology at the nucleotide level and genetic organization in these two organisms, their pattern of activation is different.

To confirm that the results described above are not confined to strain AK1, similar analyses were carried out using three clinical isolates of S. sonnei: SSOR, CR+ and CR-. PCR, Southern and nucleotide sequence analysis of the bglR region revealed that the organization and the pattern of activation are similar in these strains (data not shown). The only difference is in the case of the class II strain SSOR, in which the presence of an endogenous plasmid suppresses the mutational activation of the bgl genes (Kharat & Mahadevan, 1999 ). Activation in this case requires the loss of the plasmid. Therefore, the differences described above are not confined to a single strain, but are common to S. sonnei strains isolated from diverse sources.

Analysis of transcription in vivo in S. sonnei
The studies described above suggest that the activation of the bgl genes in S. sonnei is likely to be at the level of transcriptional enhancement, similar to that documented in E. coli. The presence of the single base pair deletion within the Rho-independent terminator preceding the bglG gene also suggests that the basal level of transcription of bglG may be higher in S. sonnei than E. coli. To test these possibilities, we analysed transcription from the bgl promoter in S. sonnei.

To monitor transcription from promoters derived from the wild-type and activated bgl operon, plasmids carrying bgl promoter fragments (-250 to +340) from strains AK1 and AK101 were used to transform the wild-type strain. RNA isolated from the transformants was hybridized to two end-labelled oligonucleotide probes, one complementary to bglG, located downstream of the terminator, and the other complementary to bla mRNA. After digestion with S1 nuclease, the products were separated by denaturing PAGE. Following autoradiography, the bgl transcripts in each case were quantitated and normalized to the bla transcript. The results obtained showed that the level of transcription from the promoter carrying an insertion element is 16-fold higher than transcription from the wild-type promoter, indicating that the higher expression of the bgl genes in the mutant is brought about by transcriptional enhancement (data not shown).

To test whether the mutation at position +100 has any effect on the basal level of transcription of bglG, transcripts from plasmids carrying the promoter fragments from wild-type strains RV (E. coli) and AK1 (S. sonnei) were quantitated using the oligonucleotides described above. The results, presented in figure 2(c), show that the basal level of transcription downstream of the terminator is significantly higher in S. sonnei than E. coli (lane 3 compared to lane 2). This is likely to result in a higher basal level of the BglG protein in S. sonnei than in E. coli. Therefore, there are intrinsic differences in the expression pattern of the bgl genes in wild-type strains of S. sonnei and E. coli. To ensure that this difference in the basal level of transcription is not related to differences at the level of initiation in the two strains, transcription initiation from the E. coli and S. sonnei wild-type promoters was compared using oligonucleotides complementary to the first 42 nt in each case. These studies show that there is no appreciable difference in the level of transcription initiation from the two promoters (Fig. 2b). Interestingly, the wild-type promoters in the case of E. coli and S. sonnei show detectable transcription. However, this level of transcription is insufficient to confer a Bgl+ phenotype, which requires further enhancement by mutation. Because of efficient termination, no readthrough can be detected downstream in the case of E. coli.



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 2. Transcriptional analysis of the S. sonnei bgl operon by S1 nuclease protection. (a) Features of pANS4-11 and pANS1-3 carrying the wild-type promoters from E. coli and S. sonnei, respectively. The positions of the probes used in the transcriptional analysis are indicated as filled bars. (b) Transcription initiation from wild-type promoters of E. coli and S. sonnei. Total RNA isolated from the wild-type E. coli strain RV, carrying the two plasmids, was used to detect the levels of initiation by the S1 nuclease protection assay (see Methods). Lane 1, probe I; lane 2, transcription initiation from plasmid pANS4-11 (E. coli strain RV) detected using probe I; lane 3, transcription initiation from plasmid pANS1-3 (S. sonnei strain AK1) detected using probe II; lane 4, probe II. (c) Basal transcription of bglG from wild-type promoters of E. coli and S. sonnei detected using probe III. Lane 1, probe III; lane 2, transcription from pANS4-11 (E. coli strain RV); lane 3, transcription from pANS1-3 (S. sonnei AK1).

