The 1-kb-repeat-encoded DNA-binding protein as repressor of an {alpha}-glucosidase operon flanking the amplifiable sequence AUD1 of Streptomyces lividans

Jean-Nicolas Volffa,1 and Josef Altenbuchner1

Institut für Industrielle Genetik, Universität Stuttgart, Allmandring 31, 70569 Stuttgart, Germany1

Author for correspondence: Josef Altenbuchner. Tel: +49 711 685 7591. Fax: +49 711 685 6973. e-mail: Josef.Altenbuchner{at}po.uni-stuttgart.de


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
High-copy-number amplification of the AUD1 element is frequently associated with the large chromosomal deletions responsible for genetic instability in Streptomyces lividans TK64. Five ORFs were found in a 7 kb region directly adjacent to AUD1. The putative products of ORF1, ORF2 and ORF3 showed similarities to ATP-binding cassette (ABC) sugar transporters, the deduced protein sequence of ORF4 displayed similarities to {alpha}-glucosidases whilst no homology to proteins with known functions was found for ORF5. ORF4 (renamed aglA) was expressed in Escherichia coli and the protein purified and characterized. An {alpha}-glucosidase activity was detected using the synthetic {alpha}-glucoside p-nitrophenyl {alpha}-D-glucopyranoside. Of the many oligosaccharides tested, only sucrose was hydrolysed at a measurable rate [specific activity 32·4 units (mg protein)-1] but no growth of S. lividans TK64 on sucrose was observed. A strain in which aglA was disrupted showed the same low {alpha}-glucosidase activity as strain TK64 and in both strains no stimulation of activity was seen by sucrose, trehalose or maltose; dextrin increased {alpha}-glucosidase activity about 10-fold. This probably resulted from induction of a second {alpha}-glucosidase-encoding gene. The AUD1 element contains three 1 kb repeats which encode DNA-binding proteins necessary for high-frequency amplification. In strains with a unique 1 kb repeat, disruption of the repeat led to a significant increase in the {alpha}-glucosidase activity. These results strongly suggest that the 1-kb-repeat-encoded proteins of AUD1 have a dual function: they are the repressors of the agl genes and they promote amplification of AUD1.

Keywords: Streptomyces, {alpha}-glucosidase, ABC transporter, repressor, amplification

Abbreviations: ABC, ATP-binding cassette; AUD, amplifiable unit of DNA; LDR, left direct repeat; LIP, left imperfect palindrome; MDR, middle direct repeat; PBS, potential binding sequence; p-NPG, p-nitrophenyl {alpha}-D-glucopyranoside; RDR, right direct repeat; RIP, right imperfect palindrome

The GenBank accession number for the sequence reported in this paper is U22894.

a Present address: Biozentrum/Physiologische Chemie I, Am Hubland, 97074 Würzburg, Germany.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
High-copy-number amplification of the chromosomally located AUD1 (amplifiable unit of DNA no. 1) is frequently observed in spontaneous mutants generated by genetic instability in Streptomyces lividans (Altenbuchner & Cullum, 1984 , 1985 ). The mutants are sensitive to chloramphenicol (CmlS) and auxotrophic for arginine (Arg-). This double mutant phenotype is due to the deletion of two chromosomal markers included in a large deletion bordering the amplified DNA sequence (ADS).

AUD1 is composed of three 1 kb and two 4·7 kb repeats alternately arranged in direct orientation (Fig. 1; Altenbuchner & Cullum, 1985 ). The 4·7 kb repeats are identical and encode putative proteins having some similarities to chitinases (Volff et al., 1996 ). They serve as substrate for the recA-dependent recombination event leading to amplification. When they are replaced by direct repeats generated from Escherichia coli DNA the modified AUD1 structure is still able to amplify (Volff et al., 1996 ). The left and middle 1 kb repeats are identical (LDR and MDR, 1009 bp in length) and differ from the right 1 kb repeat (RDR, 1012 bp in length) at 63 positions (6·2% divergence; Piendl et al., 1994 ). The 1 kb repeats encode DNA-binding proteins displaying similarities to repressors of E. coli inducible operons, such as LacI, CytR and GalR. The RDR-encoded protein binds upstream and downstream of the right 1 kb repeat and upstream of the middle 1 kb repeat (Volff et al., 1996 ). The two identical upstream binding sites named LIP (for left imperfect palindrome) are 36 bp long and are located 8 nt upstream of the start codon of MDR and RDR (Volff et al., 1996 ). The downstream binding site is located in a 147 nt fragment flanking RDR on the right side. This fragment contains another imperfect palindromic sequence, RIP (for right imperfect palindrome). RIP is 32 bp long and is found 90 nt downstream from RDR. However, RIP is not a binding site for the 1-kb-repeat-encoded proteins (Volff et al., 1996 ). A 34 nt sequence with similarities to LIP, called PBS (for potential binding sequence), is also present in the 147 nt fragment 32 nt downstream of RDR. The presence of both up- and downstream binding sites as well as a functional protein encoded by the RDR is important for efficient amplification of AUD1 (Volff et al., 1996 ).



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Physical map of the S. lividans TK64 AUD1 region with the AUD1 sequence consisting of the three 1 kb and two 4·7 kb direct repeats (open arrows) and the various ORFs (below) found inside AUD1 and in the flanking region. The location of the binding sites (LIP and PBS) of the DNA-binding protein, encoded by the right 1 kb direct repeat (RDR), as well as a further imperfect inverted repeat (RIP) are shown. The orientation of the genes with respect to the chromosomal end, argG (responsible for arginine auxotrophy), cml (chloramphenicol resistance gene) and replication origin are indicated.

