JBC PeproTech; Our Business is Cytokines!

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M414544200 on February 11, 2005

J. Biol. Chem., Vol. 280, Issue 15, 15111-15121, April 15, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/15/15111    most recent
M414544200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, C.-C.
Right arrow Articles by Wu, H.-Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, C.-C.
Right arrow Articles by Wu, H.-Y.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

LeuO Protein Delimits the Transcriptionally Active and Repressive Domains on the Bacterial Chromosome*

Chien-Chung Chen and Hai-Young Wu{ddagger}

From the Department of Pharmacology, School of Medicine, Wayne State University, Detroit, Michigan 48201

Received for publication, December 27, 2004 , and in revised form, February 3, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
LeuO protein relieves bacterial gene silencer AT8-mediated transcriptional repression as part of a promoter relay mechanism found in the ilvIH-leuO-leuABCD gene cluster. The gene silencing activity has recently been characterized as a nucleoprotein filament initiated at the gene silencer. In this gene locus, the nucleoprotein filament cis-spreads toward the target leuO promoter and results in the repression of the leuO gene. Although the cis-spreading nature of the transcriptionally repressive nucleoprotein filament has been revealed, the mechanism underlying LeuO-mediated gene silencing relief remains unknown. We have demonstrated here that LeuO functions analogously to the eukaryotic boundary element that delimits the transcriptionally active and repressive domains on the chromosome by blocking the cis-spreading pathway of the transcriptionally repressive heterochromatin. Given that one LeuO-binding site is positioned between the gene silencer and the target promoter, the simultaneous presence of a second LeuO-binding site synergistically enhances the blockade, resulting in a cooperative increase in LeuO-mediated gene silencing relief. A known DNA loop-forming protein, the lac repressor (LacI), was used to confirm that cooperative protein binding via DNA looping is responsible for the blocking synergy. Indeed, a distal LeuO site located downstream cooperates with the LeuO sites located upstream of the leuO gene, resulting in synergistic relief for the repressed leuO gene via looping out the intervening DNA between LeuO sites in the ilvIH-leuO-leuABCD gene cluster.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Gene silencing is integral to the regulation of prokaryotic and eukaryotic gene expression (1–3). In bacteria, histone-like nucleoid structuring protein (H-NS)1 was found to be responsible for the silencing of the proU gene and the {beta}-glucosidase gene (bgl) and recently for the repression of the leuO gene (4–12). Although a cis-spreading nucleoprotein filament initiated at an H-NS nucleation site is clearly involved in the aforementioned bacterial gene silencing cases, the modulation (control mechanism) for the distribution of such cis-spreading transcriptionally repressive nucleoprotein filaments on the bacterial chromosome remains unknown.

Apparently, H-NS is not the only nucleoprotein responsible for bacterial gene silencing. Centromere-like proteins such as SopB and ParB also cause repression of genes within the proximity of their binding sites, sopC and parS, respectively (3, 13–17). The common features found in these bacterial gene silencers have led to a nucleoprotein filament model that describes the formation of a cis-spreading nucleoprotein structure initiated at a nucleation site (3, 11). The proposed mechanism is similar to the gene silencing that is exerted by the eukaryotic heterochromatin and is supported by the evidence that a cis-spreading nucleoprotein filament is capable of silencing genes within its proximity (the effect ranges from several hundred base pairs to several thousand base pairs). An alternative sequestration model involving multiple protein-binding sites and a patch of DNA-binding domains localized at a specific cellular location was proposed in a study using a fusion protein consisting of the N-terminal 82 amino acids of SopB protein (for cellular localization) and the DNA-binding domain of the yeast GAL4 gene product (for binding with multiple binding sites on DNA) (13). Obviously, the exact nature of the various bacterial gene silencing mechanisms remains to be a matter of debate. The various nucleoprotein structures involved in bacterial gene silencing are, however, reminiscent of the gene silencing apparatus found in eukaryotes. A number of nucleoprotein structures such as centromeres (18), telomeres (19, 20), and heterochromatin (21) were found to be transcriptionally repressive, whereas each of the nucleoprotein structures also possesses other cellular functions (22). Despite that protein factors involved in the various gene silencing mechanisms are different, a common transcriptionally repressive molecular feature is likely conserved between these gene silencers found in both prokaryotes and eukaryotes. Hence, the control mechanism that delimits the distribution of the cis-spreading nucleoprotein structure on either the prokaryotic or eukaryotic chromosome may have mechanistic similarities.

The expression of eukaryotic genes in a gene locus is spatially controlled by either limiting a transcriptionally repressive nucleoprotein filament (e.g. heterochromatin) within the locus or by excluding the neighboring transcriptionally repressive nucleoprotein structure from entering this region (23, 24). Barrier elements/insulators that block the transcriptionally repressive heterochromatin have been found in the regulation of the expression of many genes in eukaryotic cells (25–27). Both heterochromatin and the "barriers" that define the boundaries between the transcriptionally active euchromatin and the transcriptionally inert heterochromatin are gene-nonspecific. These "elements" are responsible for many interesting position effects found in the regulation of eukaryotic genes (25, 28, 29). For example, the insulator HS4 located at the junction of active chromatin and heterochromatin is important for the regulation of the chicken {beta}-globin gene (25, 28). Boundary elements are important for the proper regulation of genes in yeast matingtype loci (26, 30). Furthermore, STAR (subtelomeric anti-silencing region) was found to block the spreading of the silencing effect mediated by the yeast telomere (27), and many other gene silencing-blocking mechanisms have been found (29). It appears that protein factors that are capable of segregating a domain may regulate gene expression by blocking/insulating the effect of a nearby gene silencing apparatus such as heterochromatin (31). Even the transcription complex at the promoter of the yeast tRNA gene was shown to be a barrier that blocks the spread of heterochromatin (32). Thus, "blocking" is clearly a theme in this type of transcriptional regulation, whereas other regional chromatin-modifying mechanisms (e.g. recruitment of histone acetyltransferase or inhibition of histone deacetyltransferase activities) may be involved in the subsequent alterations of the local chromatin structure (26). However, the mechanisms underlying this "nucleoprotein structure blocking" are largely unclear.

In a bacterial transcriptional regulation model system, we found that LeuO protein negates a gene silencing activity, which was recently demonstrated to be a cis-spreading transcriptionally repressive nucleoprotein filament consisting of H-NS (12). It appears that the novel LeuO protein modulates the gene silencing function of the nucleoprotein filament in the bacterial gene systems by blocking the cis-spreading of the nucleoprotein structure. The relatively simple bacterial model system provided clues for studying the currently incomplete molecular details that are responsible for nucleoprotein filament blocking. Indeed, we found that, in the presence of a LeuO-binding site positioned between the gene silencer and the target promoter, the gene silencing activity can be synergistically relieved by LeuO protein when a second LeuO-binding site is simultaneously present on the DNA molecule. The requirement of LeuO protein positioning between the silencer and the target promoter for LeuO-mediated gene silencing relief suggested that LeuO functions analogously to a eukaryotic boundary element that delimits the chromosomal domains.