 
To determine the effect of the mutation at position +100 on the level of bglG transcription in an activated strain, transcripts from plasmids carrying the promoter region from the Arb+ Sal+ strain AK102 were quantitated in the absence and presence of 7 mM salicin as an inducer, using the oligonucleotide probes described above. The results obtained show that the level of bglG transcription downstream of the terminator is similar in both cases, indicating constitutive expression of bglG (Fig. 3). This is likely to have an impact on the regulation of the operon in S. sonnei.



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 3. Transcription from the activated promoter of S. sonnei. (a) Features of plasmid pANS5-16 carrying the activated promoter from S. sonnei strain AK102. The position of the activating IS element is indicated by a filled triangle. (b) Transcription of bglG under inducing and non-inducing conditions. Total RNA was isolated from transformants of the activated strain AK102 carrying the plasmid pANS5-16 (S. sonnei AK102), grown in the presence and absence of 7 mM salicin. The RNA was subjected to S1 nuclease analysis as above using probe III. Lane 1, probe III; lane 2, transcription from pANS5-16 in the absence of inducer; lane 3, transcription from the plasmid in the presence of inducer.

 
Functional characterization of the bgl genes in S. sonnei
The two major differences with respect to the activity of the bgl genes in S. sonnei are the higher basal level of transcription of bglG and the two steps involved in activation to give a Sal+ phenotype. To understand the functional significance of these observations, phenotypes of different S. sonnei strains were examined under defined genetic conditions.

To test the consequence of the higher basal level of bglG in wild-type S. sonnei strains, plasmid pMN5 (Mahadevan et al., 1987 ) carrying the bglF gene was introduced into the strain. Expression of bglF is directed from a vector promoter and is dependent on BglG, as the plasmid contains the Rho-independent terminator preceding the bglF gene. The presence of the plasmid in the strain AK1 conferred a weak Arb+ phenotype to the strain confirming the higher basal level of BglG in the strain leading to enhanced expression of the BglF permease in the cell (Table 2). Similarly, when plasmid p1H (Mahadevan et al., 1987 ) carrying the bglF and bglB genes was introduced, the strain showed a weak Sal+ phenotype. As in the case of pMN5, the bglF and bglB genes present on p1H require BglG for optimal expression. Under similar conditions, the wild-type E. coli strain RV remained Arb- Sal-, indicating that the levels of BglG are lower.


View this table:
[in this window]
[in a new window]
 
Table 2. Effect of plasmids carrying different bgl genes on the phenotypes of Shigella strains

 
To investigate the reasons for the Sal- phenotype of strain AK101, plasmid p6J (Mahadevan et al., 1987 ), carrying the bglB gene expressed from a vector promoter, was introduced into the strain. This resulted in a Sal+ phenotype, suggesting that the bglB locus in AK101 carries a mutation (Table 2). The emergence of Sal+ papillae is likely to be the result of reversion or suppression of this mutation subsequent to transcriptional activation of the promoter.

Analysis of the bglB locus in AK1, AK101 and AK102
The genetic analysis described above suggests that the bglB locus in AK101 carries a mutation and appearance of Sal+ papillae is the result of reversion or suppression of this mutation. This was verified by carrying out Southern analysis of S. sonnei genomic DNA using the E. coli bglB gene as a probe. The results, presented in Fig. 4(b), show that the bglB locus in AK1 and AK101 carries an additional DNA sequence compared to E. coli. The size of this insertion is approximately 1·5 kb and is located at the 5' end of the bglB locus, as evident from the Southern analysis using different restriction enzymes, sites of which are conserved in S. sonnei. Interestingly, the Sal+ strain AK102 shows the same pattern as the wild-type and Arb+ strains, indicating the retention of the insertion. The strain has apparently regained function of the bglB gene as a result of either intragenic or extragenic suppression of the insertion mutation.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 4. Comparison of the bglB region of AK1, AK101 and AK102 by Southern analysis. (a) Restriction map of the E. coli bglB region. The positions of EcoRV and HindIII sites are indicated by vertical arrows. The inverted triangle indicates the position of the insertion element within bglB. The 3·6 kb probe used in the Southern analysis is shown as a white box. (b) Southern analysis of the bglB region of S. sonnei with the 3·6 kb probe. Lanes 1 and 4, S. sonnei strain AK1 (WT, Arb- Sal-); lanes 2 and 5, S. sonnei strain AK101 (Arb+ Sal-); lanes 3 and 6, S. sonnei strain AK102 (Arb+ Sal+); lane 7, E. coli strain RV (WT); lane 8, E. coli strain AE328 (Bgl+); lane 9, E. coli strain JF201 ({Delta}bgl).

