INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Bacteroides thetaiotaomicron and other human colonic Bacteroides species utilize a variety of polysaccharides as a source of carbon and energy (15). This trait may be important for their survival in the human colon because polysaccharides are the main form of carbohydrate available to colon bacteria. The starch utilization system of B. thetaiotaomicron is the best studied of the Bacteroides polysaccharide utilization systems. Previously, a cluster of starch utilization genes, designated sus genes, was identified and characterized (4, 5, 13, 14). This cluster contains seven structural genes (susA to susG), most of which encode proteins that mediate the initial steps in starch utilization, such as starch binding and starch hydrolysis (13, 14). These genes are organized into two transcriptional units (5, 14), one containing susA and one containing susB to susG.

Expression of the structural genes is regulated at the transcriptional level by maltose and higher oligomers of starch. Regulation is mediated by SusR, a protein encoded by a gene that is located upstream of susA. Unlike the structural genes, susR is constitutively expressed (5). Since multiple copies of susR in trans increased sus gene expression and since a disruption in susR abolished expression of susA-susG, it appeared that SusR alone was responsible for controlling expression of the structural genes. In this report, we show that there is at least one other regulatory gene that participates in control of sus gene expression.

No other starch utilization genes were found in the region of the sus gene cluster, but it was clear that B. thetaiotaomicron must have other starch utilization genes. For one thing, disruption of susB, which encodes an -glucosidase, did not eliminate all of the -glucosidase activity in cell extracts. For another, the susR disruption mutant still grew as well as wild type on maltose and maltotriose even though it could not grow on higher oligomers. Thus, other maltose utilization genes must be located elsewhere on the chromosome. We report here that although a search for mutants unable to grow on maltose and maltotriose failed to locate the second -glucosidase gene or any other genes encoding maltose utilization proteins, it did unexpectedly yield a second regulatory gene, malR, that controls expression of the sus genes, as well as expression of the second -glucosidase.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and growth conditions. The bacterial strains and plasmids used in this study are listed in Table 1. All Escherichia coli strains used in this study were grown in Luria-Bertani (LB) broth or on LB agar at 37°C. B. thetaiotaomicron 5482, transposon-generated derivatives, and some singly or doubly disrupted mutants used in this study have been described previously (1, 4, 14).

DNA methods. Isolation of plasmids was done by using a Wizard Plus DNA purification system (Promega Corp.). Dephosphorylation reactions and restriction digests were performed in accordance with the manufacturer's instructions (Bethesda Research Laboratories [Bethesda, Md.] or New England BioLabs [Beverly, Mass.]). Transformation of E. coli DH5MCR was done by the method of Lederberg and Cohen (10). Conjugations, where constructs generated in E. coli were transferred to Bacteroides recipients, were performed as described by Shoemaker et al. (17). Insertional and replicative shuttle vectors were mobilized from E. coli donors to Bacteroides recipients by transfer genes of RP4 integrated in the chromosome of S17-1 (19). Southern blotting was done as described by Maniatis et al. (11) except that a Renaissance Detection Kit (DuPont-NEN) was used for detection of the bound DNA probe.

Chemicals. Glucose, maltose, maltotriose, amylopectin, dextran, phenylmethylsulfonyl fluoride, p-nitrophenyl--D-glucopyranoside, were purchased from Sigma Corp. 4-Nitrophenyl--D-maltoheptaoside-4,6-O-ethylidene was purchased from Boehringer Mannheim Biochemicals.