The requirement of the simultaneous presence of two LeuO-binding sites on one DNA molecule for a clear boundary element activity (polarity of LeuO-mediated gene silencing relief) is a novel finding. Using a well known DNA loop-forming protein factor (LacI), we demonstrate that cooperative protein binding (protein-protein interaction via DNA looping) is responsible for the synergistic blockade of cis-spreading of the transcriptionally repressive nucleoprotein structure (the novel property of the bacterial boundary element). Apparently, LacI is capable of replacing LeuO protein to provide a similar blocking synergy for regulating the gene silencing effect despite the fact that LacI is not related to this particular gene silencing regulation. The finding is reminiscent of the potential involvement of nonspecific protein with DNA loop-forming capability in heterochromatin blocking found in eukaryotes (31). We propose a dynamic DNA-looping mechanism whereby the transcriptionally repressive and active nucleoprotein domains on the bacterial chromosome are cooperatively modulated by the boundary element activity of various DNA loop-forming proteins, including LeuO protein.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Plasmids and Bacterial Strains—The plasmids used in this study were derived from pAO, pCH301, and pWU802, which have been described previously (33, 34). DNA inserts (AT4, AT7, and AT8) were synthesized and annealed as duplex DNAs for insertion at the unique AatII site of pAO or pCH301. The DNA sequences of these DNAs have been reported previously (33). For cloning purposes, all DNA inserts contained the 4-bp protruding DNA 3'-sequence complementary to the 4-bp DNA overhang sequence generated by digestion with the AatII restriction enzyme. Each plasmid construct was verified at the DNA sequence level.

Plasmids pWU802-LA and pWU802-LH have been described previously (12). The other pWU802-based plasmids, pWU802-LAH, pWU802-LHS, pWU802-(LA)2, and pWU802-(LH)2, were also derived from pWU802 by inserting two copies of the 27-bp lacO DNA sequences at various sites on pWU802 (illustrated in Fig. 4). pWU802(AT7) was constructed by inserting the 25-bp AT7 DNA (5'-CACAATCATACACCAAGTGAATGAT-3') at the unique AatII site of pWU802.



View larger version (50K):
[in this window]
[in a new window]
  		FIG. 4.
The ternary LacI-lacO complex functions as a barrier that blocks the cis-spreading pathway of the gene silencer AT8-mediated transcriptionally repressive nucleoprotein filament in a synergistic manner. Various test plasmids derived from pWU802 were individually assayed in MC1060, an E. coli LacI deletion strain. Exogenous wild-type tetrameric LacI or dimeric active LacI was provided in trans from a pACYC-based plasmid. The sites of the inserted 27-bp lac operators are shown with the names of the restriction sites. The distance between restriction sites and the total length of pWU802 are indicated. Primer extension reactions were used for simultaneously detecting the activities of the bla (pbla) and leuO (pleuO) promoters. The quantified (Quant.) data shown at the bottom of each lane for three repeated experiments are expressed as the mean within the range of ±0.02 S.D. The activity of the leuO promoter served as an internal control throughout the experiment because the promoter of the leuO gene was not segregated from the gene silencer by LacI under the test conditions. pleu-500, leu-500 promoter.

 
Gene silencing activity was assayed in Escherichia coli leuO- strain MF1 (34) harboring one of the test plasmids and the pACYC-based LeuO overexpression vector pEV101 (35), which was used to provide LeuO in trans for testing LeuO-mediated gene silencing relief. As demonstrated previously (34), 50 µM isopropyl 1-thio-{beta}-D-galactopyranoside induction was used to provide the optimum cellular LeuO concentration for LeuO-mediated derepression as needed. E. coli MC1060, a LacI deletion strain (36), was used for assaying LacI-mediated gene silencing relief as described previously (12). The pACYC-based pSO1000 or pSO100 plasmid (37) was used for providing wild-type tetrameric LacI or dimeric active LacI, R3 (38), respectively. Bacteria were grown in LB medium at 37 °C with aeration. 50 µg/ml ampicillin and 6.25 µg/ml tetracycline were added as needed.

Primer Extension—Primer extension was carried out as described previously (39) with the following modifications. A section of DNA sequence in the coding region of the bla gene (5'-TCTGGGTGAGACAAAACAGGAAGGC-3') was used as the primer for detecting bla promoter-mediated transcripts. Another DNA oligomer (5'-CATCACCATCTAATTCAACAAGAATTGGG-3') consisting of a section of DNA sequence in the coding region of a green fluorescent gene (gfpuv) was used as the primer for detecting syn promoter-mediated transcripts. A section of plasmid DNA sequence located downstream of the leuO promoter (5'-CGGAAAACATAAAGACGCTGACAGAGAC-3') was used as the primer for detecting leuO promoter-mediated transcripts. These primers hybridize only with plasmid DNA sequences, so no transcripts from the chromosomal genes would interfere with the primer extension results. The primers were radioactively labeled at their 5'-ends using [{gamma}-32P]ATP and T4 polynucleotide kinase. The bla promoter primer was mixed with 100 µg of total RNAs in the primer extension reactions shown in Figs. 1 and 2. One of two primer sets (bla-syn promoters and bla-leuO promoters) was mixed with 100 µg of total RNAs for simultaneously detecting two promoter activities reported in Figs. 3, 4, and 8. The initiation site of RNA transcription was verified based on the specific size of the primer extension DNA products that ran off the 5'-end of the RNA transcripts. A DNA sequence ladder was prepared using the corresponding primer to verify the initiation sites at the DNA sequence level. Radioactivity incorporated in the primer extension DNA product was visualized and quantified using a Storm 840 imaging system (Amersham Biosciences). Quantification was normalized against the amount of plasmid DNA. The plasmid DNA amount in each experiment was quantified by Southern blotting as described previously (33) using a 32P-labeled probe that specifically hybridizes with the test plasmid DNA.



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 1.
Polarity of LeuO-mediated transcriptional derepression. The dimeric and monomeric test plasmids pAO-AT4R and pAO-AT4, which contain the 72-bp AT4 DNA insert in either orientation, were assayed for LeuO-mediated transcriptional derepression in MF1, a leuO- strain. AT4 DNA consists of the 25-bp LeuO-binding site AT7 (black boxes) and the 47-bp gene silencer AT8 (white boxes). A coexisting pACYC-based plasmid, pEV101, provided exogenous LeuO for testing. 50 µM isopropyl 1-thio-{beta}-D-galactopyranoside treatment was used to provided LeuO protein (+). The absence of LeuO protein is indicated (-). The transcriptional activities of the promoter of the bla gene (pbla) in the test plasmids were detected using primer extension reactions and are shown as bands marked pbla. The DNA sequence ladder is included to check the transcriptional initiation site of the bla transcript. The quantified (Quant.) bla promoter activity shown at the bottom of each lane for three repeated experiments is expressed as the mean within the range of ±0.01 S.D. The -fold increase in the effect of the LeuO-mediated gene silencing relief of each comparison pair is also indicated.