 
Our efforts towards mapping the reversion as intragenic or extragenic, by P1 transduction or by cloning the PCR product carrying the intact insertion, along with flanking bglB sequences from the wild-type and mutant strains of S. sonnei, have been unsuccessful. This is likely to be related to the instability of the clones that carry the intact DNA element. However, cloning of parts of the insertion along with flanking bglB sequences has been possible. Southern analysis of S. sonnei genomic DNA using the insertion DNA as a probe has shown the presence of multiple copies of the sequence in the genome (A. S. Kharat & S. Mahadevan, unpublished), suggesting that the insertion is a transposable element.

To identify the point of insertion, PCR amplification of the insertion was followed by cycle sequencing of the 5' end using primers designed based on the flanking bglB sequences. The point of insertion could be identified as position 3761 (coordinate based on data from Schnetz et al., 1987 ) in all S. sonnei strains examined. To test whether the element is similar to any of the known insertion sequences in the size range of 1·3–1·5 kb, the PCR product carrying the element was subjected to digestion using different restriction enzymes. These analyses suggest that the element is likely to be a novel one. Nucleotide sequence and additional characterization of the element will be published elsewhere.

Many insertion elements inactivating gene function are known to cause transcriptional polarity. To address this possibility in the case of transcription of the bglB locus in S. sonnei, in vivo transcriptional analysis of the bglB gene from the wild-type and salicin-utilizing strains of S. sonnei was carried out by Northern blot analysis. Blots of total RNA isolated under inducing conditions from strain JF201 ({Delta}bgl) of E. coli and strains AK1 (Arb- Sal-), AK101 (Arb+ Sal-) and AK102 (Arb+ Sal+) of S. sonnei were probed with a bglB-specific probe. To normalize the levels of transcription, the blot was stripped and reprobed using a probe corresponding to the E. coli ung gene as an internal control (Varshney et al., 1988 ). The transcripts were normalized by taking the ratio of counts corresponding to the bglB signal to that of ung. These results are represented in Fig. 5. There was no appreciable difference in the levels or size of transcript between strains AK101 (Sal-) and AK102 (Sal+), indicating that the insertion within bglB is not polar and reversion to the Sal+ state is not by relief of transcriptional polarity. However, these results do not rule out the possibility that the Sal+ state of AK102 is derived by the activation of another cryptic ß-glucosidase located elsewhere in the genome.



View larger version (49K):
[in this window]
[in a new window]
 
Fig. 5. Northern analysis of transcription of the bglB locus in S. sonnei. Total RNA was isolated from the strains indicated under inducing conditions and subjected to Northern analysis. The probes used for hybridization are indicated on the right side of each panel. Lane 1, JF201 ({Delta}bgl strain of E. coli); lane 2, AK1 (WT, Arb- Sal- strain of S. sonnei); lane 3, AK101 (Arb+ Sal-, derived from AK1); lane 4, AK102 (Arb+ Sal+, derived from AK101).

 
Measurement of the phospho-ß-glucosidase B activity
To confirm the Sal+ phenotype seen in the case of the strain AK102 at the level of bglB expression, measurements of phospho-ß-glucosidase B activity were carried out in the strain. The results obtained are presented in Table 3. The fully induced activity seen in strain AK102 is comparable to that of E. coli strain AE328 that carries an activated copy of the bgl operon. Interestingly, the basal level of expression seen in AK102 is fourfold higher than that of AE328, a likely consequence of the mutation in the Rho-independent terminator in the strain. These results indicate that strain AK102 can cleave salicin and the enzyme is expressed semiconstitutively.