 
The homology with several bacterial operon repressors suggested a similar function for the 1-kb-repeat-encoded proteins in S. lividans. This prompted us to search for the presence of an operon by sequencing the regions flanking AUD1. No ORFs with the typical Streptomyces codon usage were found in the 2 kb DNA to the left of AUD1 (Volff et al., 1996 ). In contrast, we report in this paper the presence of five ORFs to the right of AUD1 which are probably part of an operon negatively regulated by the proteins encoded by the 1 kb repeats.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Strains, cultures and transformations.
S. lividans TK64 is streptomycin-resistant and auxotrophic for the amino acid proline (Hopwood et al., 1985 ). Strains Jni1, WP and AJ100 are independent spontaneous CmlS Arg- mutants derived from TK64 (Altenbuchner et al., 1988 ; Volff et al., 1996 ; Fig. 3a). Strain Jni1 possesses a high-copy-number amplification of AUD1. The AUD1 sequence of strain AJ100 (Altenbuchner & Eichenseer, 1991 ) is composed of a partially deleted 4·7 kb repeat flanked on its right side by a 1 kb repeat (MDR/RDR) which was the product of a recombination event between MDR (80%) and RDR (20%) (unpublished data). S. lividans WP has AUD1 and the flanking agl genes completely deleted. Mutants Jni26 (aglA::tsr), Jni29 (LDR, MDR, RDR::tsr) and Jni30 (LDR, MDR/RDR::tsr) were constructed by gene replacement in strain TK64. Jni27 (MDR/RDR::tsr) was obtained from strain AJ100 (Fig. 3a).



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 3. Effect of disruption, deletion and amplification of the 1 kb repeats on the basal {alpha}-glucosidase activity. (a) Structure of AUD1 in the mutants studied. The black arrow represents the tsr gene and its transcriptional orientation. (b) Relative {alpha}-glucosidase activity of the mutants compared to TK64. The specific activity of TK64 (0·086±0·029 U mg-1) was set to 100%; values are the mean of at least three independent measurements having a standard error of <25%.

 
S. lividans strains were cultivated at 30 °C on Hickey-Tresner (HT) agar medium (Pridham et al., 1957 ) or M40 minimal solid medium without asparagine but plus the required amino acids (Polsinelli & Beretta, 1966 ). For protoplast regeneration, R2 plates were used (Hopwood et al., 1985 ). For genomic DNA extraction, strains were grown in HT liquid medium. To determinate the {alpha}-glucosidase activity, S. lividans strains were cultivated in 50 ml modified HT liquid medium without dextrins. Maltose, maltooligosaccharides (C2–C8), dextrin, trehalose or sucrose were added at a final concentration of 0·5%. S. lividans protoplast preparation, transformation and regeneration were done according to Hopwood et al. (1985) . Genetic instability was analysed as described by Altenbuchner & Cullum, 1985 ).

E. coli JM109 (Yanisch-Perron et al., 1984) was used as host in cloning experiments. All E. coli cultures, including those for selective expression, were done in dYT medium or on dYT agar plates (Sambrook et al., 1989 ). {lambda} phage were propagated on E. coli NM538 (Frischauf et al., 1983 ) in dYT liquid medium or dYT soft agar supplemented with 10 mM Mg2+ and 0·2% maltose. For E. coli transformation, competent cells were prepared as described by Chung et al. (1989) and transformants were selected with ampicillin (100 mg l-1).

DNA manipulations and sequencing.
S. lividans and E. coli plasmid preparations were done using the alkaline lysis procedure of Kieser (1984) . Genomic DNA was isolated according to Hopwood et al. (1985) . Southern blot experiments and DIG labelling of DNA were done as described by Volff et al. (1996) . Construction of a genomic library of S. lividans TK64 in {lambda}EMBL4, plaque hybridization with [{alpha}-32P]dCTP-labelled DNA and isolation of {lambda} DNA were done as described by Eichenseer & Altenbuchner (1994) .

Sequencing reactions were performed on double-stranded DNA using the Autoread Sequencing kit and an automated laser fluorescent ALF Sequencer (Pharmacia Biotech). Nucleotide sequences were analysed using programs of the GCG Wisconsin package (Devereux et al., 1984 ). For the codon preference program (Gribskov et al., 1984 ), a codon usage table from eight Streptomyces genes was generated as described by Pfeifer et al. (1992) . Database searches were run with the programs BLASTN, BLASTX and TBLASTN (Altschul et al., 1990 ) on the BLAST e-mail server from the National Center for Biotechnology Information, Bethesda, MD, USA.

Gel redardation assays.
LDR was amplified from plasmid pMT680 as described for RDR (Volff et al., 1996 ) using the primers S485 (5'-GCG CTA GTC TTG CGC ATA TGA CGC GAC GAC TTG C-3') and S1014 (5'-AAA AGG ATC CCG CAA GGA ACT TGC GGT-3'). The restriction sites for BamHI and NdeI in the primer sequences allowed insertion of the PCR fragment into the L-rhamnose-inducible vector pJOE2702 (Volff et al., 1996 ). The resulting plasmid, pJOE2935, and plasmid pJOE2797 (RDR in pJOE2702) constructed earlier (Volff et al., 1996 ), were transferred into E. coli JM109. Expression of the genes was induced by growing the strains at 30 °C in dYT medium to an OD600 of 0·5, followed by addition of 0·2% rhamnose and incubation for 3 h before preparation of crude extracts. Gel mobility shifts were done in 5% polyacrylamide gels using the LIP-containing HindIII fragment of plasmid pJOE2201 (Volff et al., 1996 ) or the PBS-containing HindIII fragment of pJOE2864-2. The plasmid pJOE2864-2 contains the two complementary oligonucleotides S973 (5'-GATCCATTTACGAAATCAGCGCGAGATATTGCGCCACTTGCATG-3') and S974 (5'-CAAGTGGGCGCAATATCTCGCGCTGATTTCGTAAATG-3'), inserted between the SphI and BamHI site of pIC20H. The two HindIII fragments of pJOE2201 and pJOE2864-2 were labelled by filling in the ends with [{alpha}-32P]dATP and Klenow polymerase as described by Volff et al. (1996) . Sugars (sucrose, maltose, C2–C8 maltooligosaccharides and trehalose) were included in the binding assay at a final concentration of 1 mM.