 



View larger version (50K):
[in this window]
[in a new window]
  		FIG. 2.
The polarity of LeuO-mediated gene silencing relief is dependent on the copy number of LeuO-binding sites and the relative positions of the gene silencer, the LeuO-binding site, and the target promoter on the test plasmid. The black boxes are the 25-bp LeuO-binding site AT7, and the white boxes are the 47-bp gene silencer AT8. The orientations of the AT7 and AT8 DNAs are indicated by the black and white arrows, respectively. The arrows pointing from left to right are the original orientations (5' to 3') of the two DNA elements. Various combinations of the two DNA elements as illustrated were inserted at the AatII site of monomeric pAO (see map in Fig. 1). Primer extension was employed to detect the activity of the bla promoter (pbla). The quantified (Quant.) bla promoter activity shown at the bottom of each lane for three repeated experiments is expressed as the mean within the range of ±0.01 S.D.

 



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 3.
Boundary element-like activity of LeuO protein. The black boxes are 25-bp AT7, and the white boxes are 47-bp AT8. Various combinations of AT7 and AT8 were inserted at the AatII site flanked by a pair of divergently arrayed bla and gfpuv genes on pCH301, as illustrated. The distances between the AatII site and the +1-positions of the two flanking transcription units, bla and gfpuv, are 97 and 124 bp, respectively. The activities of the promoters of the flanking genes (bla (pbla) and syn (psyn) promoters, respectively) were simultaneously detected using primer extension. +LeuO, LeuO protein was provided in trans from the coexisting plasmid pEV101; -LeuO, absence of LeuO protein in the tests. The quantified (Quant.) data shown at the bottom of each lane for three repeated experiments are expressed as the mean within the range of ±0.01 S.D. The derepression effects (-fold increase in LeuO-mediated gene silencing relief) on the bla and syn promoters are compared between the promoter activities detected in the indicated lanes.

 



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 8.
A downstream LeuO-binding site cooperates with the LeuO-binding sites located upstream of the leuO gene for synergistic relief of the repressed leuO gene. The linear map of pWU802 illustrates the positions of the two LeuO-binding sites, AT3 and AT7, and the AT8 gene silencer in LCR-I (the upstream regulatory sequence of the leuO gene). To mimic the presence of a downstream LeuO-binding site for the regulation of the leuO gene, a well defined LeuO-binding site, the 25-bp AT7 DNA, was inserted at the AatII site, which is located downstream of the leuO promoter (pleuO) in the test plasmid pWU802. The distances between the centers of the various LeuO sites are indicated. The test plasmids were individually assayed in strain MF1 in the presence of pEV101 to provide exogenous LeuO in trans upon isopropyl 1-thio-{beta}-D-galactopyranoside induction. +, presence of exogenous LeuO protein; -, absence of LeuO protein. The activities of both the leuO and bla (pbla) promoters in the test plasmids were simultaneously monitored in the primer extension studies. The quantified (Quant.) data shown at the bottom of each lane for three repeated experiments are expressed as the mean within the range of ±0.04 S.D. pleuABCD, leuABCD promoter.

 
Electrophoretic Mobility Shift Assay (EMSA)—All test DNA segments were PCR products obtained using primers that were 5'-end-labeled with [{gamma}-32P]ATP. Those DNA segments that showed affinity to LeuO protein were further digested with restriction enzymes, as illustrated in the figure legends, to focus the LeuO affinity to a smaller region. The radioactively labeled DNAs (30 pg/reaction) were mixed with purified His10-tagged Salmonella typhimurium LeuO in binding buffer (40 mM Tris-Cl (pH 8.0), 4 mM MgCl2, 70 mM KCl, 0.1 mM EDTA, and 0.1 mM dithiothreitol). The preparation of His10-tagged S. typhimurium LeuO has been described previously (33). The binding reaction mixtures were incubated at 37 °C for 30 min to reach a protein binding equilibrium. The mixtures were then electrophoresed on native polyacrylamide gels in 89 mM Tris borate (pH 8.3) and 1 mM EDTA. Either 4 or 5% polyacrylamide was used depending on the size of the DNA being tested.

DNase I Footprinting Assay—A locus control region (LCR) I DNA segment consisting of a 292-bp DNA sequence that includes the promoter and upstream regulatory region (-255 to +36) of the S. typhimurium leuO gene was generated by PCR using primer set 5'-TCATTGTTTGAATTCTTTAGGCATTTTTG-3' and 5'-TCCTGAGTCACACCATTG-3'. An LCR-II DNA segment consisting of a 289-bp DNA sequence that includes the promoter and upstream regulatory region (-255 to +33) of the S. typhimurium leuO gene was generated by PCR using primer set 5'-CTGGTTTATTCTG-3' and 5'-GCTTGCTCCACTTTATAC-3'. The PCR products were unique end-labeled using one 32P-5'-end-labeled primer of the primer sets used in each PCR. The exact position of the labeled end is indicated at the bottom of each DNase I footprinting result. The entire S. typhimurium DNA sequence relevant to the promoter relay mechanism is available in the GenBankTM Data Bank under accession number AF106956 [GenBank] .

Reactions were carried out in a total volume of 50 µl. 1 ng of radio-actively labeled DNA (~20,000 cpm/reaction) was mixed with increasing amounts of purified His10-tagged S. typhimurium LeuO in binding buffer. The binding reactions were incubated at room temperature for 30 min to reach protein binding equilibrium. Subsequently, 5 µlof1mM MgCl2 and 0.5 mM CaCl2 were added to the reactions. 1 min later, DNase I (0.4 units/reaction) was added to the reactions for a 90-s incubation at room temperature. The entire reaction was then halted by adding 140 µl of stop solution (192 mM sodium acetate, 32 mM EDTA, 0.14% SDS, and 64 µg/ml yeast RNA). The samples were phenol-extracted, ethanol-precipitated, and analyzed on 7 M urea and 7% polyacrylamide denaturing gels. DNA sequence ladders were prepared using dideoxy-DNA sequencing reactions (Sequenase DNA sequencing kit, United States Biochemical Corp., Cleveland, OH) with one of the primers. Position markers for marking restriction sites of the test DNA templates were prepared by digesting the unique end-labeled DNA with appropriate restriction enzymes as indicated in the figure legends.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
A "Boundary Element-like" Activity of LeuO Protein Is Responsible for LeuO-mediated Gene Silencing Relief—A 72-bp DNA element (AT4) was identified as one of the cis-acting elements important for coordinating the expression of genes in the S. typhimurium ilvIH-leuO-leuABCD gene cluster (34, 35). As a trans-acting protein factor important for gene expression coordination (35), LeuO protein, the gene product of leuO, is capable of negating AT4-mediated transcriptional repression (34). A subsequent study (33) has successfully divided the 72-bp AT4 DNA into two elements: a 47-bp gene silencer (AT8) and a 25-bp derepression element (AT7), which is a LeuO-binding site (collectively, the AT4 element illustrated in Fig. 1). A recent study revealed that the bacterial gene silencer AT8 is an H-NS nucleation site. A cis-spreading nucleoprotein structure was found to be responsible for AT8-mediated gene silencing (12). Upon providing LeuO protein in trans, a LeuO-binding site located within the proximity of the AT8 gene silencer is required for LeuO-mediated relief of the gene silencing activity (33). However, the molecular mechanism whereby LeuO protein relieves AT8-mediated gene silencing remains unclear.