View this table:
[in this window]
[in a new window]
 
Table 3. Measurement of phospho-ß-glucosidase B activity in different strains

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The experiments described above were undertaken to address the differences in the pattern of activation of the bgl genes in E. coli and S. sonnei, two closely related organisms, with the objective of understanding the possible evolutionary reasons for the crypticity of the bgl genes. The results presented above highlight three major differences in the organization and function of the bgl genes in S. sonnei compared to E. coli. i) The bglB gene encoding the phospho-ß-glucosidase B enzyme is insertionally inactivated in S. sonnei. ii) Despite the loss of function of the major structural gene involved in ß-glucoside utilization, the basal level of transcription of the operon is enhanced. iii) Although the promoter structure and organization are conserved, the spectrum of activating mutations obtained under similar conditions is different.

The most striking difference in the expression of the bgl genes in S. sonnei is the higher basal level of transcription of bglG. Interestingly, the higher basal level of expression is not achieved by the activation of the cryptic promoter, but by a mutation in a downstream regulatory element. However, this increase is not sufficient to give an Arb+ phenotype to the cell. Therefore it is unlikely that the mutation within the regulatory site arose as a direct result of selection for arbutin utilization. The presence of the mutation also makes expression of the operon constitutive once the promoter is activated, which is seen at the level of transcription and enzyme activity.

A majority of the activating mutations (87%), isolated under nutritionally rich conditions, mapped within the hns locus in the case of S. sonnei. The complementation analysis using a multicopy plasmid carrying the wild-type copy of the hns gene does not rule out the possibility that the mutations map not within the hns gene, but at loci under the control of hns. Since H-NS is known to regulate bgl expression directly, this possibility is unlikely. Activating mutations isolated under identical conditions in E. coli map within the bglR locus and consist predominantly (98%) of insertions. Interestingly, late-arising Arb+ mutants of E. coli, isolated after prolonged incubation on minimal arbutin plates, showed a higher incidence (21%) of lesions in the hns gene (Hall, 1998 ). The basis for this difference between the two strains is not known at this stage. One possibility is that hns mutations may occur at the same frequency in E. coli, but are not selected because of their weaker Bgl phenotype. Since the basal level of expression of the operon is higher in the S. sonnei strains as a result of the regulatory-site mutation, even weaker hns mutations may be able to confer a stronger Bgl phenotype in these strains. This is likely to have an impact on the expression of the bgl operon in the natural environments of the two organisms.

Activation of the operon by two steps, involving reversion of an insertion mutation in bglF, has been reported in the case of a specific strain of E. coli (Parker et al., 1988 ). The difference between this report and our findings is that the insertional inactivation of bglB is not confined to a single strain, but is a property of all S. sonnei strains examined. Partial characterization of the insertion within the bglB locus has indicated that the point of insertion is identical in all strains examined, suggesting that the insertional inactivation of the bglB locus occurred early in the evolutionary history of S. sonnei.

It is interesting that the insertion element is retained in the Sal+ revertants. Transcriptional analysis of the bglB locus reported here shows that bglB transcript is present even in the Sal- strains. Therefore, the reversion is likely to be at a step involving the translation of the bglB message. Efforts to identify the nature of reversion have been hampered by our inability to clone the insertion element in its entirety. Genetic approaches such as P1 transduction, to map the location of the reversion, have also been unsuccessful. However, cloning of partial sequences of the element has been possible. A comparison of the nucleotide sequences of the bglB locus in the mutant and wild-type strains is likely to provide information regarding the nature of the reversion. Alternatively, the mutation in AK102 may lie in an unlinked locus and suppress the effect of the insertion. The results of the transcriptional analysis also do not exclude the possibility that the Sal+ status of AK102 is the result of activation of another gene encoding a cryptic phospho-ß-glucosidase specific for salicin.