Selective expression of aglA.
aglA was PCR-amplified in the presence of 10% DMSO (30 cycles, annealing temperature 58 °C) from plasmid pJOE683. Primers (S1069, 5'-AAA ACA TAT GGC AGC CCC CAA CTC GCA CC-3'; and S1070, 5'-AAA AAG ATC TGG CGC GCA GCC ACA CCG TGT-3') were designed to introduce a NdeI site overlapping the ATG start codon of aglA and a BglII site directly flanking its last C-terminal codon. For selective expression in E. coli, aglA was inserted as an NdeI–BglII fragment into pJOE2775 (Krebsfänger et al., 1998 ) under the control of the rhaB promoter. The six histidine codons present in the vector were fused to aglA. The correct sequence of the amplified aglA gene was verified by DNA sequencing of the resulting plasmid, pJOE2951. E. coli JM109 cells transformed with pJOE2951 were induced with 0·2% rhamnose for 3 h at 30 °C.

Determination of {alpha}-glucosidase activities with p-nitrophenyl {alpha}-D-glucopyranoside (p-NPG).
Enzyme activity was measured spectrophotometrically by monitoring the release of p-nitrophenol from p-NPG. Cells of S. lividans and E. coli were grown in 50 ml medium, washed twice and resuspended in enzyme buffer (20 mM Tris, 100 mM NaCl, 0·02% NaN3, pH 8·0). Crude extracts (20 µl, prepared either by sonication or by French press) or purified enzyme were added to 600 µl 0·1 M sodium phosphate pH 7·0 and 400 µl of a 13 mM pNPG solution in the same buffer and incubated at 25 °C until a yellow colour appeared. The reaction was stopped by addition of 500 µl 0·4 M Na2BO3 pH 9·8 and the absorbance at 405 nm was measured. For Km determination, the final p-NPG concentration was varied from 0·1 to 10 mM. Protein concentration was determined using the method of Bradford (1976) . One unit of {alpha}-glucosidase activity was defined as the amount of enzyme hydrolysing 1 µmol p-NPG under these conditions. The temperature stability of AglA was determined by incubation of the enzyme for 15 min at temperatures between 25 °C and 60 °C and cooling on ice for several minutes before adding it to the reaction mixture. The pH optimum of AglA was determined by using 600 µl 0·66 M potassium/sodium phosphate buffer pH 5·0–9·0 or 0·05 M Tris/HCl pH 7·0–9·0 as reaction buffer.

Gene disruption experiments.
To disrupt the chromosomal aglA gene in S. lividans TK64, the tsr gene conferring resistance to thiostrepton (TsR, Thompson et al., 1980 ) was inserted into the BamHI site of aglA contained in the 3·8 kb insert of pJOE683 (this work). This insert ligated with tsr was then cloned into the E. coliStreptomyces shuttle vector pJOE1082. This temperature-sensitive vector carries the minimal replicon and kanamycin resistance gene of Streptomyces plasmid pGM11 (Muth et al., 1989 ) ligated to the E. coli plasmid pUC19 (Vieira & Messing, 1982 ). The resulting plasmid, pJni5, was used to disrupt aglA. To inactivate the 1 kb repeat(s), the tsr gene was inserted into the PstI site of the right 1 kb repeat RDR (Piendl et al., 1994 ). The resulting disrupted repeat was cloned into pJOE1082 and the resulting plasmid was called pJOE2369. Protoplasts of S. lividans were transformed with pJni5 and pJOE2369. After selection of the transformants with thiostrepton (50 µg ml-1) at 30 °C, spores were separated from mycelium by filtration (Hopwood et al., 1985 ) and plated on HT medium at 40 °C to inhibit the replication of the temperature-sensitive vector pJOE1082. After a further plating of spores on HT medium plus thiostrepton at 40 °C, single colonies were tested for their growth on HT medium plus kanamycin (15 µg ml-1). Thiostrepton-resistant and kanamycin-sensitive colonies were retained and further cultivated. The success of the gene disruption and the loss of vector pJOE1082 were confirmed by Southern blot hybridization using pJni5, pJOE2369, the tsr gene and pJOE1082 as probes (data not shown).

Construction of plasmids pJOE3072-9, pJOE3073-1 and pJOE3076-2.
These three plasmids are derivatives of the E. coliS. lividans shuttle vector pJOE3069, which is similar to pEI16 described previously (Volff et al., 1996 ). The plasmid pJOE3069 contains the SCP2 origin of replication on a SalI–XhoI fragment and the SCP2 stability region on a BamHI–SacI fragment, both inserted into pUC13. Plasmid pJOE3076-2 contains the tsr gene inserted into pJOE3069 as a HindIII fragment from plasmid pJOE803 (Altenbuchner & Eichenseer, 1991 ). The agl genes, together with the C-terminal end of RDR, were obtained from {lambda}559-1 as an EcoRI–PstI fragment and inserted into pJOE803 (pJOE3063-2). The agl genes were isolated again from pJOE3063-2 together with the tsr gene as a HindIII–PstI fragment and combined with a HindIII–PstI fragment from pJOE2434 (Volff et al., 1996 ) containing the N-terminal end of RDR together with 73 bp upstream sequence. This HindIII fragment was inserted into pJOE3069 to give pJOE3072-9. For pJOE3073-1, the PstI site in the RDR was destroyed by Klenow polymerase. Nucleotide sequencing of the deleted region showed that instead of a 4 bp deletion, actually 11 bp were missing.

Substrate specificity of the {alpha}-glucosidase.
To 30 µl of reaction buffer (100 mM sodium phosphate pH 7·0) 10 µl substrate (500 mM stock solution in 100 mM sodium phosphate) and 10 µl enzyme (usually 2·4 µg purified enzyme, but when sucrose was used as substrate, the enzyme solution was tenfold diluted) were added and incubated for 60 min at 25 °C. To determine the Km for sucrose, the final concentration of sucrose in the reaction mixture was varied from 10 mM up to 900 mM. The reaction was stopped by heating to 90 °C for 2 min. Granutest 250 (E. Merk; 1 ml) was added and after 45 min at 25 °C the absorbance at 340 nm was determined on a Kontron spectrophotometer. The test is based on reduction of NAD+ by glucose dehydrogenase oxidizing the glucose which is generated by the enzyme reaction to gluconolactone. The amounts of glucose were obtained from a calibration curve generated with defined glucose concentrations. In addition, the products of enzyme reactions were determined and quantified by HPLC using a Hamilton RCX-10 column (250x4·1 mm) and an electrochemical detection unit, CoulochemII (ESA), with the analytical cell 5040. The mobile phase was 0·1 M NaOH and the flow rate was 0·75 ml min-1. For tri- and tetrasaccharides and phosphorylated disaccharides, the mobile phase was 0·22 M NaOH/0·02 M sodium acetate.

