Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (190K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (19)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Bourgerie, S. J.
Right arrow Articles by Hyde, E. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bourgerie, S. J.
Right arrow Articles by Hyde, E. I.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 1995 Oxford University Press 1685-1693

Footnote

DNA binding and DNA bending by the MelR transcription activator protein from Escherichia coli

DNA binding and DNA bending by the MelR transcription activator protein from Escherichia coli Sylvain J. Bourgerie, Carmen M. Michán, Mark S. Thomas1, Stephen J. W. Busby* and Eva I. Hyde

School of Biochemistry, The University of Birmingham, PO Box 363, Birmingham B15 2TT, UK and 1Department of Medical Microbiology, University of Sheffield Medical School, Sheffield, UK

Received February 14, 1997; Revised and Accepted March 18, 1997

ABSTRACT

The Escherichia colimelR gene encodes MelR protein which is a member of the AraC/XylS family of bacterial transcription activators. The function of MelR was investigated by making a targeted deletion in the melR gene of the Escherichia coli chromosome. MelR is a transcription activator essential for melibiose- dependent expression of the melAB operon which is needed for bacterial growth with melibiose as a carbon source. To investigate the interactions of MelR at the melAB promoter, both full length MelR and a shortened derivative, MelR173, containing the C-terminal DNA-binding domain, were purified as fusions to glutathione- S-transferase. Circular permutation studies show that both full-length MelR and MelR173 induce an apparent bend upon binding to target sites at the melAB promoter. Bound full-length MelR, but not MelR173, can oligomerise to form larger complexes that are likely to be involved in transcription activation.

INTRODUCTION

Most transcription factors belong to a relatively small number of `families', with different proteins sharing many common features (1 ). Thus, in-depth study of a particular transcription factor can provide information that is applicable to other factors found in a range of organisms. Considerable attention has recently been focused on the AraC/XylS family of bacterial transcription activators and, to date, the genes encoding more than 50 family members have been sequenced (2 ). Most of these genes are essential for the triggering of transcription in response to particular metabolites. In recent work (3 ,4 ), we have studied one typical member of the AraC/XylS family, the MelR transcription activator protein from Escherichia coli. Webster et al. (5 ) originally identified MelR as a 302 amino acid protein encoded by an open reading frame located immediately adjacent to the E.coli melAB operon (encoding the genes required for melibiose metabolism) (6 ). The melR gene and the melAB operon are transcribed from two divergent promoters, pmelR and pmelAB, the transcription start points of which are separated by 225 bp (Fig. 1 ). It was shown that expression from pmelAB is activated by MelR and melibiose, and two binding sites for MelR upstream of the melAB transcription start were identified: these are two identical 18 bp sequences, organised as an inverted repeat separated by 20 bp (5 ,7 ). Both 18 bp target sequences are essential for optimal pmelAB activity (4 ). Like most members of the AraC/XylS family, MelR appears to consist of two domains of approximately equal size: the N-terminal domain is likely to be involved in triggering by melibiose, whilst the C-terminal domain is responsible for DNA binding. In support of this, it has been shown that a fragment of MelR carrying the 173 C-terminal amino acids can recognise each of the DNA sites for MelR at pmelAB (3 ).

From the few cases that have been studied in detail, it appears that AraC/XylS family members bind to their target promoters both in the presence and in the absence of the trigger ligand, with one protein subunit binding to one ~18 bp operator (8 ). Most target promoters contain at least two sites for binding of the AraC/XylS family member: it is generally assumed that the ligand triggers some sort of rearrangement of bound activator and this involves interactions between different sites. However, few details have been determined, since AraC/XylS family members are difficult to study in vitro, since they are poorly soluble and are difficult to purify (9 ).

In this work, we first investigated the function of MelR directly by making a targeted deletion in the melR gene on the E.coli chromosome. Next, to study fundamental properties of MelR (and, by inference, other members of the AraC/XylS family), we devised a simple purification for MelR, exploiting glutathione-S-transferase (GST) fusion technology. Study of the binding of MelR to pmelAB DNA reveals two interesting properties: first, upon binding to target sites, MelR appears to induce a bend in the DNA, and second, bound MelR subunits can oligomerise with other MelR subunits. We argue that the combination of bending and aggregation leads to the formation of multi-subunit nucleoprotein complexes that play a role in transcription activation by MelR (and by other family members). The GST fusion technology was also used to purify a C-terminal fragment of MelR carrying the DNA binding domain. This fragment can bind and distort target sites but is unable to oligomerise.

MATERIALS AND METHODS

Bacterial strains, plasmids and oligonucleotides

Standard methods for recombinant DNA manipulations were used (10 ). The bacterial strains, plasmids, DNA fragments and synthetic oligodeoxynucleotides used are listed in Table 1 .

Table 1 Bacterial strains, DNA fragments, plasmids and oligodeoxynucleotides
Escherichia coli strains
BL21 (lDE3)

 T7RNApol+ F- ompT rB- mB-

 ref. 26
M182

 [Delta]lac mel

 ref. 27
Y1089

 [Delta]lacU169 melR+

 ref. 28
Y1089[Delta]melR

 [Delta]lacU169[Delta]melR

 This work
DNA fragments
KK3

 1620 bp HindIII-BglII fragment carrying pmelAB, pMelR and MelR sequences ligated to BglII-HindIIIlinker to give a HindIIIfragment.

 ref. 5
 

 The nucleotide sequence of this fragment is shown in figure 5 of ref. 5.