Surrounded by the foreign plasmid DNA context, gene silencer AT8-mediated gene repression activity is bidirectional on a pAO-based test plasmid (33). However, in its natural chromosomal DNA environment (AT-rich LCR-I of the ilvIH-leuO-leuABCD gene cluster), the AT8-mediated transcriptionally repressive nucleoprotein structure was shown to cis-spread preferentially toward the target leuO promoter (12). In this natural setting, the LeuO-binding site AT7 is located between the AT8 gene silencer and the target leuO promoter. Whether or not the relative positions of these three involved components (AT8, AT7, and leuO promoter) are important for the LeuO-mediated gene silencing relief was uncertain and was thus tested using pAO-based plasmids (Fig. 1).

Indeed, we found a polarized effect for LeuO-mediated transcriptional derepression (Fig. 1, lanes 1–4). When the LeuO-binding site AT7 was located between the target bla promoter and the gene silencer, we observed a significant (11-fold) derepression effect on the bla promoter upon providing LeuO in trans (pAO-AT4 plasmid map) (Fig. 1, compare lanes 3 and 4). Upon inversion of the AT4 element, the LeuO-binding site AT7 was no longer located between the bla promoter and the gene silencer on the pAO-AT4R test plasmid. Consequently, LeuO-mediated gene silencing relief on the repressed bla promoter was reduced (pAO-AT4R plasmid map) (Fig. 1, compare lanes 1 and 2). Clearly, under this test condition, LeuO-mediated transcriptional derepression is highly polarized, and the efficiency of LeuO-mediated gene silencing relief is determined by the relative positions of the LeuO-binding site, gene silencer, and target promoter.

However, the striking polarization of the transcriptional derepression mediated by LeuO protein is true only if dimeric DNA templates are used in the study. A small difference in LeuO-mediated derepression between the AT4 and AT4R settings was observed when monomeric DNA templates were used in a similar set of experiments (Fig. 1, lanes 6–9). This rather puzzling result (Fig. 1) provided the following clues in investigating the mechanism underlying LeuO-mediated transcriptional derepression. The major differences between the dimeric and monomeric DNA templates (Fig. 1) were the copy number of the repression and derepression elements per DNA molecule and the size of the plasmid. Additionally, the orientations of the repression and derepression elements were altered in the AT4R versus AT4 setting.

To determine which of the above factors were responsible for the polarization of LeuO-mediated transcriptional derepression that occurred on the dimeric DNA templates (Fig. 1), we first tested all possible combinations of single units of both repression and derepression elements with either orientation for their effects on LeuO-mediated derepression using a monomeric pAO DNA template (Fig. 2A). Once again, there was no significant difference between combinations. However, we observed an overall trend that LeuO-mediated gene silencing relief was relatively significant when the derepression element (LeuO-binding site) was located between the target promoter and the gene silencer regardless of the orientation of the elements in each combination (Fig. 2A, compare lanes 2–5 and lanes 6–9).

Because the copy number of the involved elements is also one of the potential determining factors, we further tested all possible combinations of (AT8)2, a tandem repeat of the repression element AT8, and (AT7)2, a tandem repeat of the derepression element AT7, for their effects on LeuO-mediated gene silencing relief (Fig. 2B). Because the effect of the orientation of each element on LeuO-mediated derepression had been ruled out (Fig. 2A), only one orientation of the elements was used in testing the combinations of (AT8)2 and (AT7)2. Strikingly, the combination of (AT8)2 and (AT7)2 exerted a significant polarization (position effect) on LeuO-mediated derepression. Given that LeuO protein was provided in trans, (AT7)2 located between the target promoter and (AT8)2 supported significant transcriptional derepression (Fig. 2B, lane 3). In contrast, LeuO-mediated gene silencing relief was essentially undetectable when (AT7)2 was not located between the target promoter and (AT8)2 (Fig. 2B, lane 5). A partial derepression was observed when (AT7)2 was located between the two gene silencers (Fig. 2B, lane 4). Most interestingly, enhanced LeuO-mediated derepression was observed when (AT8)2 was flanked by an AT7 pair (Fig. 2B, lane 6). Altogether, the data support the conclusion that, although the presence of two copies of the derepression element on one DNA molecule is important for the synergy (cooperative increase) of LeuO-mediated derepression, LeuO-mediated gene silencing relief is achieved only if at least one of the two derepression element copies is positioned between the gene silencer and the target promoter. The specific position requirement suggests that LeuO may negate gene silencing activity via blocking the cis-spreading pathway of the transcriptionally repressive nucleoprotein structure mediated by gene silencer AT8 (12). If so, LeuO protein may exert a boundary element-like activity in bacterial transcriptional regulation.

LeuO Protein Functions as a Barrier, Which Is a Transcriptionally Inert Element—The possible boundary element-like activity of LeuO protein was further tested using various combinations of the AT8 repression element and the AT7 derepression element, which are flanked by the promoters (bla and syn) of two divergently transcribed genes (plasmid map illustrated in Fig. 3). The activities of the flanking promoters were simultaneously monitored in a primer extension study. Upon providing LeuO in trans, the (AT8)2-mediated gene silencing effect on the flanking promoters was relieved only if a LeuO-binding site was positioned between the gene silencer(s) and one of the two promoters (the opposite trends of the derepression effects on the syn and bla promoters shown in the quantified results of Fig. 3). This effect is also evidenced by the fact that both flanking promoters were partially released from the gene silencing activity when a LeuO-binding site (AT7) was located between the two gene silencers (Fig. 3, lane 6). The repression of both flanking promoters was simultaneously relieved when the gene silencers were flanked by a pair of LeuO-binding sites (Fig. 3, lane 8). These are LeuO protein-dependent relievers of gene silencing because the silencing states of both promoters were not affected if LeuO protein was not provided in trans (Fig. 3, lanes 1–4).

Fig. 3 clearly shows that, although one of the flanking promoters was released from the gene silencing activity due to the presence of the LeuO-binding site AT7 between the silencer(s) and the particular promoter, the gene silencers remained active and were capable of repressing the activity of another flanking promoter that was not protected (blocked) by the derepression element (AT7). This is a strong piece of evidence that LeuO protein does not directly affect the gene silencing activity mediated by the (AT8)2 repression element (e.g. disruption of the nucleoprotein structure). Instead, the binding of LeuO to its binding site (AT7) must merely block the cis-spreading pathway of the nucleoprotein structure, so the transcriptionally repressive nucleoprotein filament is not able to reach the promoter, which is protected by the LeuO-binding site AT7, whereas the nucleoprotein filament maintains its ability to reach and thus repress the promoter located on the opposite side (the AT7-unprotected promoter). This is typical boundary element activity, which blocks the cis-spreading pathway of a transcriptionally repressive nucleoprotein structure on the chromosome as a barrier. Otherwise, the barrier itself is transcriptionally inert (28). Although the boundary element property of LeuO protein is now clear, the underlying mechanism of the synergy (cooperative increase) of LeuO-mediated gene silencing relief in the presence of two LeuO-binding sites on the test plasmid (Fig. 3, lane 8) remains undetermined.