Purification of AglA.
E. coli JM109(pJOE2951) was grown in 50 ml dYT at 37 °C to an OD600 of 0·5, the culture shifted to 30 °C, L-rhamnose added to a final concentration of 0·1% and incubation continued for a further 3 h. The cells were harvested by centrifugation (Sorvall SS34, 10 min, 6000 r.p.m., 4 °C), washed in enzyme buffer (20 mM Tris, 100 mM NaCl, 0·02% NaN3, pH 8·0) and resuspended in 3 ml enzyme buffer. The cells were disrupted using a French press (two cycles at approx. 800 bar), the crude extract clarified by centrifugation (Sorvall SS34, 30 min, 20000 r.p.m., 4 °C) and the supernatant (1 ml) applied to a metal affinity spin column (Talon, Clontech Laboratories) equilibrated with enzyme buffer. After centrifugation at 600 r.p.m. for 2 min in a Heraeus Megafuge with a swingout rotor, the column was washed twice with 1 ml enzyme buffer and the enzyme eluted with 2x1 ml enzyme buffer supplemented with 50 mM imidazole. Finally, the two fractions were combined and dialysed twice against enzyme buffer.

Gel filtration of AglA.
The native molecular mass of AglA was determined by gel filtration on a Superdex 75, PC 3.2/30 column (Smart system, Pharmacia), calibrated with ribonuclease A (13700 Da), chymotrypsin (25000 Da), ovalbumin (43000 Da) and bovine serum albumin (67000 Da). The column was equilibrated with enzyme buffer and the enzyme eluted with this buffer at a flow rate of 0·04 ml min-1. The elution profile was monitored at 280 nm and the presence of active enzyme in the fractions by enzyme assays with p-NPG as substrate.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Nucleotide sequence of the region flanking AUD1 on the right side
The AUD1 region, isolated on a 15 kb BclI fragment, has been completely sequenced (Volff et al., 1996 ). To extend this sequence to the right side, a genomic library of S. lividans TK64 in {lambda}EMBL4 (Frischauf et al., 1983 ) was screened by plaque hybridization with pMT682, a pEMBL8 plasmid containing the right end of the 15 kb BclI fragment (Altenbuchner & Cullum, 1985 ). Two recombinant phages were obtained ({lambda}559-1 and {lambda}559-5) which carried DNA from beyond the right BclI site. This DNA was inserted into the E. coli pIC20H vector (Marsh et al., 1984 ) and sequenced by subcloning restriction fragments or by transposon tagging. About 7 kb of DNA was sequenced in this way, extending the nucleotide sequence of the AUD1 region to 20996 bp. The sequence was examined by codon usage analysis and similarity searches in various databases, which identified five new ORFs, all with the same orientation. The deduced gene products of the first three ORFs showed similarities to the periplasmic-binding-protein-dependent ABC transporters and the product of the fourth ORF to {alpha}-glucosidases. To the last one, no possible function could be attributed.

The similarities indicated that the ORFs may form an operon for uptake and degradation of an {alpha}-glycosidic sugar and were named tentatively aglE, aglF, aglG, aglA and aglX (Fig. 1 and Table 1). The first ORF, aglE, begins 151 bp downstream of the stop codon of RDR (right 1 kb repeat of AUD1) and encodes a protein of 471 amino acids. The deduced protein shows similarities to periplasmic sugar-binding proteins. The N-terminal amino acid sequence is in accordance with consensus sequences of signal peptides of secreted proteins with a hydrophilic region containing six arginine residues followed by a hydrophobic region of 14 aa and a signal petide cleavage site, LAAC, which is recognized by signal peptidase II (van Heijne, 1989 ). This indicates that AglE is a lipoprotein. The next two predicted gene products, AglF and AglG, have sizes of 231 and 291 aa, respectively, and show similarity to the MalF-MalG proteins, integral membrane components of the E. coli maltose transport system (Boos & Lucht, 1996 ). AglF and AglG are hydrophobic proteins according to a Kyte–Doolittle plot as expected (data not shown). AglA (534 aa) is an {alpha}-glucosidase, as indicated by high aa sequence identity with a maltooligosaccharide hydrolase (61% identity, Table 1) and the E. coli trehalose-6-phosphate hydrolase (38% identity), and later proved by enzyme assays. The aglX-encoded protein of 547 aa shows significant similarities to a putative mycobacterial protein of unknown function.


View this table:
[in this window]
[in a new window]
 
Table 1. ORFs on the right side of AUD1

 
Expression of aglA in E. coli, and purification and biochemical analysis of the enzyme
The homology of the putative product of aglA with different bacterial {alpha}-glucosidases and eukaryotic maltases predicted that AglA would have a similar enzyme activity. For further biochemical characterization, the gene was amplified by PCR and placed under the control of the L-rhamnose-inducible promoter in the plasmid pJOE2775 (Krebsfänger et al., 1998 ). In addition, the C-terminal end of the gene was fused to six histidine codons in the vector for purification of the gene product. The resulting plasmid pJOE2951 was transferred into E. coli JM109 and the {alpha}-glucosidic activity was measured spectrophotometrically by monitoring the release of p-nitrophenol from p-NPG. After 3 h induction with rhamnose at 30 °C, a 60-fold increase in enzyme specific activity from 0·07 U mg-1 to 4·03 U mg-1 was determined. The protein was purified (5·6-fold) from crude extract in one step by immobilized metal affinity chromatography, taking advantage of the His-tag attached to the end of the enzyme. The protein was more than 95% pure as estimated by densitometrical scanning of an SDS-polyacrylamide gel (Fig. 2). The molecular mass of the enzyme determined by SDS-PAGE was 57000 Da, which is in good agreement with the calculated molecular mass of 59148 Da. Gel filtration of the protein in comparison with known standards showed a molecular mass of 56000 Da for the native enzyme, which indicates that AglA has a monomeric structure. AglA activity was tested from pH 5·0 to 9·0, raising the pH in steps of 0·5. The pH optimum was at 7·5; there was no activity below pH 5·5 and at pH 9·0 the residual activity was about 15%. The enzyme was stable up to 35 °C; incubation for 15 min at 40 °C reduced activity to 60% and 15 min at 45 °C abolished activity completely.