  
KK3NPA

 KK3 fragment cut at the unique NsiIsite within MelR, and filled in with Klenow enzyme and religated to yield a gene encoding inactive protein.

 This work

KK15

 1200 bp BglII-EcoRV fragment carrying melR, pmelR and pmelAB (with EcoRI and HindIII linkers attached at the EcoRV and BglII ends, respectively).

 ref. 7
 

 The fragment is described fully in figure 1 of ref. 7.

  
KK33

 EcoRI-HindIII fragment containing pmelAB sequences from -136 to +21 (from HaeIII-BglIIfragment of KK3).

 refs 5 and 9
KK43

 EcoRI-HindIII fragment carrying pmelAB sequences from -136 to +21.

 ref. 16
 

 The KK43 fragment carries a GC to AT transversion at -73, creating a unique BglII site between MelR binding sites 1 and 2.

  
 

 The fragment is described fully in figure 1 of ref. 16.

  
KK67

 EcoRI-HindIII fragment carrying pmelAB sequences from -136 to +7.

 ref. 4
 

 A KK43 derivative with BamHIsite from -52 to -47 downstream of MelR binding site 2 and HindIII linker at +8, lacking the melA translation initiation region. The sequence of the fragment is shown in figure 1 of ref. 4.

  
NJS1

 Fragment carrying consensus CRPsite.

 ref. 20
 

 The sequence of this fragment is shown in figure 2 of ref. 20.

  
Plasmids
pAA121

 pBR322 derivative for cloning EcoRI-HindIII fragments.

 ref. 29
 

 Different fragments (above) were cloned in this vector.

  
pAA182

 Promoter probe vector allowing EcoRI-HindIII fragments to be cloned upstream of the promoter-less lac genes as a trpB trpA fusion.

 ref. 5
 

 Different fragments (above) were cloned in this vector.

  
pAA224

 Promoter probe vector allowing EcoRI-HindIII fragments to be cloned upstream of the promoter-less galE translation initiation region fused to lac. The fusion is shown in figure 2 of ref. 4. Different fragments (above) were cloned in this vector.

 ref. 4
pBend2

 Derivative of pBR322 containing 17 restriction sites in a direct repeat spanning a central region containing cloning sites (XbaI and SalI sites).

 ref. 17
pBend2-XS

 pBend2 derivative with MelR binding site 1 cloned between the XbaI and SalI sites.

 This work
pBend2-CRP

 pBend2 derivative with CRP site cloned into SalI site.

 This work
pCM117-303

 pET9d derivative encoding MelR on an NcoI-BamHI fragment controlled by T7 promoter (MelR303 contains an additional alanine at codon 2).

 ref. 3
pCM117-173

 pET9d derivative encoding MelR173 on an NcoI-BamHI fragment under the control of a T7 promoter.

 ref. 3
pGEX-5X-1

 Vector that allows the cloning of genes fused to GST controlled by the tac promoter.

 ref. 13
pGEX-NB

 Derivative of pGEX-5X-1 with an NcoI linker cloned into the BamHI site and a BamHI linker cloned into the EcoRI site.

 This work
pGEX-NB173

 Derivative of pGEXNB containing melR173 from pCM117-173 as an NcoI-BamHI fragment giving a GST fusion.

 This work
pGEX-NB303

 Derivative of pGEXNB containing melR from pCM117-303 as an NcoI-BamHI fragment giving a GST fusion.

 This work
pJW12

 pAA121 derivative carrying the melR gene.

 ref. 9
pJW12[Delta]BclI

 pAA121 derivative carrying the melR gene with 320 bp internal deletion between two BclI sites.

 This work
pMAK705

 Vector containing thermosensitive replication origin.

 ref. 11
 

 Encodes resistance to chloramphenicol.

  
pLysS

 pACYC184 derivative which supplies low levels of T7 lysozyme.

 ref. 30
Oligodeoxynucleotides
D9007

 5'-TGCTCTAGACGAAAACTCTGCTTTTC-3'

 This work
 

 This anneals between -133 and -107 upstream of pmelAB and places an XbaI site at -141 (upstream of site 1).

  
D9008

 5'-ACGCGTCGACTGCGTGAAGCAGCAGT-3'

 This work

 

 This anneals between -66 and -87 bases upstream of pmelAB and places a SalI site at -68 (downstream of site 1).

  


Figure 1. Organisation of the regulatory region of the E.coli mel operon. The diagram shows the divergent melR and melAB transcription start points. The location of the pmelAB -10 hexamer and MelR-binding sites 1 and 2 are indicated. The lower part of the figure shows the base sequence of the upper strand around the two MelR-binding sites in the KK43 fragment. Coordinates refer to base pairs upstream from the melAB transcript start point. Note the BglII site (AGATCT) from positions -68 to -73: restriction at this site permits the generation of a fragment carrying solely DNA site 1 for MelR.