One possible explanation for the structural gene inactivation is that S. sonnei represents a state where the bgl operon is in the process of becoming a pseudogene by accumulating mutations. This conclusion has two major limitations. Although the bglB gene has accumulated a mutation, the bglG and bglF genes are intact in the strain, conferring on it the ability to mutate to Arb+. Secondly, if the operon is in the process of being converted to a pseudogene, it is difficult to explain the mutation in the regulatory site leading to an enhanced basal level of transcription. In the absence of the structural gene function, any selection for an enhanced basal level is unlikely. It is also unlikely that the mutation is selectively neutral since it is conserved in all strains of Shigella including Shigella dysenteriae, Shigella flexneri and Shigella boydii (A. S. Kharat & S. Mahadevan, unpublished).

This seemingly paradoxical situation can be addressed by considering the possibility that the regulators of the bgl operon, bglG and bglF, have additional roles to play in regulating cellular functions not directly related to ß-glucoside utilization (Mahadevan, 1997 ). The bglG–bglF combine represents an efficient regulatory system in which the phosphotransferase component encoded by bglF acts as a sensor and the bglG gene product acts as a response regulator by mediating antitermination of transcription in response to the availability of ß-glucosides. Similar systems that share considerable functional and sequence similarities have been reported in different organisms. The situation in the case of Bacillus subtilis is particularly illuminative because of the presence of several paralogues of the BglG regulator (Stulke et al. 1998 ). It has been shown that the LicT regulator, a homologue of the BglG antiterminator, regulates at least two relatively unrelated functions in B. subtilis (Schnetz et al., 1996 ). Therefore it is conceivable that the bgl genes of E. coli perform regulatory functions elsewhere within the cell and are conserved for this reason. The basal level of the regulators may be sufficient for this putative function. Another possibility is that the mutation in S. sonnei that results in an increased basal level was selected under specific conditions that demanded higher levels of the bgl regulators. This is consistent with the results of a recent study demonstrating the in vivo expression of the wild-type bgl operon in a pathogenic strain of E. coli recovered from infected mouse liver (Khan & Isaacson, 1998 ). Such a possibility is also consistent with our results that the bglR and bglG loci are conserved in pathogenic strains of Shigella and Salmonella although these strains are incapable of mutating to Bgl+ (A. S. Kharat & S. Mahadevan, unpublished). However, until the putative targets for the regulators are identified, the possibility for alternative roles for the bgl genes, such as in pathogenicity, remains speculative.

In summary, our studies have helped to highlight the high degree of conservation of the ‘cryptic’ bgl genes in the two closely related micro-organisms, with interesting differences in terms of their organization and function. These results, along with a number of recent observations by different investigators (Keyhani & Roseman, 1997 ; Khan & Isaacson, 1998 ), prompt a re-examination of the concept of crypticity of genes in general.


   ACKNOWLEDGEMENTS
 
We thank B. E. Uhlin, C. Ueguchi, J. Lopilato and U. Varshney for the gifts of hns, leuO, bglJ and ung constructs, respectively, and R. Roy and V. Ramchandran for the gift of clinical isolates of S. sonnei. We wish to thank members of the laboratory for fruitful discussions. A.S.K. acknowledges the help of N. R. Kulkarni in the isolation and identification of the strain AK1. This work was supported by a grant to S.M. from the Council for Scientific and Industrial Research, Government of India. A programme support to the Indian Institute of Science from the Department of Biotechnology is also acknowledged.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Amster-Choder, O. & Wright, A. (1993). Transcriptional regulation of the bgl operon of Escherichia coli involves phosphotransferase system-mediated phosphorylation of a transcriptional antiterminator.J Cell Biochem 51, 83-90.[Medline]

Caramel, A. & Schentz, K. (1998). Lac and lambda repressors relieve silencing of the Escherichia coli bgl promoter: activation by alteration of a repressing nucleoprotein complex.J Mol Biol 284, 875-883.[Medline]

Defez, R. & DeFelice, M. (1981). Cryptic operon for ß-glucoside metabolism in Escherichia coli K12: genetic evidence for a regulatory protein.Genetics 97, 11-25.[Abstract/Free Full Text]

DiNardo, S., Voelkel, K. A., Sternglanz, R., Reynolds, A. E. & Wright, A. (1982). Escherichia coli DNA topoisomerase I mutants have compensatory mutations in DNA gyrase genes.Cell 31, 43-51.[Medline]

Falkow, S. & Mekalanos, J. (1990). The enteric bacilli and vibrios. In Microbiology, pp. 567-587. Edited by B. D. Davis, R. Dulbecco, H. N. Eisen & H. S. Ginsberg. Philadelphia: Lippincott.