LeuO-LeuO Interaction-dependent LeuO Protein Binding Cooperativity May be Responsible for the Synergy of LeuO-mediated Gene Silencing Relief—Based on the revealed barrier function of LeuO protein, we hypothesized that the interaction between the two LeuO proteins located at the two binding sites may be responsible for the observed blocking synergy (cooperative increase) of LeuO-mediated gene silencing relief. This hypothesis is consistent with the observation that LeuO protein forms oligomers in solution.2 If LeuO-LeuO interaction is indeed responsible for the blocking synergy, such a functional property of LeuO protein is reminiscent of the well known ternary LacI-lacO complex via DNA looping. Theoretically, a LacI tetramer can bind to two lac operators because LacI binds to lacO via two subunits (40). It is also well known that, in the presence of two lac operators, the binding of tetrameric LacI to the lac operators is cooperatively increased (41, 42).

Without our knowing the exact mechanism involved, LacI was able to block the cis-spreading pathway of the transcriptionally repressive nucleoprotein filament found in our model system (12) and in bgl gene silencing (43). Hence, LacI was used here to test whether protein-protein interaction via DNA looping is responsible for the blocking synergy observed in the gene silencing relief mediated by LeuO protein.

We took advantage of the test plasmid pWU802 (linear map illustrated in Fig. 4), in which the AT8 repression element (H-NS nucleation site)-mediated gene silencing activity was shown to be capable of extending to a far distance (~400 bp), so both promoters of the target leuO gene and the downstream bla gene were simultaneously affected by the cis-spreading H-NS-mediated nucleoprotein filament (12). As demonstrated previously, the gene silencing effect on the bla promoter was reduced (2-fold reduction) when lacO was placed at the AatII site. The leuO promoter was not affected because lacO was not located between AT8 and the leuO promoter (Fig. 4, lane 2). This was presumably due to the LacI binding-mediated blockage of the cis-spreading pathway of the nucleoprotein filament (12). Strikingly, the gene silencing relief mediated by LacI was cooperatively increased when an additional lac operator was placed at the HindIII site of the pWU802 test plasmid (Fig. 4, lane 4), whereas the leuO promoter remained unaffected under this condition (Fig. 4, lane 15). The cooperative relief of the gene silencing activity must be due to the simultaneous presence of two lac operators on the test plasmid because a single lac operator at the HindIII site was insufficient to cause any LacI-dependent transcriptional derepression (Fig. 4, lane 3). It is also clear that not only is the simultaneous presence of two lac operators important for the synergistic effect, but one of the two lac operators must be located between the gene silencer and the target promoter. (In this study, the target promoter is the promoter of the bla gene.) This was concluded because LacI did not affect either of the two promoters when neither one of the two lac operators was located between the AT8 gene silencer and the promoters (Fig. 4, lanes 5 and 16). The gene silencing relief is LacI-dependent because both promoters remained silent under all test conditions provided that LacI was not provided in trans (Fig. 4, lanes 7–11 and lanes 18–22).

The striking LacI-dependent cooperative relief of gene silencing (Fig. 4, lane 4) must be due to the interaction between LacI proteins located at the two lac operators on the test plasmid because the synergistic effect was abolished (Fig. 4, lane 6) when dimeric active LacI, R3 (38), was used to replace wild-type LacI in this study. R3 resulted in a basal level (2-fold) of gene silencing relief (the quantified result of Fig. 4, lane 6), which is similar to the degree of gene silencing relief in the plasmid containing only one lac operator at the AatII site (the quantified result of Fig. 4, lane 2). This is consistent with the fact that R3 (defect in tetramerization) maintains its binding to lacO, whereas LacI-LacI interaction capability is impaired (44, 45). Hence, the binding of LacI at the AatII site alone must be responsible for the basal level effect (the 2-fold effect on LacI-mediated gene silencing relief). LacI-LacI interaction cooperatively increased the gene silencing relief (synergistic effect) as observed in Fig. 4 (lane 4). The synergistic effect is most likely due to the cooperative increase in the binding stability of the ternary LacI-lacO complex via DNA looping between two lac operators.

Interestingly, the required LacI-LacI interaction could also be achieved when two lac operators were juxtapositioned (Fig. 4, lane 24). The lacO tandem repeat (zero base pair spacing length between the two lac operators) is expected to allow the interaction between LacI proteins to form the ternary LacI-lacO complex. However, the lacO tandem repeat has to be located between the gene silencer and the target promoter for LacI-mediated transcriptional derepression. This is evidenced by the fact that a lacO tandem repeat located at the HindIII site in pWU802 failed to affect the repressed bla promoter (Fig. 4, lane 25). Serving as an internal control in this study, the leuO promoter remain unaffected under all test conditions (Fig. 4, lanes 12–22 and lanes 26–28) because no lacO was placed between the gene silencer and the leuO promoter. It is therefore clear that it is the ternary LacI-lacO complex rather than the DNA looping that is important for the synergistic blockade of the cis-spreading nucleoprotein filament.

The effect of the lacO tandem repeat on gene silencing relief is strikingly similar to that of the LeuO-binding site tandem repeat as found in a previous experiment (Fig. 2B, lane 5). Clearly, LacI is capable of negating gene silencing activity in a manner similar to that of LeuO protein regarding gene silencing relief. Not many biochemical properties of LeuO protein are presently known because LeuO is a relatively new transcription factor. However, the fact that LeuO proteins form oligomers in solution2 supports the possibility that LeuO-LeuO interaction may result in the required increase in LeuO binding stability, which is then responsible for the synergistic blockade of the pathway of the cis-spreading transcriptionally repressive nucleoprotein filament, as in the outcome demonstrated using LacI (Fig. 4). The detailed biochemical and biophysical properties of LeuO protein will be required for further study of the molecular details involved in LeuO-mediated cooperative relief of gene silencing activity in the presence of two or more LeuO-binding sites. Nonetheless, the present data support the possibility that there may be more than one LeuO-binding site in the LCRs of the ilvIH-leuO-leuABCD gene cluster for LeuO to cooperatively modulate gene silencing, which may be essential for regulating the expression of genes at this locus.