Targeted disruption of the E.coli melR gene

The melR gene of E.coli strain Y1089 was disrupted using the gene replacement method of Hamilton et al. (11 ). To do this, a 320 bp internal deletion was created in the melR gene by digesting plasmid pJW12 (9 ) with BclI and religating to give pJW12[Delta]BclI (plasmid pJW12 contains no BclIsites other than those within the cloned melR gene). The BamHI fragment carrying the shortened melR gene from pJW12[Delta]BclI was then transferred to plasmid pMAK705, which carries a temperature sensitive replication origin and resistance to chloramphenicol (11 ). After transformation of the resulting pMAK705 derivative into strain Y1089 at 30oC, chloramphenicol resistant survivors were isolated following subculture at non-permissive temperature (42oC). Some of these survivors result from integration of the pMAK705 derivative into the chromosome of Y1089 via the segments of melR flanking the internal deletion, and these were identified by their cold-sensitivity. To promote a second recombination event resulting in excision of the integrated plasmid, low temperature survivors were selected after subculture at 30oC. Derivatives in which the chromosomal melR gene had been replaced by the plasmid-borne [Delta]melR allele were recognised by their Mel- phenotype. The plasmid was then cured from one such derivative by streaking at high temperature in the absence of chloramphenicol to give the derivative strain Y1089[Delta]melR. The Mel phenotype of Y1089 and the [Delta]melR derivative was checked using Maconkey melibiose plates and by monitoring growth in media carrying melibiose as the sole carbon source. Expression from the pmelAB promoter was measured by cloning DNA fragments carrying different lengths of melR, pmelR and pmelAB into the lac expression vectors, pAA182 (5 ) or pAA224 (4 ). The resultant recombinant plasmids, carrying pmelAB-lac fusions, were transformed into Y1089 and the Y1089[Delta]melR derivative, and melibiose-induced lac expression was measured as in our previous studies (5 ). Cells were grown aerobically in minimal medium containing fructose as a carbon source and 80 [mu]g/ml ampicillin, with or without 0.05% (w/v) melibiose. Cells were harvested during exponential growth and [beta]-galactosidase activities were assayed by the Miller method (12 ).

Overexpression and purification of MelR and MelR173

To overexpress and purify full length or shortened MelR protein, we used the vector, pGEX-5X-1 from Pharmacia, designed for inducible, high level expression of proteins as fusions with Schistosoma japonicum GST (13 ). The polylinker of pGEX-5X-1 was adapted with BamHI and NcoI linkers to give pGEX-NB. For this, pGEX-5X-1 was linearised at the BamHI site and the overhanging ends generated were filled in with DNA polymerase Klenow fragment and religated with an NcoI linker to give pGEX-N. Similarly, pGEX-N was linearised at the EcoRI site, treated with Klenow enzyme and ligated with a BamHI linker to give pGEX-NB. The NcoI-BamHI fragment from pCM117-303, encoding full-length MelR (3 ), was ligated between the NcoI and BamHI sites of pGEX-NB to give pGEX-NB303 encoding a GST-MelR fusion protein. The NcoI-BamHI fragment from pCM117-173, encoding the 173 C-terminal amino acids of MelR, (3 ), was also cloned into pGEX-NB to give pGEX-NB173. Cultures of E.coli M182 cells (50 ml) carrying pGEX-NB303 or pGEX-NB173 were grown with shaking at 37oC to an OD650 of 0.7 in L broth containing 80 [mu]g/ml ampicillin. Cultures were induced by adding 50 [mu]l of 100 mM IPTG and incubated for a further 4 h. Cells were harvested by centrifugation, resuspended in 5 ml of 100 mM K2HPO4 (pH 7.4), 50 mM KCl, 10% (v/v) glycerol, 1 mM EDTA and 1 mM DTT (lysis buffer) and lysed on ice by sonication. PMSF (0.1 M in acetone) was then added to give a final concentration of 1 mM. Cell debris was removed by centrifugation and the supernatant was applied to a column containing 1 ml of glutathione-Sepharose (Pharmacia) which had been pre-equilibrated in PBS (0.14 M NaCl, 2.7 mM KCl, 10.1 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.3). The column was washed with 5 ml of PBS. Fusion protein was eluted by competition with free glutathione using 1 bed vol of 50 mM Tris-HCl (pH 8.0) containing 5 mM reduced glutathione. Products were stored at -20oC. Cleavage of the product by Factor Xa (Biolabs) was carried out at room temperature for 30 min in 20 mM Tris-HCl, 100 mM NaCl, 2 mM CaCl2 (pH 8.0) with an enzyme-substrate ratio of 1:500, according to Nagai and Thogersen (14 ). Protein concentrations were determined by the Bradford assay (15 ).

Table 2 In vivo activities of pmelAB-lacZ fusions in strains Y1089 and Y1089[Delta]melR
Plasmid

 [beta]-galactosidase activity (Miller units)
 

 Y1089

 Y1089[Delta]melR
 

 minus

plus

 fold

 minus

plus

 fold
 

 melibiose

 melibiose

 induction

 melibiose

 melibiose

 induction
pAA182

 115

 98

 0.75

 131

 124

 0.94
KK3/182

 125

 1699

 13.5

 130

 2252

 17.0
KK3NPA/182

 110

 1095

 9.9

 98

 98

 1.0
KK15/182

 148

 4021

 27.0

 191

 2476

 12.0
KK33/182

 110

 1434

 13.0

 187

 169

 0.90
pAA224

 100

 75

 0.75

 137

 131

 1.0
KK67/224

 34

 1290

 38

 31

 34

 1.1
[beta]-Galactosidase levels in two strains of E.coli containing different plasmids are shown. Assays were performed as in (5). Assays vary by <10% when repeated independently.