Isolation of a B. thetaiotaomicron mutant deficient in maltose and maltotriose utilization. To isolate a mutant of B. thetaiotaomicron that was deficient in maltose and maltotriose utilization, transposon mutagenesis was done by introducing the Bacteroides transposon Tn4351 into two different hosts: B. thetaiotaomicron 5482 (wild type) and the susR disruption mutant B. thetaiotaomicron susR (BTsusR). In the mutant strain, BTsusR, the susR gene had been disrupted by a single crossover insertion of the suicide vector, pBT-1, into which an internal segment of susR had been cloned (5). The selectable marker on pBT-1 was a tetracycline resistance gene, tetQ, so that the selectable marker on Tn4351, the erythromycin resistance gene ermF, could be used. Tn4351 was introduced into the wild-type strain or the BTsusR mutant by conjugation, by using as a donor E. coli HB101 containing Tn4351 on the self-transmissible IncP plasmid, R751 (2, 12). Transconjugants harboring Tn4351 insertions were selected by growth on TYG agar plates containing erythromycin (10 µg/ml) and gentamicin (200 µg/ml). The gentamicin selection eliminated the E. coli donors. Tetracycline (3 µg/ml) was also included in the medium to ensure retention of the pBT-1 insertion in BTsusR. Transconjugants were screened for growth on maltose-defined medium agar plates.

Cloning of DNA adjacent to the Tn4351 insertion in B. thetaiotaomicron MAL (BTMAL). The strategy used to clone DNA adjacent to the Tn4351 insertion is shown in Fig. 1. Chromosomal DNA was isolated from the mutant and digested with EcoRI. EcoRI cuts Tn4351 near one end of each of the directly repeated insertion sequence (IS) elements that flank Tn4351 but nowhere else in the transposon. The EcoRI fragments were ligated into the EcoRI site of pUC19. E. coli DH5 MCR transformants were plated onto Luria agar containing ampicillin (100 µg/ml) and X-Gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside; 40 µg/ml). White colonies on X-Gal plates were screened for the Tn4351 junction fragment by colony hybridization, by using a probe that contained IS element DNA. To clone the other chromosomal junction, we first recloned the EcoRI fragment containing IS DNA into pGERM. This clone was transferred to wild-type B. thetaiotaomicron to create a single crossover insertion in the cloned region. Chromosomal DNA from the resulting strain was digested with PstI, religated, and then transformed into E. coli with selection for ampicillin. PstI cuts only once in pGERM and not at all in the IS element. Thus, ampicillin-resistant transformants should contain chromosomal DNA adjacent to the insertion site on both sides of the transposon insertion.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Cloning of pMAL-Right and pMAL-Left to sequence the region where Tn4351 is inserted in BTMAL. The position of the transposon insertion in BTMAL is indicated by the black rectangle. The heavy horizontal arrows in Tn4351 indicate the direct repeat insertion sequence (IS4351). Tn4351 carries an erythromycin resistance gene (ermF), which is expressed only in Bacteroides strains, and a tetracycline resistance gene (tetX), which is expressed only in aerobically grown E. coli strains. pGERM, a suicide vector in Bacteroides, has a different erythromycin resistance gene (ermG). The double lines (=) indicate Bacteroides chromosomal DNA. Abbreviations: RI, EcoRI; P, PstI

Enzyme assays. -Glucosidase activity and amylase activity in sonically disrupted cell extracts were measured by determining the rate of hydrolysis of p-nitrophenyl--D-glucopyranoside and 4-nitrophenyl--D-maltoheptaoside-4,6-O-ethylidene, respectively. -Glucosidase activity was measured as described by Smith and Salyers (20). Amylase activity was measured with 2 mM 4-nitrophenyl--D-maltoheptaoside-4,6-O-ethylidene in potassium phosphate buffer (20 mM, pH 6.5) at 37°C (Boehringer Mannheim Biochemicals). The protein concentration in each extract was measured by using the Bio-Rad DC protein kit. The K_m values of the -glucosidases were calculated from a Lineweaver-Burke plot. BTsusB was used to determine the K_m of the second -glucosidase, and BTmalR was used to determine the K_m of the susB-encoded -glucosidase. -Glucuronidase (GUS) assays, used to monitor expression of uidA (GUS) fusions, were done as described by Feldhaus et al. (6).

malR gene expression. To provide malR in trans on a multicopy plasmid, malR was amplified by PCR with primers TCAAAGTACTGGATCCCGAAATGACC, which lies ca. 300 bp upstream of the first start codon in the ORF, and TATATTGACAGGATCCATGTACTTGT, which lies ca. 150 bp downstream of the first stop codon in the ORF. This product (ca. 1.2 kb) was cloned into pT-COW, which has a copy number in Bacteroides of 5 to 10 (22). The resulting vector was called pMALR.