View larger version (105K):
[in this window]
[in a new window]
 
Fig. 2. SDS-PAGE analysis of the purification of AglA from E. coli JM109 pJOE2951. Lanes: 1, protein size standard; 2, crude extract of E. coli JM109 pJOE2951 induced with rhamnose; 3, protein fraction which did not bind to the metal affinity spin column; 4, first washing fraction of the metal affinity spin column; 5, purified AglA eluted with 50 mM imidazole.

 
Substrate specificity of AglA
To find out more about the biological function of aglA and the neighbouring ORFs, AglA was incubated with a series of natural products containing an {alpha}-glucosidic bond, and with some phosphorylated disaccharides which might be intermediates in the catabolism of disaccharides. The reaction was stopped by heating, and the release of glucose was monitored by HPLC analysis using an electrochemical detection unit and also by measuring the NADH produced by glucose dehydrogenase in a second reaction using a spectrophotometer. Furthermore, a series of sugars linked by an {alpha}-glycosidic bond with p-nitrophenol were tested to see if sugars different from glucose might be accepted as substrate. The results are summarized in Table 2 and demonstrate a high substrate specificity of AglA. From the p-nitrophenol-containing compounds only p-nitrophenyl {alpha}-D-glucopyranoside, and from the oligosaccharides only sucrose, were hydrolysed at a fairly high rate. No activity was seen with phosphorylated sucrose or other phosphorylated disaccharides. With p-NPG or sucrose as substrate the specific activity of AglA was nearly the same (22·60 and 14·90 U mg-1, respectively) under standard conditions (5·2 mM p-NPG or 100 mM sucrose). When the Km values for the two substrates were determined, a nearly 100-fold difference was found (Km for p-NPG=1·5 mM and Km for sucrose=110 mM). At non-limiting sucrose concentrations (600–900 mM), the specific activity of AglA was 32·4 U mg-1.


View this table:
[in this window]
[in a new window]
 
Table 2. Substrate specificity of AglA

 
Disruption of aglA in S. lividans
Release of p-nitrophenol from p-NPG was detected in the presence of crude extracts of S. lividans TK64 after culturing in M40 liquid minimal medium plus glycerol or in modified HT medium where dextrin was replaced by glycerol. To see if this activity was due to AglA, the aglA gene was disrupted by insertion of the thiostrepton resistance gene, tsr (Thompson et al., 1980 ), and the {alpha}-glucosidase activity in the resulting strain Jni26 was compared with TK64. In addition, to see if the {alpha}-glucosidase activity could be increased by induction with {alpha}-glucosides, both strains were grown in modified HT-medium supplemented with either glycerol, dextrin, maltose, sucrose or trehalose for 24 h and the {alpha}-glucosidase activity determined in crude extracts with p-NPG as substrate. No difference was found between S. lividans TK64 and Jni26. Both strains exhibited about the same basal {alpha}-glucosidase activity, and the presence of maltose, sucrose or trehalose did not lead to a significant increase in activity. Only with dextrin was an about 10-fold higher activity observed in both Jni26 and TK64 (Table 3). This indicated that there is a second, dextrin-inducible {alpha}-glucosidase in S. lividans.


View this table:
[in this window]
[in a new window]
 
Table 3. {alpha}-Glucosidase activity in S. lividans TK64 and Jni26 grown in modified HT medium with different carbon sources

 
The 1-kb-repeat-encoded protein as repressor of the agl genes
The 1 kb repeat RDR was disrupted in TK64 by the tsr gene and the resulting strain was called Jni29 (Fig. 3). In strain Jni30, which was also constructed from TK64, one event of homologous recombination necessary for disruption had occurred in RDR and the second in MDR. The resulting strain had an intact LDR, a disrupted second 1 kb repeat consisting of the recombined MDR and RDR, and a deletion of one 4·7 kb repeat. Jni1 and AJ100 are two CmlS Arg- spontaneous mutants isolated from TK64. Strain Jni1 contains a high amplification of AUD1 and therefore presumably of MDR. Crude extracts of this strain retarded DNA fragments containing the binding sites of the 1-kb-repeat-encoded proteins, whilst crude extracts of strain TK64 did not (Volff et al., 1996 ). Strain AJ100 contains a total deletion of one 4·7 kb repeat, a partial deletion of the other, a deletion of LDR and one unique 1 kb repeat recombined from the MDR sequence (about 80%) and the RDR sequence (20%). Strain Jni27 was constructed by disruption of the unique 1 kb repeat of AJ100 (Fig. 3). All strains were grown in modified HT medium with glycerol and the {alpha}-glucosidase activity determined from crude extracts with p-NPG as substrate. Jni27 showed an {alpha}-glucosidase activity higher than that of TK64 (about 10-fold) and of its more closely related parent AJ100 (about 17-fold). This strongly suggested that the 1 kb repeat encodes the repressor of the agl genes. Strains Jni29, Jni30 and AJ100 showed no increased activity, indicating that LDR, MDR and the composite repeat of AJ100 control the expression of aglA. No reduction in enzyme activity was observed in the amplified strain Jni1, probably indicating that most of the activity observed derives from the second gene present in S. lividans strains (Fig. 3).