Preparation of crude extracts

Protein extracts were prepared as described by Michán et al. (3 ). Escherichia coli BL21([lambda]DE3) [pLysS] was transformed with pCM117-303 or pCM117-173 and grown overnight in L Broth with kanamycin (50 [mu]g/ml) and chloramphenicol (30 [mu]g/ml) at 37oC, with aeration. An aliquot of the pre-culture (50 [mu]l) was used to inoculate 5 ml of fresh medium and incubated at 37oC. When the OD650 reached 0.5, 100 [mu]M IPTG was added and incubation was continued for 1 h before the addition of rifampicin (50 [mu]g/ml). After 3 h, 3 ml of the culture was harvested by centrifugation, cells were resuspended in 300 [mu]l of lysis buffer and disrupted by sonication. An aliquot (5 [mu]l) of 0.1 M PMSF was added and the samples were centrifuged for 10 min at 4oC. Aliquots of the supernatant (100 [mu]l) were mixed with 40 [mu]l 50% (v/v) glycerol and samples were stored at -20oC. Extracts from cultures of M182 carrying pGEXNB303 and pGEXNB173 were made in a similar way, growing 5 ml cultures in L Broth containing ampicillin (80 [mu]g/ml) and inducing with 0.1 M IPTG when an OD650 of 0.7 was reached. Concentrations of protein in extracts were determined by the Bradford assay (15 ).

Gel retardation assays

For the titration experiments with a single MelR-binding site, the EcoRI-BglII fragment from KK43/121 carrying MelR-binding site 1 (Table 1 and Fig. 1 ) (16 ) was purified, labelled and used in gel retardation assays as in our previous work (3 ,16 ). For experiments with both sites 1 and 2 at pmelAB, the EcoRI-HindIII fragment from KK43/121 was used. Melibiose (5 mM) was included in all buffers. Cell extracts and end-labelled DNA fragments were incubated at room temperature for 5 min. Samples were loaded onto a 5% polyacrylamide gel and electrophoresed at 10 mA in TBE (89 mM Tris-borate, 2 mM EDTA) buffer. Bands were detected by autoradiography.

Cyclic permutation assays

The polymerase chain reaction (PCR) was used to prepare DNA fragments containing MelR binding site 1, using KK43/121 as template, 0.125 [mu]mol each of primers D9007 and D9008 (Table 1 ) and the following conditions: 33 cycles of 30 s at 94oC, followed by 15 s at 58oC and 1 min at 72oC. The products were digested by XbaI and SalI, purified by polyacrylamide gel electrophoresis and ligated between the XbaI and SalI sites of pBend2. This is a plasmid derived from pBR322 which contains two identical DNA segments with 17 restriction sites in a direct repeat on either side of XbaI and SalI cloning sites (17 ). The resulting plasmid, pBend2-XS was digested with different enzymes to yield a series of DNA fragments with the position of MelR binding site 1 permuted. These fragments were end-labelled and gel retardation assays were performed as above. To estimate the apparent bend angle ([alpha]) and to locate the centre of the bend, we used the method of Thompson and Landy (18 ). Changes in the mobility of the protein-bound DNA fragment (relative to the mobility of the free fragment) as the position of the MelR-binding site was varied, were fitted to a cosine function. The amplitude of the cosine function, Acp, is related to the bend angle by the equation Acp = 1 - cos([alpha]/2) (19 ). Values quoted are the average of six independent determinations. As a control, similar experiments were performed with a consensus CRP binding site cloned into pBend2. For this, the EcoRI-BamHI fragment from NJS1/121 (20 ) was treated with the Klenow fragment of DNA polymerase and ligated to SalI linkers (Promega). The resulting fragment was cleaved with SalIand cloned into the unique SalI site of pBend2 to give pBend2-CRP.

RESULTS

Targeted disruption of the gene encoding MelR

To investigate the function of MelR, we made a targeted deletion in the melR gene of the [Delta]lac melR+ E.coli strain Y1089 (see Materials and Methods). Whilst Y1089, the starting strain, is phenotypically Mel+ and can grow in media with melibiose as the sole carbon source, the [Delta]melR derivative is phenotypically Mel- and is unable to metabolise melibiose, confirming that the function of MelR is essential for melAB expression. The Mel- phenotype of the [Delta]melR derivative strain can, however, be complemented by transformation with plasmids carrying the KK15 fragment, that encodes MelR (see Table 1 and Table 2 ). To confirm that disruption of melR affects transcription initiation at pmelAB, the Y1089[Delta]melR derivative and the starting Y1089 strain were transformed with a series of plasmids carrying pmelAB-lacZ fusions with and without the upstream MelR gene. In the absence of melibiose there is very little expression of lacZ in either strain with any of the plasmids. In the presence of plasmids carrying the fragments KK3 and KK15 which contain the whole melR gene, both strains of E.coli show induction of pmelAB in the presence of melibiose. In contrast, in the presence of plasmids containing KK33 which does not carry the MelR gene, or KK3NPA, in which the melR gene has been disrupted, melibiose dependent induction of the promoter is only found in the melR+ background. Results similar to those for KK33/182 were obtained with KK67/224 a different pmelAB-lacZ fusion in a different vector, so the result is not dependent on the gene fusion or plasmid used. From these results, we deduce that MelR is essential for expression from pmelAB, and that the function of MelR is to act as melibiose- triggered activator of pmelAB.