Nucleotide sequence accession number. The nucleotide sequence of malR region (Fig. 2A) has been deposited in GenBank under accession number AF391102.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   (A) Map showing the relative locations of ORFs in the malR gene area. The position of the transposon insertion in mutant BTMAL is indicated by the vertical arrow above the map. DNA segments used to make insertional disruptions are shown as horizontal lines under the map and marked with an "." The sizes of the DNA fragments used to make the disruptions are also indicated under the lines. gs, putative glutamine synthetase gene; malR, putative regulatory protein gene; -nagA, putative -N-acetylglucosaminidase gene. (B) Deduced amino acid sequence of malR. A possible carboxy-terminal helix-turn-helix motif is underlined. The vertical arrow indicates the transposon insertion site in the transposon-generated mutant, BTMAL.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of a mutant with reduced ability to grow on maltose and maltotriose. A screen of thousands of transposon-generated mutants of wild-type B. thetaiotaomicron was done to find mutants that had lost the ability to utilize maltose or maltotriose. No such mutants were found. A possible explanation for the failure to find such mutants was that the sus genes were contributing to the utilization of maltose and maltotriose, and this redundancy with the other maltose or maltotriose utilization genes made it impossible for a single transposon insertion to abolish maltose utilization. Accordingly, a mutant with a disruption in susR (BTsusR) was used as the background for transposon mutagenesis. This mutant did not produce any of the Sus structural proteins. Over 15,000 transposon-generated mutants were screened for maltose and maltotriose utilization. One mutant was found that was deficient in the ability to grow on maltose and maltotriose. The mutant was named BTMAL. The growth rate of BTMAL on glucose (0.45 h¹) was the same as that of the parent strain, BTsusR, but BTMAL had one-fourth the growth rate of the parent strain on maltose (0.11 h¹) and did not grow at all on maltotriose.

View larger version (49K): [in this window] [in a new window]   FIG. 3.   Growth rates of various mutants on glucose, maltose, and maltotriose. The medium used was defined medium that contained glucose, maltose, and maltotriose, and the results are indicated by the three shaded bars in each set from left to right, respectively. The concentrations of antibiotics used in this experiment were 3 µg/ml for tetracycline (Tc) and 10 µg/ml for erythromycin (Em). These measurements were done in triplicate; the range of values is indicated by the error bars. B. thetaiotaomicron BT4009 was used as the wild-type control because it contains a single copy of tetQ and ermF. Thus, either tetracycline or erythromycin or both (Tc, Em) can be added to the medium used to grow both the control and the mutant strains. This eliminates the slight differences in growth rate that can sometimes occur due to the presence of antibiotics in the medium (note the difference between "Tc" columns versus the "Tc, Em" columns). The selectable marker used to create the BTmalR strain was ermF, and the marker used to create the BTsusR strain was tetQ. The double mutant BTsusRmalR contained both resistance genes.

The malR gene controls expression of the second -glucosidase. -Glucosidase and amylase activities in the various mutant strains were measured. Amylase activity was used as an indicator of the sus gene expression because the only amylases produced by B. thetaiotaomicron are encoded by susA and susG (16). Thus, the activity of these proteins is a good indicator of the activity of proteins encoded in this region. As expected, BTsusR had no detectable amylase activity and had lower -glucosidase activity than the wild type (Table 2). The -glucosidase activity remaining in extracts from the BTsusR mutant did not come from SusB; when we disrupted susB, the level of -glucosidase activity was similar to that of BTsusR. When malR in BTsusR was disrupted to create BTsusRmalR, virtually all of the -glucosidase activity disappeared. This was also true of BTsusBmalR. Hence, malR controls expression of the second -glucosidase.