Cmls Arg- mutants of S. lividans are highly unstable and therefore the possibility that there were additional rearrangements in strain Jni27 leading to increased expression of the agl genes could not be completely excluded. A further experiment was performed to definitively corroborate the role of the RDR-encoded protein as repressor of the agl genes. All five agl genes together with RDR and a thiostrepton resistance gene were inserted into a low-copy SCP2 plasmid derivative to give plasmid pJOE3072-9. A second plasmid was generated by deleting the PstI site in pJOE3072-9, which introduced a frameshift mutation in RDR (pJOE3073-1), and for a control, the tsr gene alone was inserted into the SCP2 vector (pJOE3076-2). All three plasmids were introduced into S. lividans WP, a Cmls Arg- mutant of S. lividans which has AUD1 and the agl genes deleted. After growth in modified HT medium without dextrin, the {alpha}-glucosidase activity was determined. As can be seen from Fig. 4, there was a slight increase in enzyme activity in S. lividans WP with pJOE3072-9 compared to the vector control and a 10-fold increase when RDR was inactivated, which confirms the repressor function of the RDR-encoded protein.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 4. Restriction map of the plasmids pJOE3072-9, pJOE3073-1 and pJOE3076-2. The {alpha}-glucosidase activities (mean values±standard deviations) obtained after transformation of S. lividans WP with these plasmids are shown on the right side [U (mg protein) -1]. From each transformation, three colonies were grown separately in modified (without dextrin and other additional carbohydrates) HT medium with thiostrepton (10 µg ml-1) for 3 d. Enzyme activity was determined with p-NPG as substrate. Activity of each colony was measured at least twice in independent experiments.

 
To demonstrate a binding of the protein encoded by the left and middle 1 kb direct repeat, LDR was amplified by PCR from a plasmid containing only LDR (pMT680; Altenbuchner & Cullum, 1985 ) and inserted into a rhamnose-inducible vector, pJOE2702, as described for RDR (Volff et al., 1996 ). Crude extracts from rhamnose-induced cultures of E. coli JM109 containing the vector pJOE2702, pJOE2797 (RDR) and pJOE2935 (LDR) were incubated with a radioactively labelled fragment containing the already identified binding site LIP, which was bound by the protein encoded by RDR (Volff et al., 1996 ) and analysed by mobility shift assay (Fig. 5a). To determine whether the 1-kb-repeat-encoded proteins respond to maltose, maltooligosaccharides or sucrose, binding was done in the presence or in the absence of these sugars (in Fig. 5 only the results of binding without sugar and with 10 mM maltooligosaccharides are shown). Fig. 5 shows that both proteins bind to LIP. Interestingly, despite the very low difference in amino acid sequence between the LDR- and RDR-encoded proteins, there is a clear difference in migration of the retarded bands. The presence of the sugar sources did not modify the binding of the 1-kb-repeat-encoded proteins. This is consistent with the results previously obtained and reinforces the conclusion that sucrose, maltose and maltooligosaccharides are not inducers of the putative agl operon.



View larger version (86K):
[in this window]
[in a new window]
 
Fig. 5. (a) Mobility shift assay with the LDR- and RDR-encoded proteins and the DNA-binding site LIP located upstream of the 1 kb repeats. A HindIII fragment from plasmid pJOE2201 containing the LIP sequence was labelled with [{alpha}-32P]dATP and incubated with crude extract from E. coli JM109(pJOE2702) (vector control; lanes 1, 4), E. coli JM109(pJOE2797) (RDR-protein; lanes 2, 5), and E. coli JM10(pJOE2935) (LDR-protein; lanes 3, 6). In lanes 4, 5 and 6, 10 mM of a mixture of maltooligosaccharides (maltose–maltooctaose) was added to the binding reaction. (b) Mobility shift assay with the DNA-binding site PBS obtained from plasmid pJOE2864-2 as a HindIII fragment. The labelled fragment was incubated with crude extract from E. coli JM109(pJOE2702) (vector control; lane 1) and from E. coli JM109(pJOE2797) (RDR-encoded protein; lane 2).

 
LIP is located upstream of RDR. A second binding site was located in the intercistronic region between the end of RDR and the beginning of aglE. It was shown previously that a palindromic sequence called RIP in this region was not bound by the 1-kb-repeat-encoded protein. Another candidate for binding was a 36 bp sequence which showed homology to LIP. This nucleotide sequence, called PBS (potential binding site), was synthesized as two complementary oligonucleotides and inserted into the polylinker sequence of pIC20H (pJOE2864-2). From there it was cut out again by HindIII, labelled with Klenow polymerase and [{alpha}-32P]dCTP and used in mobility shift experiments. As can be seen in Fig. 5(b), the sequence is indeed shifted by the RDR-encoded protein.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We report here that the amplifiable element AUD1 of S. lividans TK64 is on one side flanked by five ORFs. The S. lividans AUD1 right 1 kb repeat and the following ORFs show similarity to the maltose operon of S. coelicolor A3(2) (van Wezel et al., 1997a , b ). There is a lacI-like repressor gene, an intercistronic promoter/operator region, where the repressor binds, genes for an ABC transport system (composed of an extracellular sugar-binding protein and two membrane proteins for the uptake of the sugar) and a gene encoding an {alpha}-glucosidase. The lack of a gene encoding a MalK-homologous ATP-binding protein, which is necessary to provide energy to the sugar transport systems, is not unusual. It is also missing in the S. coelicolor maltose operon (van Wezel et al., 1997a ). It seems that there is a malK-homologous gene msiK in S. lividans (and S. coelicolor) which is used by different ABC transport systems responsible for uptake of xylobiose, cellobiose and other disaccharides (Hurtubise et al., 1995 ; Schlösser et al., 1997 ).

No function could be attributed to the fifth gene, aglX, found in S. lividans. It is not known whether aglX belongs to the putative agl operon and if there are even more agl genes downstream since no transcription terminator sequence was found behind aglA and aglX.