Overexpression and purification of MelR as fusion protein

In our previous work, we had cloned DNA fragments encoding full-length MelR or MelR fragments into diverse expression vectors and had devised a partial purification procedure based upon chromatography on phosphocellulose followed by heparin-agarose (9 ). To obtain pure protein, a new protocol was devised, following the methods of Smith and Johnson (13 ), in which MelR and MelR derivatives were made and purified as GST fusion proteins. Briefly, DNA segments encoding either full length MelR or the 173 C-terminal amino acids (MelR173) were cloned into the vector pGEX-5X-1 encoding GST to give GST-MelR fusions under the control of the tac promoter. After expression of the fusion proteins, cell lysates were applied to a glutathione-agarose affinity matrix. Purified fusion protein was then eluted by washing the column with buffer containing glutathione and the GST `tag' was removed from the fusion protein by cleavage with factor Xa.


Figure 2. Purification of MelR and MelR173 as GST fusion proteins. The figure shows SDS polyacrylamide gels stained with Coomassie blue. (A) Purification of GST-MelR173 and (B) purification of GST-MelR. Lanes marked P show the fusion protein obtained after elution from a glutathione-Sepharose column. Lanes marked C show the product after proteolysis with factor Xa. Lanes marked M show size markers (calibration in kDa) and the lane marked G shows purified GST alone.


Figure 3. Gel retardation analysis of MelR173 and MelR binding to a single site. The figure shows autoradiograms of gel retardation assays performed with a labelled DNA fragment carrying MelR-binding site 1. The fragment was incubated with increasing amounts of MelR173 (A) or MelR (B). In lane 1 of each gel, no protein was added to the fragment. In lanes 2, 3, 4, 5, 6, 7 and 8, the volume of added protein was 4, 2, 1, 0.5, 0.2, 0.1 and 0.02 [mu]l, respectively. The migration of free fragment (F) and the different complexes (C1 and C2) are indicated by arrows.

Figure 2 shows typical results from our protocol. Although the starting lysates contained large amounts of fusion proteins, most of the material was insoluble and was lost in the pellet. This material could be solubilised by detergent but, despite several attempts, we were unable to recover this material as functional protein. In addition, some of the soluble fusion protein did not bind to the matrix, although it was active in gel retardation experiments, and a further fraction could not be eluted from the matrix. As a result, yields of final product were low with our procedure (see Materials and Methods) typically giving ~0.5 mg of 62 kDa GST-MelR fusion protein or 48 kDa GST-MelR173 fusion protein from 50 ml of cell culture. The fusion proteins were cleaved with factor Xa: efficient cleavage occurred at enzyme: substrate ratios of 1:500, although a small proportion of the GST-MelR173 fusion protein was resistant to cleavage. However, after cleavage we were unable to separate MelR or MelR173 from GST with the affinity column. Thus, the resulting unfractionated mixtures (or fusion proteins) were used in subsequent experiments. Attempts to perform the cleavage step whilst the fusion protein was bound to the column also failed: after cleavage, neither MelR nor MelR173 could be eluted from the column.

Binding of MelR to a single binding site

The binding of the GST-fusion proteins to a single MelR binding site was investigated using gel retardation assays and compared to that of MelR and MelR173. For these experiments the EcoRI-BglII DNA fragment from KK43/121, carrying only MelR-binding site 1 (Fig. 1 ), was used. In previous experiments with crude extracts from cells overexpressing MelR and MelR173, we had shown that titration of MelR173 into a fragment carrying one MelR-binding site results in the formation of a single complex (C1 in Fig. 3 A). Titration of full length MelR into this DNA fragment, results in the formation of a complex (C1 in Fig. 3 B) which is supplanted by a less mobile complex (C2) at higher MelR concentrations. The uncleaved GST-MelR173 fusion binds to the DNA fragment carrying one binding site to give a single complex (Fig. 4 A), similar to that for MelR173. In titrations of the fragment with the GST-MelR fusion, as with full length MelR, two bands are observed, the second less mobile complex being observed at higher protein concentrations (Fig. 4 B). However, with the GST-MelR fusion protein, this higher complex is less stable than with MelR. Experiments with MelR or MelR173 preparations derived from the GST fusions gave results identical to those obtained either with crude cell extracts or with MelR or MelR173 that had been partially purified by our original phosphocellulose chromatography method. This shows that the GST fusions are active and that both bands C1 and C2 arise from MelR and not from interactions of MelR with another protein.

Binding of MelR to tandem binding sites at the melAB promoter

Using gel retardation assays, we investigated the binding of MelR, MelR173, GST-MelR and GST-MelR173 to the EcoRI-HindIII fragment from KK43/121 carrying both the MelR-binding sites of pmelAB (Table 1 ). With MelR173, as found in our previous study (3 ), two retarded bands appear, with the more mobile band (C1 in Fig. 5 A) being replaced by the less mobile band (C2 in Fig. 5 A) at higher concentrations of MelR173. Similar results were obtained with GST-MelR173. With full length MelR, the situation is more complicated and, as found in our previous studies (3 ,9 ,21 ), four retarded bands are observed (C1-C4 in Fig. 5 B). Four bands are also observed with GST-MelR fusion protein.


Figure 4. Analysis of GST-MelR173 and GST-MelR binding to a single site. The figure shows autoradiograms of gel retardation assays performed with a labelled DNA fragment carrying MelR-binding site 1. The fragment was incubated with increasing amounts of GST-MelR173 (A) or GST-MelR (B). In lanes 1, 2, 3, 4, 5, 6, 7 and 8, the volume of added protein was 4, 2, 1, 0.5, 0.2, 0.1, 0.02 and zero [mu]l, respectively. The migration of free fragment (F) and the different complexes (C1 and C2) are indicated by arrows.