Multiple LeuO-binding Sites in the LCRs of the ilvIH-leuO-leuABCD Gene Cluster Cooperate for Control of LeuO-mediated Expression of the leuO Gene—Thus far, one LeuO-binding site, AT7 (Fig. 5, black box), was identified in LCR-I of the S. typhimurium ilvIH-leuO-leuABCD gene locus. Using a gel mobility shift assay, we systematically searched for additional LeuO-binding site(s) in LCR-I (Fig. 5). LeuO binding affinity was observed in the 188-bp AT2 fragment. LeuO binding affinity was further traced to both the AT3 and AT4 fragments. The AT4 fragment contains AT7, which has been previously characterized and shows preferential binding to LeuO protein (Fig. 5C). Although binding weakly to LeuO, the AT3 DNA clearly interacted with LeuO protein at higher LeuO protein concentration titration points (Fig. 5C, lanes 8–10). DNase I footprinting was used to map the LeuO-binding site(s) in the AT3 DNA. LeuO-dependent DNase I protection was found in the AT3 DNA region, whereas the original LeuO-binding site (AT7) was also detected in the DNase I footprinting assay (Fig. 6). Due to the far distance from the labeled end of the DNA template used in the DNase I footprinting experiments, the exact location of the LeuO-binding site in the AT3 DNA region was not as clear (Fig. 6, left panel). However, a similar DNase I footprinting assay using the DNA template labeled at the opposite end (Fig. 6, right panel) clearly showed that there is a single protected region (LeuO-binding site) in the AT3 DNA region. The two LeuO-binding sites, AT7 and AT3, flank the bacterial gene silencer AT8, as illustrated in Fig. 6. The DNase I footprinting result also clearly demonstrated that the LeuO-binding sites, AT3 and AT7, on LCR-I are located 70 bp apart (the distance between the centers of the two LeuO-dependent DNase I-protected regions as determined in Fig. 6).



View larger version (49K):
[in this window]
[in a new window]
  		FIG. 5.
Identification of all LeuO-binding sites in LCR-I DNA. The 318-bp LCR-I DNA (-335 to -18 of the S. typhimurium leuO gene) was cleaved with restriction enzyme EcoNI into two fragments (130-bp AT1 and 188-bp AT2 DNAs). The two DNA fragments were individually tested for their affinity for LeuO protein by EMSA (A and B). The AT2 DNA was further digested with restriction enzyme HpaI, which resulted in the production of AT3 and AT4 DNAs. The mixture of the AT3 and AT4 DNAs along with the residual AT2 DNA was incubated with increasing amounts of LeuO protein for EMSA (C). The positions of the free and LeuO-bound DNAs are indicated. pleuABCD, leuABCD promoter; pleuO, leuO promoter.

 



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 6.
LeuO protein-dependent DNase I protection in the AT2 region of LCR-I. The 32P-unique end-labeled DNA segments consisting of the DNA sequence from -269 to +40 of the S. typhimurium leuO gene were generated by PCR. The DNA segments were end-labeled either at the +40-position (left panel) or at the -269-position (right panel). The labeled DNAs were incubated with an increasing amount of LeuO protein (as indicated at the top of each lane) for DNase I footprinting studies. To mark the positions of a number of key restriction enzyme sites on the test DNAs, the radioactively labeled DNAs were digested with restriction enzymes and used as position markers (M lanes). The positions of these restriction enzyme sites and the gene silencer AT8 are illustrated in the linear map of LCR-I.

 
According to the previous studies using LacI (42) or other DNA loop-forming proteins such as AraC (46), 70 bp is a spacing length that is unfavorable for supporting protein-protein interaction via DNA looping due to the rigidity of DNA helix. Hence, we continued searching for more LeuO-binding sites in the LCRs of the ilvIH-leuO-leuABCD gene cluster. By EMSA, we indeed detected another LeuO-binding site in LCR-II (Fig. 7A), which is an AT-rich DNA, located downstream of the leuO gene and upstream of the ilvIH promoter (Fig. 7). The AT-rich LCR-II DNA is where multiple Lrp (leucine-responsive regulatory protein)-binding sites are located (47). The gel mobility shift assay further revealed LeuO-binding site(s) located in the ilvIH promoter end of the AT-rich DNA sequence (Fig. 7B). The DNase I footprinting assay showed that LeuO protein-dependent DNase I protection was located primarily between -90 and -113 of the ilvIH operon (Fig. 7, C and D). Although both EMSA and DNase I footprinting results (Fig. 7) indicated that the LeuO-binding site in LCR-II is a relatively weak LeuO-binding site (compared with the LeuO binding affinities of AT3 and AT7 observed in Fig. 5), the study thus far provided key information that an additional LeuO-binding site is present in the downstream region of the leuO gene. The downstream LeuO-binding site is located at a distance of ~1.5 kb from the LeuO-binding sites (AT3 and AT7) in LCR-I. Because the two LeuO-binding sites in LCR-I are unlikely to support LeuO-LeuO interaction for transcriptional regulation due to the small spacing (70 bp) length between the two sites, the downstream LeuO site is very likely to be important for LeuO-LeuO interaction via DNA looping to provide the crucial blocking synergy and to result in cooperative relief of AT8-mediated gene silencing.



View larger version (49K):
[in this window]
[in a new window]
  		FIG. 7.
Identification a LeuO-binding site downstream of the leuO gene in LCR-II. The relative positions of the identified LeuO-binding sites AT3 and AT7 in LCR-I, the leuO gene, and LCR-II are illustrated in the linear map of the S. typhimurium ilvIH-leuO-leuABCD gene cluster. The 254-bp LCR-II DNA (-284 to -31 of the ilvIH operon) was tested for LeuO binding affinity by EMSA, and the results are shown in A. The LCR-II DNA was further digested with restriction enzyme HphI, resulting in 92- and 162-bp DNA segments. The mixture of the two DNA segments was incubated with an increasing amount of LeuO protein for EMSA, and the results are shown in B. The LeuO-bound DNAs are indicated with asterisks. The radioactively unique end-labeled DNA segments consisting of the DNA sequence from -255 to +33 of the S. typhimurium leuO gene were generated by PCR. The DNA segments were end-labeled either at the +33-position (C) or the -255-position (D). The labeled DNAs were incubated with an increasing amount of LeuO protein (as indicated at the top of each lane) for DNase I footprinting assays (C and D). The DNA sequence ladders are included to mark the positions of the LeuO protein-mediated DNase I protection region. pleuO, leuO promoter.

 
This possibility was tested using plasmid pWU802 carrying LCR-I, which contains the gene silencer, the flanking LeuO-binding sites (AT3 and AT7), and the promoters of the target leuO gene and the downstream bla gene, as illustrated in Fig. 8. Indeed, upon providing LeuO protein in trans, the two LeuO-binding sites (AT7 and AT3) in LCR-I supported merely a basal level relief of gene silencing in both the leuO and bla promoters (Fig. 8). A 2-fold relief was found in the leuO promoter (Fig. 8, compare lanes 1 and 2), and a 4-fold relief was found in the bla promoter (compare lanes 6 and 7). Both LeuO-mediated silencing relievers were strikingly enhanced in the presence of an additional LeuO-binding site located 260 bp downstream at the AatII site (Fig. 8, lanes 4 and 8). In this experiment, the additional LeuO-binding site at the AatII site functionally mimicked the downstream LeuO-binding site found in LCR-II. This is because, like the 1.5-kb distance, the 260-bp distance is theoretically sufficient to allow efficient LeuO-LeuO interaction via DNA looping (42). Hence, this is strong evidence that, via DNA looping, the downstream LeuO-binding site in LCR-II must be functionally crucial in LeuO-mediated synergistic relief of the silenced leuO gene as part of the promoter relay mechanism (35). This synergy is likely to be responsible for the coordinated expression of genes in the ilvIH-leuO-leuABCD gene cluster (39, 48).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Although the revealed boundary element-like activity is unprecedented for bacterial gene expression regulation, it is basically consistent with the barrier function of the boundary elements/insulators found in eukaryotes (26, 28, 49). A new molecular detail regarding this bacterial boundary element activity is that the barrier activity is synergistically increased in the presence of more than one binding site of the boundary element. Such blocking polarity and synergy supported solely by multiple LeuO-binding sites may help explain how transcriptionally active and inert domains on the chromosome are delimited by various barrier proteins (50). This is especially true for DNA loop-forming barrier proteins (31) because the formation of boundaries that spatially separate chromatin domains is a common feature suggested in the mechanism of the eukaryotic barrier (51–53). In fact, many eukaryotic boundary elements/insulators are found in pairs or in multiple copies (54).