The 1-kb-repeat-encoded protein binds 8 nt upstream of its start codon (for MDR and RDR) at a sequence called LIP which would allow autoregulation of its own synthesis. It also binds in the region between RDR and aglE at a sequence called PBS which shows significant similarity to LIP. A nucleotide sequence fitting the consensus sequence of a Streptomyces vegetative promoter was identified within the 151 bp intercistronic region (Fig. 6). PBS overlaps the -35 sequence of this putative promoter sequence and therefore the binding of the RDR-encoded protein to PBS would inhibit binding of the RNA polymerase to this promoter. When the agl genes are transcribed by this promoter the mRNA would start at the beginning of RIP. The function of RIP remains unclear. It could be the target of a global regulatory system. As a part of the leader region of the agl mRNA it could also stabilize the mRNA by a stem-and-loop secondary structure as described by Deng et al. (1990) .



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 6. DNA sequence between the end of RDR and aglE. The putative vegetative promoter (-10, -35) and ribosome-binding sites are in bold and underlined, the PBS and RIP sequence are in boxes and the palindromic sequence of RIP is indicated by arrows. Start and stop codons are in bold.

 
Genetic and biochemical studies suggested that the agl genes do not constitute a maltose or trehalose operon. Two {alpha}-glucosidase activities could be identified in S. lividans. The activity of AglA was determined by sequencing and expression of the gene in E. coli. But the basal and dextrin-inducible {alpha}-glucosidase activity found in crude extracts of S. lividans was due to a second gene since inactivation or even deletion of aglA had no influence on this activity. This second gene might be part of a genuine maltose operon. The gene reg1, which was isolated from S. lividans as a regulatory gene of {alpha}-amylase and chitinase activity, is nearly identical to the malR gene of the S. coelicolor maltose operon (Nguyen et al., 1997 ) and might be in fact the regulator of the S. lividans maltose operon. S. coelicolor and S. lividans grow very poorly on maltose (van Wezel et al., 1997b a nd our own results). The reason might be that maltooligosaccharides, i.e. degradation products of dextrin, but not maltose are the inducers of the maltose operons in the two strains. Indeed, inactivation of MalR in S. coelicolor improved growth on maltose considerably (van Wezel et al., 1997b ) and addition of maltooligosaccharides (C2–C8) to the growth medium of S. lividans had about the same effect on {alpha}-glucosidase activity as dextrin (data not shown).

According to its amino acid sequence, AglA clearly belongs to the family 13 of glycoside hydrolases. Members of this family retain the configuration at the anomeric centre of the substrate during hydrolysis and the catalytic domain of the enzymes shows a characteristic (ß/{alpha})8 barrel structure. Secreted and intracellular enzymes with various substrate specificities, such as {alpha}-glucosidases, {alpha}-amylases, pullulanases, cyclodextrinases or trehalose-6-phosphate hydrolase, belong to this family (Henrissat, 1991 ; Henrissat & Bairoch, 1993 ; Warren, 1996 ). For the {alpha}-glucosidase MalL of Bacillus subtilis, a member of this family, sucrose was also the preferred substrate (Schönert et al., 1998 ). But the specific activity of MalL was about 10-fold higher and the Km for sucrose 10-fold lower than for AglA. Despite the preference for sucrose, the expression of the corresponding gene was induced by maltose and not by sucrose and the biological function of MalL seems to be in the breakdown of internal storage polysaccharides (Schönert et al., 1998 ).

The function of AglA and of the other Agl proteins remains unclear since S. lividans does not grow on sucrose. Colonies of S. lividans grew equally well on minimal medium agar plates with sucrose or without any carbon source. Also, addition of sucrose to complete medium did not induce expression of aglA. The most obvious explanation would be that we have not identified the natural substrate of the agl genes. The very high Km of AglA for sucrose (100 mM) might point in this direction.

Another reason might be that AUD1 and the agl genes, which are located in the unstable region of the S. lividans chromosome, were accidentally fused. This would mean that the inducer is different from the substrate of AglA. A further possibility would be that there was originally only one regulatory gene for the agl genes. This regulatory gene was duplicated twice during the evolution of the strain and by mutagenesis evolved a different function, for example to protect the chromosome in unstable mutants from further deletions by promoting amplification of the AUD1 sequence. When strain Jni27 with the inactivated regulatory gene was plated on agar plates with minimal medium and sucrose as carbon source there was no improvement in growth compared to TK64, AJ100 or Jni26, which makes the latter two possibilities unlikely. Alternatively, the agl genes may have lost their function by mutation in a similar manner to pseudogenes in higher eukaryotes.


   ACKNOWLEDGEMENTS
 
J.-N. Volff was a recipient of a long-term fellowship from the European Molecular Biology Organization (EMBO). We thank Hilde Watzlawick and Gisela Kwiatkowski for their help in purification and biochemical characterization of AglA, Mervin Bibb for critical reading of the manuscript and Ralf Mattes for support and helpful discussions.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Altenbuchner, J. & Cullum, J. (1984). DNA amplification and an unstable arginine gene in Streptomyces lividans 66.Mol Gen Genet 195, 134-138.[Medline]

Altenbuchner, J. & Cullum, J. (1985). Structure of an amplifiable DNA sequence in Streptomyces lividans 66.Mol Gen Genet 201, 192-197.[Medline]

Altenbuchner, J. & Eichenseer, C. (1991). A new system to study DNA amplification in Streptomyces lividans. In Genetics and Product Formation, pp. 253-263. Edited by S. Baumberg, H. Krügel & D. Noack. New York & London: Plenum.