Figure 5. Gel retardation analysis of MelR and MelR173 binding to pmelAB. The figure shows autoradiograms of gel retardation assays performed with a labelled DNA fragment carrying both MelR-binding sites 1 and 2. The fragment was incubated with increasing amounts of MelR173 (A) or MelR (B). In lane 1 of each gel, no protein was added to the fragment. In lanes 2, 3, 4, 5, 6 and 7, the volume of added protein was 4, 2, 1, 0.5, 0.2 and 0.1 [mu]l, respectively. The migration of free fragment (F) and the different complexes are indicated by arrows.

Binding of MelR to a single binding site induces bending

To investigate whether binding of MelR induces distortion of the DNA, we employed the circular permutation assay, pioneered by Crothers and others (22 ,23 ), exploiting the vector pBend2 which had been designed to facilitate the assay (17 ). MelR-binding site 1 was amplified by PCR and cloned on an XbaI-SalI fragment into pBend2 to give pBend2-XS (Fig. 6 A). As a control, a consensus DNA site for the E.coli cyclic AMP receptor protein (CRP), a transcription factor that induces a well-characterised bend (24 ), was also cloned into pBend2.


Figure 6. Cyclic permutation assays to measure MelR-induced DNA bending. (A) Schematic representation of circularly permuted DNA fragments. The top line represents the EcoRI-HindIII fragment of pBend2-XS containing MelR binding site 1 (as a XbaI-SalI insert) flanked by two direct repeats. Cutting with the following restriction enzymes generated circularly permuted fragments each 187 bp long: MluI (a), BglII (b), NheI (c), SpeI (d), XhoI (e), PvuI (f), SmaI (g) and SspI (h). (B) Electrophoretic analysis of circularly permuted MelR173-DNA complexes. 32P-labelled DNA-fragments (a-h) were separately mixed with MelR173 (from crude extracts containing 2 mg total protein/ml) and electrophoresed through a 5% polyacrylamide gel. The locations of free (F) and bound (C) fragment are indicated. (C) Electrophoretic analysis of circularly permuted MelR-DNA complexes. 32P-labelled DNA-fragments (a-h) were separately mixed with full length MelR (from crude extracts containing 1.75 mg total protein/ml) and electrophoresed through a 5% polyacrylamide gel. The location of free (F) and bound (C1 and C2) fragment are indicated. (D) Mapping the centre of bending within MelR-binding site 1. The graph plots the relative mobility of the MelR173-DNA complex or the MelR-DNA complexes, C1 and C2, as a function of the position of the binding site (in bp, measured from the centre of the MelR binding site to one end of the fragment). (E) Base sequence around MelR-binding site 1. The positions of the centre of apparent bending are indicated by vertical arrows for the MelR173 complex and the two MelR303 complexes, C1 and C2, at binding site 1.

A series of fragments containing MelR-binding site 1 at different positions relative to the end of the fragment were isolated from pBend2-XS (Fig. 6 A). MelR or MelR173 was incubated with these fragments and the resulting complexes were analysed by electrophoresis (Fig. 6 B and C). Whilst the free DNA fragments all migrated at the same position in the gel, complexes with both MelR and MelR173 had a lower mobility when the MelR-binding site was located towards the centre of the fragment: this behaviour is suggestive of a protein-induced bend in the DNA (17 ,22 ,23 ). Although the experiments in Figure 6 were performed with all buffers and solutions containing 5 mM melibiose, identical results were obtained in the absence of melibiose: thus, binding and bending of DNA targets by MelR is not affected by the inducer.

To quantify the apparent bending angle, [alpha], induced by MelR and MelR173, we used the method of Thompson and Landy (18 ) (see Materials and Methods). For MelR173 bound at site 1, the apparent bending angle is 62o (+-7o). For full-length MelR, in the higher mobility complex at site 1 (C1 in Figs 3 B and 6 C), the apparent bending angle is the same within experimental error, 63o (+-5o). The same result was obtained with GST-MelR173. For MelR173 and GST-MelR173 at site 1 and for MelR, in the higher mobility complex at site 1, the centre of the apparent bend falls at the same position within the 18 bp MelR-binding sequence (Fig. 6 E). The analysis was repeated for the less mobile complex forming at higher MelR concentrations (C2 in Figs 3 B and 6 C). The apparent bending of the DNA in this complex is increased to 93o (+-4o) and the centre of the bend is still within the 18 bp target binding site (Fig. 6 E). As a control, we compared the DNA distortion induced by MelR in our assay with that induced by CRP (not shown). The apparent bending angle induced by CRP, deduced from our experiments performed under the same conditions, was 92o (+-5o), and accords with previous determinations (17 ).


Figure 7. MelR and MelR173 binding to a single site (A) and tandem sites at pmelAB (B). MelR subunits are shown as full ovals whilst MelR173 is shown as half ovals. In the absence of detailed information about stoichiometries, the figure is drawn to show the simplest scenario. Additional interactions between MelR subunits bound at sites 1 and 2 are probable and are likely to be facilitated by MelR-induced distortions of the DNA.