Theoretically, the presence of more than two binding sites of DNA loop-forming barrier proteins should result in multiple blocking scenarios for regional gene expression regulation. Due to the relatively high LeuO binding affinity, the LeuO-binding site AT7 was the first LeuO site identified in this transcriptional regulation model system (33). According to the EMSA results of the present study, the order of LeuO binding affinity of regional multiple sites is AT7 > AT3 > downstream LCR-II LeuO site. The differential LeuO binding affinities of these three sites may be crucial in determining the outcome of LeuO-mediated relief of leuO silencing. If the AT7 LeuO-binding site is the one interacting with the downstream LCR-II LeuO site, the AT8 gene silencer and the leuO promoter are sequestered by this LeuO-LeuO interaction-mediated DNA loop, so the cis-spreading transcriptionally repressive nucleoprotein structure mediated by AT8 is prevented from reaching the leuO promoter (Fig. 9A), thus derepressing it. On the other hand, if the AT3 LeuO-binding site is the one interacting with the downstream site(s), the AT8 gene silencer-mediated nucleoprotein filament remains able to repress the leuO promoter because both AT8 and the leuO promoter are in the same domain (Fig. 9B).



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 9.
Dynamic DNA looping-mediated LeuO-LeuO interaction scenarios that control the expression of the leuO gene as part of the promoter relay mechanism via the boundary element-like activity of LeuO protein. See "Discussion" for detailed descriptions of the three scenarios. The revealed LeuO affinities (AT7 > AT3 > downstream LeuO-binding site in LCR-II) of the three involved LeuO-binding sites in the ilvIH-leuO-leuABCD gene cluster may determine the frequency and stability of the three proposed DNA looping conditions during gene expression coordination (viz. the promoter relay mechanism) in the gene cluster. The facilitation or competition between LeuO protein and other chromosomal architectural proteins (viz. Lrp) for binding sites in the region may also determine the frequency and stability of the three scenarios. This is because the downstream LeuO-binding site in LCR-II is co-localized with the binding sites of Lrp, which is the positive regulator of the expression of the ilvIH operon. If Lrp facilitates the LeuO protein binding at the downstream LeuO-binding site in LCR-II, the scenario shown in A could be the mechanism that explains why the leuO gene is activated by ilvIH transcriptional activity located 1.5 kb downstream as the first step in the promoter relay mechanism (35, 70). pleuABCD, pilvIH, and pleuO, leuABCD, ilvIH, and leuO promoters, respectively; RNAP, RNA polymerase.

 
The spacing length of 70 bp between LeuO-binding sites AT3 and AT7 makes LeuO-LeuO interaction via forming such a small DNA loop very unlikely under normal conditions. However, DNA bending (55, 56) and negative DNA supercoiling (57) are known to provide the required DNA flexibility for the formation of a small DNA loop (58–60). Given the facts that the AT8 gene silencer possesses an intrinsic curvature DNA sequence (12) and that transcription-driven DNA supercoiling actively participates in the coordination of gene expression (viz. the promoter relay mechanism), it is very likely that the 70-bp DNA loop is formed under certain condition(s). This will result in a completely active domain for gene expression in the entire locus when AT8 is isolated and bent in the small DNA loop (Fig. 9C). Perhaps the final activation of either the mutant leu-500 promoter or the wild-type leuABCD promoter in the promoter relay mechanism (35) is due to the isolation of AT8 in a small DNA loop as described.

The boundary element activity of LeuO that blocks the cis-spreading pathway of an H-NS-associated nucleoprotein filament is consistent with the fact that LeuO plays either a positive or negative role in effecting a number of hns- phenotypes (61–65). Among these examples, the most interesting case is that LeuO overexpression relieves H-NS-mediated bgl silencing in E. coli (63). LeuO may indeed prevent the cis-spreading nucleoprotein filament from reaching the bgl promoter as proposed. Interestingly, LacI is also capable of affecting bgl silencing (43). The barrier function of LeuO also explains the negative regulatory role of LeuO found in the control of expression of E. coli dsrA (65). As a barrier protein, LeuO blocks the cis-spreading of the transcriptionally repressive nucleoprotein filament; otherwise, it is a transcriptionally inert element. Hence, the exact transcriptional effect exerted depends on the nature of the transcriptional apparatus blocked by LeuO. With such a neutral chromosomal domain-delimiting function, it is likely that the barrier protein LeuO plays a negative role in dsrA regulation (65) and a positive role in leuO regulation (33).

Although LacI replaces LeuO in the gene silencing blockage assay (Fig. 4), the revealed protein-protein interaction via DNA looping is unlikely the only property important for the transcriptional regulatory function of LeuO. This is because LeuO rather than LacI is evolutionarily conserved for taking part in the promoter relay mechanism. LeuO must possess some not yet identified properties that are also important for coordinating gene expression in the locus and perhaps throughout the bacterial chromosome. For instance, localized LeuO may recruit other proteins to the region for further modification of the local nucleoprotein structure. The exact post-blocking modifications mediated by LeuO may not be the same as the known eukaryotic modifications (e.g. histone acetylation and deacetylation, etc.). These additional LeuO-specific properties may, however, be the reason why LacI cannot replace LeuO in the promoter relay mechanism.

The juxtapositioned LeuO sites or lac operators provided the synergy in blocking the transcriptionally repressive nucleoprotein filament. This result suggested that, instead of DNA looping, the protein-protein interaction-dependent protein binding cooperativity must be responsible for the blocking synergy. However, DNA looping may still be required for the sequestration of DNA supercoiling in the region and may provide a higher ordered transcriptional control for regional genes. After all, transcription-driven negative supercoiling was shown to be important for bacterial gene silencing. The adjacent transcriptional activities transcribing away from the gene silencer are crucial for AT8 to exert its full gene silencing activity (34, 66). If AT8 is isolated in a small DNA loop (70 bp) as described (Fig. 9C), the sequestration of the gene silencer from the effect of DNA supercoiling driven by flanking genes (e.g. leuABCD and leuO in this case) must also be an important mechanism for maintaining the isolated gene silencer in its inactive state, e.g. H-NS interacts much less efficiently with the sequestered gene silencer. Similar DNA supercoiling segregation may also be applicable in the eukaryotic boundary elements/insulators that delimit active and inactive chromatin domains by looping out DNA on the chromosome.