View this table: [in this window] [in a new window]   TABLE 2.   -Glucosidase and amylase activities of whole-cell extracts of strainsa

Comparison of properties of the two -glucosidases. Cell extracts from BTmalR and BTsusB were used to assay SusB and the second -glucosidase, respectively. The K_m values of the two -glucosidases were similar: 132 µM for SusB and 103 µM for the second -glucosidase. Also, the cellular location of these two -glucosidase were the same. SusB partitioned to the inner membrane fraction but could be eluted by washing the membrane with 0.5 M NaCl (4). This was also the case for the second -glucosidase; about two-thirds of the -glucosidase activity partitioned with the membrane fraction. There was, however, a difference in the stability of the two enzymes in cell extracts. The second -glucosidase was not stable even at 4°C. The enzyme activity that was detectable in a cell extract from BTsusB after 14 h at 4°C was just 33% of the initial activity. In contrast, 87% of the original activity was detectable in an extract from BTmalR that had been stored for 14 h at 4°C.

malR regulates expression of sus genes. If malR regulates only the expression of the second -glucosidase, BTmalR should still have full SusB and amylase (SusA and SusG) activity. That is, the -glucosidase activity of the malR disruption strain (BTmalR) should be ca. 53 U/g of cell protein, the value of the activity in BT5482 extracts minus the activity in BTsusR extracts, and the amylase activity should be 49 to 50 U/g of cell protein. However, the -glucosidase activity in BTmalR extracts was only 14 U/g cell protein, one-fourth of the expected value (Table 2). Moreover, the amylase activity in BTmalR extracts was much lower than in extracts from wild type. Thus, malR seems to be necessary for full expression of the sus genes. To make sure that the -glucosidase of BTmalR was due to SusB, we disrupted susB in BTmalR to create BTsusBmalR. BTsusBmalR has no detectable -glucosidase activity. Thus, the -glucosidase activity in extracts from BTmalR came from SusB. The -glucosidase specific activity in BTsusR extracts was almost the same as in BTsusB extracts, so SusR does not control the expression of the -glucosidase controlled by MalR.

1

1

1

Effect of providing malR in trans. The data from disruption mutants indicated that malR had its own promoter, because disruptions in the adjacent ORFs had no effect on maltose utilization. Accordingly, malR plus ca. 300 bp of upstream DNA was cloned to produce pMALR and was introduced into wild type and BTmalR. The plasmid in which malR was cloned has a copy number of ca. 5 to 10 copies per cell (22). When pMALR was present in BTmalR, -glucosidase and amylase activities were twofold higher than those of BTmalR (Table 3), so malR in trans complemented the disrupted chromosomal malR and increased expression of starch utilization genes. This same effect was seen when pMalR was introduced into the wild-type strain (Table 3). These results showed that MalR is not an -glucosidase but is a regulatory protein that regulates positively genes encoding -glucosidase, amylase genes (susA and susG) and presumably susC to susF as well. The growth rate of BT5482(pMALR) on amylopectin was 0.61 h¹, a value slightly higher than that of a control strain, BT5482(pT-COW), which contained only the vector into which malR was cloned (0.52 h¹).

View this table: [in this window] [in a new window]   TABLE 3.   -Glucosidase and amylase activities of whole-cell extracts of strainsa

malR is expressed constitutively, and its expression is not autoregulated. RT-PCR was used to determine whether malR is expressed constitutively or is induced by maltose. Expression of susD was used as a control. RT-PCR detected susD mRNA only when cells were grown on maltose (Fig. 4). In contrast, malR mRNA was detected in cells grown on glucose as well as in cells grown on maltose (Fig. 4). Thus, malR is expressed constitutively.