DISCUSSION

The first aim of this work was to investigate the function of MelR in vivo: this was achieved by targeting a deletion into the melR gene on the E.coli chromosome. As anticipated from our previous studies (5 ,7 ), this deletion conferred a Mel- phenotype on cells, apparently because the melAB promoter cannot be activated. The phenotype can be complemented by tranforming the strain with plasmids containing the melR gene. The second aim was to investigate some of the properties of MelR involved in the function of transcription activation. Thus, we devised a simple purification procedure involving GST-MelR fusions, and studied the binding of MelR to melAB promoter DNA. Gel retardation assays performed with pure GST-fusion proteins gave results identical to those obtained with crude cell extracts. Similarly, MelR and MelR173 derived from cleavage of pure GST-fusion proteins gave results identical to those obtained with crude extracts of cells overexpressing the proteins or from partially purified MelR or MelR173 made by phosphocellulose chromatography. This shows that all the bands observed in these studies arise from MelR and not from interactions of MelR with another component in the cell extract.

The gel retardation experiments with MelR binding to a DNA fragment containing a single MelR binding site initially show a high mobility complex C1, which we interpret as corresponding to the 1:1 complex between MelR and the 18 bp MelR-binding target carried by the fragment. This is chased into a lower mobility complex, C2, at higher protein concentrations which probably corresponds to a 2:1 complex due to oligomerisation of MelR. The absence of a second band with MelR173 and GST-MelR173 shows that this oligomerisation is dependent on the N-terminal domain of MelR. The lower stability of the C2 complex in the GST-MelR fusion is probably due to the presence of GST attached to the N-terminus of MelR interfering with the ability of MelR to form the higher order structure found in complex C2. Figure 7 A shows a schematic representation of the different complexes with MelR and MelR173 likely to form at a DNA fragment carrying a single DNA site for MelR. MelR and MelR173 give identical DNAse I footprints at this site, suggesting that only one MelR subunit makes direct contact with the DNA (3 ). With fragments of DNA containing two MelR binding sites, MelR173 and GST-MelR173 give two complexes in gel retardation assays. We interpret complex C1 as being due to occupation of one MelR-binding site, whilst complex C2 is due to occupation of both MelR-binding sites. Full length MelR and GST-MelR give four retarded bands. Figure 7 B illustrates a possible explanation for these complexes: as with the fragment carrying a single MelR-binding site, the complexity is probably due to oligomerisation of free MelR subunits with subunits bound at either site 1 or site 2.

The cyclic permutation assay with MelR and MelR173 at a single site suggests that MelR distorts target DNA, and that the bend is dependent on the C-terminal domain of the protein but independent of the N-terminal. In contrast, oligomerisation of MelR at a single DNA-binding site is dependent on the N-terminal domain. We suppose that the dimers forming at the adjacent MelR binding sites must also interact in some way, and, we suggest that the combination of MelR-induced DNA bending and N-terminal domain-dependent oligomerisation leads to the formation of a large nucleoprotein complex that plays an essential role in the process of transcription activation. The current challenge is to understand the organisation of this complex, to discover how MelR subunits interact with the different RNA polymerase subunits, and to understand the role of melibiose in triggering the activation of the melAB promoter.

Most promoters activated by members of the AraC/XylS family of transcription factors depend on binding of the factor to two separated operator sites and are likely to require interactions between bound subunits. To our knowledge, this is the first report of DNA bending induced by a member of the AraC/XylS family, and our conclusions from this study with MelR are likely to be applicable to other family members. Most text books draw target promoters as a linear array of recognition elements for transcription factors and RNA polymerase, but it is certain that the final complexes are far from linear, with the promoter DNA being highly distorted (reviewed in ref. 25 ). Similarly, most text books regard transcription activators as proteins that interact with target promoters and then make some sort of simple contact with the transcription machinery. We suspect that, perhaps in the majority of cases, complex protein-protein and protein-DNA interactions play crucial roles in transcription activation, as in the case of the E.coli MelR protein.

ACKNOWLEDGEMENTS

We are grateful to Christine Webster and Jenny Keen for help throughout the study. This work was funded by the UK BBSRC with grant number G04260.

REFERENCES

1 Henikoff, S., Wallace, J. and Brown, J. (1990) Methods Enzymol. 183, 111-130. MEDLINE Abstract

2 Gallegos, M.T., Michán, C. and Ramos, J. (1993) Nucleic Acids Res. 21, 807-810. MEDLINE Abstract

3 Michán, C., Busby, S.J.W. and Hyde, E.I. (1995) Nucleic Acids Res. 23, 1518-1523. MEDLINE Abstract

4 Keen, J., Williams, J. and Busby, S. (1996) Biochem. J. 318, 443-449. MEDLINE Abstract

5 Webster, C., Kempsell, K., Booth, I. and Busby, S. (1987) Gene 59, 253-263. MEDLINE Abstract

6 Hanatani, M., Yazyu, H., Shiotaniiya, S., Moriyama, Y., Kanazawa, H., Futai, M. and Tsuchiya, T. (1984) J. Biol. Chem. 259, 1807-1812. MEDLINE Abstract

7 Webster, C., Gardner, L. and Busby, S. (1989) Gene 83, 207-213. MEDLINE Abstract

8 Lobell, R. and Schleif, R. (1990) Science 250, 528-532. MEDLINE Abstract

9 Caswell, R., Williams, J., Lyddiatt, A. and Busby, S. (1992) Biochem. J. 287, 493-499. MEDLINE Abstract

10 Maniatis, T., Fritsch, E. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

11 Hamilton, C., Aldea, M., Washburn, B., Babitzke, P. and Kushner, S. (1989) J. Bacteriol. 171, 4617-4622. MEDLINE Abstract