Altenbuchner, J., Eichenseer, C. & Brüderlein, M. (1988). DNA amplification and deletion in Streptomyces lividans. In Biology of Actinomycetes ’88, pp. 139-144. Edited by Y. Okami, T. Beppu & H. Ogawara. Tokyo: Japan Scientific Societies.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990). Basic local alignment search tool.J Mol Biol 215, 403-410.[Medline]

Boos, W. & Lucht, J. M. (1996). Periplasmic binding protein-dependent ABC transporters. In Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, 2nd edn, pp. 1175–1209. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of protein–dye binding.Anal Biochem 72, 248-254.[Medline]

Chung, C. T., Niemela, S. L. & Miller, R. H. (1989). One-step preparation of competent Escherichia coli: transformation and storage of bacterial cells in the same solution.Proc Natl Acad Sci USA 86, 2172-2175.[Abstract]

Deng, T., Noel, J. & Tsai, M.-D. (1990). A novel expression vector for high-level synthesis and secretion of foreign proteins in Escherichia coli: overproduction of bovine pancreatic phospholipase A2.Gene 93, 229-234.[Medline]

Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX.Nucleic Acids Res 12, 387-395.[Abstract]

Eichenseer, C. & Altenbuchner, J. (1994). The very large amplifiable element AUD2 from Streptomyces lividans 66 has insertion sequence-like repeats at its ends.J Bacteriol 176, 7017-7112.[Abstract]

Frischauf, A.-M., Lehrach, H., Poustka, A. & Murray, N. E. (1983). Lambda replacement vectors carrying polylinker sequences.J Mol Biol 170, 827-842.[Medline]

Gribskov, M., Devereux, J. & Burgess, R. R. (1984). The codon preference plot: graphic analysis of protein coding sequences and prediction of gene expression.Nucleic Acids Res 12, 539-549.[Abstract]

van Heijne, G. (1989). The structure of signal peptides from bacterial lipoproteins.Protein Eng 2, 531-534.[Abstract]

Henrissat, B. (1991). A classification of glycosyl hydrolases based on amino acid sequence similarities.Biochem J 280, 309-316.[Medline]

Henrissat, B. & Bairoch, A. (1993). New families in the classification of glycosyl hydrolases based on amino acid sequence similarities. Biochem J 293, 781-788.[Medline]

Hopwood, D. A., Bibb, M. J., Chater, K. F. & 7 other authors (1985). Genetic Manipulation of Streptomyces: a Laboratory Manual. Norwich: John Innes Foundation.

Hurtubise, Y., Shareck, F., Kluepfel, D. & Morosoli, R. (1995). A cellulase/xylanase-negative mutant of Streptomyces lividans 1326 defective in cellobiose and xylobiose uptake is mutated in a gene encoding a protein homologous to ATP-binding proteins.Mol Microbiol 17, 367-377.[Medline]

Kieser, T. (1984). Factors affecting the isolation of CCC DNA from Streptomyces lividans and Escherichia coli.Plasmid 12, 19-36.[Medline]

Krebsfänger, N., Zocher, F., Altenbuchner, J. & Bornscheuer, U. T. (1998). Characterization and enantioselectivity of a recombinant esterase from Pseudomonas fluorescens.Enzyme Microb Technol 22, 641-646.

Marsh, L. J., Erfle, M. & Wykes, E. J. (1984). The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation.Gene 32, 481-485.[Medline]

Muth, G., Nußbaumer, B., Wohlleben, W. & Pühler, A. (1989). A vector system with temperature-sensitive replication for gene disruption and mutational cloning in streptomycetes.Mol Gen Genet 219, 341-348.

Nguyen, J., Francou, F., Virolle, M.-J. & Guerineau, M. (1997). Amylase and chitinase genes in Streptomyces lividans are regulated by reg1, a pleiotropic regulatory gene.J Bacteriol 179, 6383-6390.[Abstract]

Pfeifer, O., Pelletier, I., Altenbuchner, J. & van Pée, K.-H. (1992). Molecular cloning and sequencing of a non-haem bromoperoxidase gene from Streptomyces aureofaciens ATCC 10762.J Gen Microbiol 138, 1123-1131.[Medline]

Piendl, W., Eichenseer, C., Viell, P., Altenbuchner, J. & Cullum, J. (1994). Analysis of putative DNA amplification genes in the element AUD1 of Streptomyces lividans 66.Mol Gen Genet 244, 439-443.[Medline]

Polsinelli, M. & Beretta, M. (1966). Genetic recombination in crosses between Streptomyces aureofaciens and Streptomyces rimosus.J Bacteriol 91, 63-68.

Pridham, T. G., Anderson, P., Foly, C., Linderfelser, C. W., Hesseltime, C. W. & Benedict, R. C. (1957). A selection of media for maintenance and taxonomic studies of Streptomyces. Antibiotics Annu, 947–953.

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

Schlösser, A., Kampers, T. & Schrempf, H. (1997). The Streptomyces ATP binding component MsiK assists in cellobiose and maltose transport. J Bacteriol 179, 2092-2095.[Abstract]

Schönert, S., Buder, T. & Dahl, M. K. (1998). Identification and enzymatic characterization of the maltose-inducible {alpha}-glucosidase MalL (sucrase-isomaltase-maltase) of Bacillus subtilis. J Bacteriol 180, 2574-2578.[Abstract/Free Full Text]

Thompson, C. J., Ward, J. M. & Hopwood, D. A. (1980). DNA cloning in Streptomyces: resistance genes from antibiotic-producing species.Nature 286, 525-527.[Medline]

Vieira, J. & Messing, J. (1982). The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers.Gene 19, 259-268.[Medline]

Volff, J. N., Eichenseer, C., Viell, P., Piendl, W. & Altenbuchner, J. (1996). Nucleotide sequence and role in DNA amplification of the direct repeats composing the amplifiable element AUD1 of Streptomyces lividans 66.Mol Microbiol 21, 1037-1047.[Medline]

Warren, R. A. J. (1996). Microbial hydrolysis of polysaccharides.Annu Rev Microbiol 50, 183-212.[Medline]

van Wezel, G. P., White, J., Bibb, M. J. & Postma, P. W. (1997a). The malEFG gene cluster of Streptomyces coelicolor A3(2): characterization, disruption and transcriptional analysis.Mol Gen Genet 254, 604-608.[Medline]

van Wezel, G. P., White, J., Young, P., Postma, P. W. & Bibb, M. J. (1997b). Substrate induction and glucose repression of maltose utilization by Streptomyces coelicolor A3(2) is controlled by malR, a member of the lacI-galR family of regulatory genes.Mol Microbiol 23, 537-549.[Medline]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors.Gene 33, 103-119.[Medline]

Received 25 October 1999; revised 31 December 1999; accepted 14 January 2000.