View larger version (54K): [in this window] [in a new window]   FIG. 4.   Detection of malR expression on glucose and maltose by using RT-PCR. Odd-numbered lanes contain negative controls in which the reaction was done without the RT step. The expression of the maltose-regulated susD gene was used as a control to assess whether we could detect regulated gene expression (lane 2, glucose [G]-grown cells; lane 4, maltose [M]-grown cells). The region amplified from susD was 0.5 kb in size. To check malR expression (lanes 6 and 8), a 0.4-kbp region of the mRNA was amplified. The 1-kb ladder (Gibco-BRL) is seen in the two outside lanes.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Our results suggest that the genes controlled by susR and malR encode most or all of the proteins responsible for utilization of maltotriose. Disrupting both of these genes abolished growth on maltotriose and severely reduced growth on maltose. Proteins encoded by the sus genes aid in maltotriose utilization but do not contribute significantly to utilization of maltose because disruption of susR decreases growth on maltotriose but not growth on maltose. In contrast, disruption of malR decreases the rate of growth on both maltose and maltotriose. The malR gene was not linked genetically to the gene encoding the second -glucosidase or to any other structural genes that may be under MalR control, so there is as yet little information about the genes MalR controls other than the sus genes. An attempt to purify the second -glucosidase failed due to the instability of the enzyme (data not shown), so it was not possible to use N-terminal sequencing as a basis for designing a probe to locate the gene encoding it. The K_m for this enzyme was virtually identical to that of SusB. In fact, the only difference between the two enzymes was stability in cell extracts. This raises the question of why there are two -glucosidases, which appear to have redundant characteristics, and why one is associated with the sus system and one is not.

MalR appears to be a regulatory protein. Not only does it have amino acid similarity to known regulatory proteins, but its loss affects the activities of more than one protein. This latter observation supports the hypothesis that MalR controls the transcription of the genes encoding these proteins and probably other genes as well. The fact that the transcription of susB is affected by a malR disruption indicates that transcription of the other genes in the susB-susG operon would also be affected. An unexpected role of malR is its effect on sus gene expression. The sus genes were thought to be controlled only by susR, but results reported here show that disruption of malR reduced expression of the sus structural genes by 5- to 10-fold. Loss of SusR, however, did not seem to affect the expression of the -glucosidase controlled by MalR. Since both susR and malR are expressed constitutively, it is likely that SusR and Mal R proteins interact with each other rather than one controlling the other's expression.

An alternative possibility is that the second -glucosidase or one of the other proteins encoded by MalR-controlled genes is not involved in catabolism of maltose but rather converts maltose to a derivative that is a more effective inducer than maltose itself. In this case, the drop in the expression of susA and susB in the malR disruption strain would be due to the fact that maltose is now the only inducer available to interact with SusR. This would explain why MalR seems to influence expression of the sus genes but SusR does not seem to influence expression of the MalR-controlled -glucosidase. If SusR and MalR proteins formed a complex that increases transcription of the sus genes, one would expect this same complex to be responsible for expression of the MalR-controlled genes. The hypothesis that a MalR-controlled gene encodes an enzyme that makes a better inducer would also explain why expression of the sus genes is decreased in the malR disruption strain but does not fall to zero.

The results reported here suggest that the MalR regulon and the SusR regulon together account for all of the genes needed to grow on maltotriose and starch. The very low rate of growth of the BTsusBmalR strain on maltose could be due to uptake and breakdown of maltose by transporters and enzymes that normally handle a closely related substrate such as melibiose or lactose. Given that the mal and sus systems account for most or all of the utilization of small glucosides, the finding that the BTmalR strain still grew on starch confirms that the susA-susG gene products are solely responsible for the processing of starch. The fact that the malR disruption strain was able to grow on starch nearly as well as wild type was surprising since expression of the sus genes was decreased five- to ninefold in that strain. This observation can be explained by noting that B. thetaiotaomicron probably never encounters in nature concentrations of starch as high as those it encounters in laboratory medium. The bacteria may be optimized to operate with enzyme levels lower than those seen in bacteria growing on high concentrations of maltose. Whatever the explanation for the effect of the disruption in malR on sus gene expression, it is clear that the starch utilization pathway and the pathway controlled by MalR are linked at the metabolic and/or regulatory level.