12 Miller, J. (1972) in Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

13 Smith, D. and Johnson K. (1988) Gene 67, 31-40. MEDLINE Abstract

14 Nagai, K. and Thogersen, H.C. (1984) Nature309, 810-812. MEDLINE Abstract

15 Bradford, M. (1976) Anal. Biochem. 72, 248-254. MEDLINE Abstract

16 Williams, J., Michán, C., Webster, C. and Busby, S. (1994) Biochem. J. 300, 757-763. MEDLINE Abstract

17 Kim, J., Zwieb, C., Wu, C. and Adhya, S. (1989) Gene 85, 15-23. MEDLINE Abstract

18 Thompson, T. and Landy, A. (1988) Nucleic Acid Res. 16, 9687-9705.

19 Kerppola, T. and Curran, T. (1993) Nucleic Acids Mol. Biol. 7, 70-105.

20 Savery, N.J., Rhodius, V.A., Wing, H.J. and Busby, S.J.W. (1995) Biochem. J. 309, 77-83. MEDLINE Abstract

21 Caswell, R., Webster, C. and Busby, S. (1992) Biochem. J. 287, 501-508. MEDLINE Abstract

22 Wu, H.M. and Crothers, D.M. (1984) Nature 308, 509-513. MEDLINE Abstract

23 Prentki, P., Pham, M.-H. and Galas, D.J. (1987) Nucleic Acids Res. 15, 10060. MEDLINE Abstract

24 Schultz, S., Shields, G. and Steitz, T. (1991) Science 253, 1001-1007. MEDLINE Abstract

25 Perez-Martin, J., Rojo, F. and de Lorenzo, V. (1994) Microbiol. Rev. 58, 268-290. MEDLINE Abstract

26 Grodberg, J. and Dunn, J. (1988) J. Bacteriol. 170, 1245-1253. MEDLINE Abstract

27 Casadaban, M. and Cohen, S. (1980) J. Mol. Biol. 138, 179-207. MEDLINE Abstract

28 Huynh, T., Young, R. and Davis, R. (1983) in Glover, D. (ed.) DNA Cloning: A Practical Approach. Volume I, chapter 2, IRL Press, pp 49-80.

29 Kelsall, A., Evans, C. and Busby, S. (1985) FEBS Lett. 180, 155-159.

30 Studier, F.W. (1991) J. Mol. Biol. 219, 37-44. MEDLINE Abstract

Return

*To whom correspondence should be addressed. Tel: +44 121 414 5439; Fax: +44 121 414 7366; Email: s.j.w.busby@bham.ac.uk
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nucleic Acids ResHome page
S. Samarasinghe, M. S. El-Robh, D. C. Grainger, W. Zhang, P. Soultanas, and S. J. W. Busby
Autoregulation of the Escherichia coli melR promoter: repression involves four molecules of MelR
Nucleic Acids Res., May 1, 2008; 36(8): 2667 - 2676.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
A. Kolin, V. Jevtic, L. Swint-Kruse, and S. M. Egan
Linker Regions of the RhaS and RhaR Proteins
J. Bacteriol., January 1, 2007; 189(1): 269 - 271.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
L. Pérez-Lago, M. Salas, and A. Camacho
A precise DNA bend angle is essential for the function of the phage {phi}29 transcriptional regulator
Nucleic Acids Res., January 7, 2005; 33(1): 126 - 134.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
J. A. Ibarra, M. I. Villalba, and J. L. Puente
Identification of the DNA Binding Sites of PerA, the Transcriptional Activator of the bfp and per Operons in Enteropathogenic Escherichia coli
J. Bacteriol., May 1, 2003; 185(9): 2835 - 2847.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
V. J. Howard, T. A. Belyaeva, S. J. W. Busby, and E. I. Hyde
DNA binding of the transcription activator protein MelR from Escherichia coli and its C-terminal domain
Nucleic Acids Res., June 15, 2002; 30(12): 2692 - 2700.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
M. E. Porter and C. J. Dorman
In Vivo DNA-Binding and Oligomerization Properties of the Shigella flexneri AraC-Like Transcriptional Regulator VirF as Identified by Random and Site-Specific Mutagenesis
J. Bacteriol., January 15, 2002; 184(2): 531 - 539.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
P. M. Bhende and S. M. Egan
Genetic Evidence that Transcription Activation by RhaS Involves Specific Amino Acid Contacts with Sigma 70
J. Bacteriol., September 1, 2000; 182(17): 4959 - 4969.
[Abstract] [Full Text]

Home page
 J. Bacteriol.Home page
P. M. Bhende and S. M. Egan
Amino Acid-DNA Contacts by RhaS: an AraC Family Transcription Activator
J. Bacteriol., September 1, 1999; 181(17): 5185 - 5192.
[Abstract] [Full Text]

Home page
 Microbiol. Mol. Biol. Rev.Home page
M. K. B. Berlyn
Linkage Map of Escherichia coli K-12, Edition 10: The Traditional Map
Microbiol. Mol. Biol. Rev., September 1, 1998; 62(3): 814 - 984.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
E. Tamai, T. A. Belyaeva, S. J. W. Busby, and T. Tsuchiya
Mutations That Increase the Activity of the Promoter of the Escherichia coli Melibiose Operon Improve the Binding of MelR, a Transcription Activator Triggered by Melibiose
J. Biol. Chem., May 26, 2000; 275(22): 17058 - 17063.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (190K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (19)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Bourgerie, S. J.
Right arrow Articles by Hyde, E. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bourgerie, S. J.
Right arrow Articles by Hyde, E. I.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?