Fox, F. C. & Wilson, G. (1967). The role of a phospho-enol-pyruvate-dependent kinase system in ß-glucoside catabolism in Escherichia coli.Proc Natl Acad Sci USA 59, 988-995.

Giel, M., Desnoyer, M. & Lopilato, J. (1996). A mutation in a new gene, bglJ, activates the bgl operon in Escherichia coli K12.Genetics 143, 627-635.[Abstract/Free Full Text]

Goransson, M., Sonden, B., Nilsson, P., Dagberg, B., Forsman, K., Emanuelsson, K. & Uhlin, B. E. (1990). Transcriptional silencing and thermoregulation of gene expression in Escherichia coli.Nature 344, 682-685.[Medline]

Hall, B. G. (1998). Activation of the bgl operon by adaptive mutation.Mol Biol Evol 15, 1-5.[Abstract]

Hall, B. G. & Xu, L. (1992). Nucleotide sequence, function, activation and evolution of the cryptic asc operon of Escherichia coli K12.Mol Biol Evol 9, 688-706.[Abstract]

Keyhani, N. O. & Roseman, S. (1997). Wild type Escherichia coli grows on the chitin disaccharide N,N'-diacetylchitobiose by expressing the cel operon.Proc Natl Acad Sci USA 94, 14367-14371.[Abstract/Free Full Text]

Khan, N. O. & Isaacson, R. E. (1998). In vivo expression of the ß-glucoside (bgl) operon of Escherichia coli occurs in mouse liver.J Bacteriol 180, 4746-4749.[Abstract/Free Full Text]

Kharat, A. S. & Mahadevan, S. (1999). Plasmid-mediated suppression of the mutational activation of the bgl operon of Shigella sonnei.Acta Biochim Polon 46, 853-861.[Medline]

Kricker, M. & Hall, B. G. (1984). Directed evolution of cellobiose utilisation in Escherichia coli K-12. Mol Biol Evol 1, 171-182.[Abstract]

Kricker, M. & Hall, B. G. (1987). Biochemical genetics of the cryptic gene system for cellobiose utilisation in Escherichia coli K12.Genetics 115, 419-429.[Abstract/Free Full Text]

Lawrence, J. G., Ochaman, H. & Hartl, D. L. (1992). The evolution of insertion sequences within enteric bacteria.Genetics 131, 505-513.

Lopilato, J. & Wright, A. (1990). Mechanism of activation of the cryptic bgl operon of Escherichia coli K12. In The Bacterial Chromosome, pp. 435-444. Edited by C. Drlica & M. Riley. Washington, DC: American Society for Microbiology.

Mahadevan, S. (1997). The BglG group of antiterminators: a growing family of bacterial regulators. J Biosci 22, 505-513.

Mahadevan, S. & Wright, A. (1987). A bacterial gene involved in transcription antitermination: regulation at a Rho-independent terminator in the bgl operon of Escherichia coli.Cell 50, 485-494.[Medline]

Mahadevan, S., Reynolds, A. E. & Wright, A. (1987). Positive and negative regulation of the bgl operon in Escherichia coli.J Bacteriol 169, 2570-2578.[Medline]

Mukerji, M. & Mahadevan, S. (1997). Characterisation of the negative elements involved in silencing the bgl operon of Escherichia coli: possible roles for DNA gyrase, H-NS, and CRP-cAMP in regulation.Mol Microbiol 24, 617-627.[Medline]

Owen, R. J. & Borman, P. (1987). A rapid biochemical method for purifying high molecular weight bacterial chromosomal DNA for restriction enzyme analysis.Nucleic Acid Res 15, 3631.[Medline]

Parker, L. L. & Hall, B. G. (1988). A fourth Escherichia coli gene system with the potential to evolve ß-glucoside utilisation.Genetics 119, 485-490.[Abstract/Free Full Text]

Parker, L. L., Betts, P. W. & Hall, B. G. (1988). Activation of a cryptic gene by excision of DNA fragment.J Bacteriol 170, 218-222.[Medline]

Prasad, I. & Schaefler, S. (1974). Regulation of the ß-glucoside system in Escherichia coli K12.J Bacteriol 120, 638-650.[Medline]

Prasad, I., Young, B. & Schaefler, S. (1973). Genetic determination of the constitutive biosynthesis of phospho-ß-glucosidase A in Escherichia coli K12.J Bacteriol 114, 909-915.[Medline]

Reynolds, A. E., Felton, J. & Wright, A. (1981). Insertion of DNA activates the cryptic bgl operon in Escherichia coli K12. Nature 293, 625-629.[Medline]

Reynolds, A. E., Mahadevan, S., Le Grice, S. F. J. & Wright, A. (1986). Enhancement of bacterial gene expression by insertion elements or by mutation in a CAP-cAMP binding site.J Mol Biol 191, 85-95.[Medline]

Rutberg, B. (1997). Antitermination of transcription of catabolic operons.Mol Microbiol 23, 413-421.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schaefler, S. & Malamy, A. (1969). Taxonomic investigations on expressed and cryptic phospho-ß-glucosidases in Enterobacteriaceae.J Bacteriol 99, 422-433.[Medline]

Schnetz, K. (1995). Silencing of Escherichia coli bgl promoter by flanking sequence elements.EMBO J 14, 2545-2550.[Abstract]

Schnetz, K. & Rak, B. (1992). IS5: a mobile enhancer of transcription in Escherichia coli. Proc Natl Acad Sci USA 89, 1244-1248.[Abstract]

Schnetz, K., Toloczyki, C. & Rak, B. (1987). ß-Glucoside (bgl) operon of Escherichia coli K12: nucleotide sequence, genetic organisation and possible evolutionary relationship to regulatory components of two Bacillus subtilis genes.J Bacteriol 169, 2579-2590.[Medline]

Schnetz, K., Stulke, J., Gertz, S., Kruger, S., Krieg, M., Hecker, M. & Rak, B. (1996). LicT, a Bacillus subtilis transcriptional antiterminator protein of the BglG family. J Bacteriol 178, 1971-1978.[Abstract]

Singh, J., Mukerji, M. & Mahadevan, S. (1995). Transcriptional activation of the Escherichia coli bgl operon: negative regulation by DNA structural elements near the promoter.Mol Microbiol 17, 1085-1092.[Medline]

Stulke, J., Arnaud, M., Rapoport, G. & Martin-Verstraete, I. (1998). PRD - a protein domain involved in PTS-dependent induction and carbon catabolite repression of catabolic operons in bacteria. Mol Microbiol 28, 865-874.[Medline]

Ueguchi, C., Ohta, T., Set, C., Suzuki, T. & Mizuno, T. (1998). The leuO gene product has a latent ability to relieve bgl silencing in Escherichia coli.J Bacteriol 180, 190-193.[Abstract/Free Full Text]

Varshney, U., Hutcheon, T. & Van de Sande, J. H. (1988). Sequence analysis, expression, and conservation of Escherichia coli uracil DNA glycosylase and its gene (ung).J Biol Chem 263, 7776-7784.[Abstract/Free Full Text]

Watanabe, H. & Okamura, N. (1991). The genus Shigella. In The Prokaryotes, 2nd edn, vol. III, A handbook on the biology of bacteria: ecophysiology, isolation, identification and applications, pp. 2754–2759. Edited by A. Balows & others. New York: Springer-Verlag.

Woodcock, D. M., Crowther, P. J., Doherty, J., Jefferson, S., DeCruz, E., Noyer-Weidner, M., Smith, S. S., Michael, M. Z. & Graham, M. W. (1989). Quantitative evaluation of Escherichia coli host strains for tolerance to cytosine methylation in plasmid and phage recombinants. Nucleic Acid Res 17, 3469-3478.[Abstract]

Received 21 December 1999; revised 28 March 2000; accepted 11 April 2000.