Furthermore, LeuO-LeuO interaction between LeuO-binding sites may not be limited to a particular gene locus for transcriptional regulation. A long-range interaction of LeuO proteins between LeuO-binding sites throughout the bacterial chromosome is also possible. Such long-range chromosomal modulation is seen in the higher order domains of chromosomal architecture in eukaryotes. A Drosophila insulator, the Su(Hw)-binding site gypsy, determines the co-localization of the Su(Hw)-Mod(mdg4) protein complex and its attachment to nuclear lamina (53, 67, 68). These processes result in the formation of Su(Hw) protein-dependent domains that sequester the enhancer upstream of and the target gene (promoter) downstream of the gypsy element insulator (69). This level of chromosome-wide sequestration may affect not only the regulation of expression of a particular gene, but also other DNA-based processes such as replication and recombination.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grant GM-53617. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed: Dept. of Pharmacology, School of Medicine, Wayne State University, 540 E. Canfield Ave., Detroit, MI 48201. Tel.: 313-577-1584; Fax: 313-577-6739; E-mail: haiwu{at}med.wayne.edu.

1 The abbreviations used are: H-NS, histone-like nucleoid structuring protein; EMSA, electrophoretic mobility shift assay; LCR, locus control region. Back

2 H.-Y. Wu, unpublished data. Back


   ACKNOWLEDGMENTS
 
We thank Dr. Victoria Kimler for critical reading of the manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Hendrich, B. D., and Willard, H. F. (1995) Hum. Mol. Genet. 4, 1765-1777[Abstract]
  2. Festenstein, R., and Kioussis, D. (2000) Curr. Opin. Genet. Dev. 10, 199-203[CrossRef][Medline] [Order article via Infotrieve]
  3. Yarmolinsky, M. (2000) Curr. Opin. Microbiol. 3, 138-143[CrossRef][Medline] [Order article via Infotrieve]
  4. Schnetz, K. (1995) EMBO J. 14, 2545-2550[Medline] [Order article via Infotrieve]
  5. Schnetz, K., and Wang, J. C. (1996) Nucleic Acids Res. 24, 2422-2428[Abstract/Free Full Text]
  6. Mukerji, M., and Mahadevan, S. (1997) Mol. Microbiol. 24, 617-627[CrossRef][Medline] [Order article via Infotrieve]
  7. Ueguchi, C., and Mizuno, T. (1993) EMBO J. 12, 1039-1046[Medline] [Order article via Infotrieve]
  8. Jordi, B. J., Fielder, A. E., Burns, C. M., Hinton, J. C., Dover, N., Ussery, D. W., and Higgins, C. F. (1997) J. Biol. Chem. 272, 12083-12090[Abstract/Free Full Text]
  9. Jordi, B. J., and Higgins, C. F. (2000) J. Biol. Chem. 275, 12123-12128[Abstract/Free Full Text]
  10. Dorman, C. J. (2004) Nat. Rev. Microbiol. 2, 391-400[CrossRef][Medline] [Order article via Infotrieve]
  11. Rimsky, S. (2004) Curr. Opin. Microbiol. 7, 109-114[CrossRef][Medline] [Order article via Infotrieve]
  12. Chen, C.-C., Chou, M. Y., Huang, C. H., Majumder, A., and Wu, H.-Y. (2005) J. Biol. Chem. 280, 5101-5112[Abstract/Free Full Text]
  13. Kim, S. K., and Wang, J. C. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 8557-8561[Abstract/Free Full Text]
  14. Hao, J. J., and Yarmolinsky, M. (2002) J. Bacteriol. 184, 4857-4867[Abstract/Free Full Text]
  15. Rodionov, O., Lobocka, M., and Yarmolinsky, M. (1999) Science 283, 546-549[Abstract/Free Full Text]
  16. Lynch, A. S., and Wang, J. C. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 1896-1900[Abstract/Free Full Text]
  17. Hanai, R., Liu, R., Benedetti, P., Caron, P. R., Lynch, A. S., and Wang, J. C. (1996) J. Biol. Chem. 271, 17469-17475[Abstract/Free Full Text]
  18. Murphy, T. D., and Karpen, G. H. (1998) Cell 93, 317-320[CrossRef][Medline] [Order article via Infotrieve]
  19. Perrod, S., and Gasser, S. M. (2003) CMLS Cell. Mol. Life Sci. 60, 2303-2318
  20. Tham, W. H., and Zakian, V. A. (2002) Oncogene 21, 512-521[CrossRef][Medline] [Order article via Infotrieve]
  21. Zhimulev, I. F., and Beliaeva, E. S. (2003) Genetika 39, 187-201[Medline] [Order article via Infotrieve]
  22. Kamakaka, R. T. (2003) Curr. Biol. 13, R317-R319[CrossRef][Medline] [Order article via Infotrieve]
  23. Bi, X., Braunstein, M., Shei, G. J., and Broach, J. R. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 11934-11939[Abstract/Free Full Text]
  24. Fourel, G., Revardel, E., Koering, C. E., and Gilson, E. (1999) EMBO J. 18, 2522-2537[CrossRef][Medline] [Order article via Infotrieve]
  25. Bell, A. C., West, A. G., and Felsenfeld, G. (2001) Science 291, 447-450[Abstract/Free Full Text]
  26. Donze, D., and Kamakaka, R. T. (2002) BioEssays 24, 344-349[CrossRef][Medline] [Order article via Infotrieve]
  27. Pryde, F. E., and Louis, E. J. (1999) EMBO J. 18, 2538-2550[CrossRef][Medline] [Order article via Infotrieve]
  28. Bell, A. C., and Felsenfeld, G. (1999) Curr. Opin. Genet. Dev. 9, 191-198[CrossRef][Medline] [Order article via Infotrieve]
  29. Oki, M., and Kamakaka, R. T. (2002) Curr. Opin. Cell Biol. 14, 299-304[CrossRef][Medline] [Order article via Infotrieve]
  30. Haber, J. E. (1998) Annu. Rev. Genet. 32, 561-599[CrossRef][Medline] [Order article via Infotrieve]
  31. Kamakaka, R. T. (2002) Curr. Biol. 12, R535-R537[CrossRef][Medline] [Order article via Infotrieve]
  32. Donze, D., and Kamakaka, R. T. (2001) EMBO J. 20, 520-531[CrossRef][Medline] [Order article via Infotrieve]
  33. Chen, C.-C., Ghole, M., Majumder, A., Wang, Z., Chandana, S., and Wu, H.-Y. (2003) J. Biol. Chem. 278, 38094-38103[Abstract/Free Full Text]
  34. Chen, C.-C., Fang, M., Majumder, A., and Wu, H.-Y. (2001) J. Biol. Chem. 276, 9478-9485[Abstract/Free Full Text]
  35. Fang, M., and Wu, H.-Y. (1998) J. Bacteriol. 180, 626-633[Abstract/Free Full Text]
  36. Casadaban, M. J., and Cohen, S. N. (1980) J. Mol. Biol. 138, 179-207[CrossRef][Medline] [Order article via Infotrieve]
  37. Oehler, S., Eismann, E. R., Kramer, H., and Muller-Hill, B. (1990) EMBO J. 9, 973-979[Medline] [Order article via Infotrieve]
  38. Lehming, N., Sartorius, J., Niemoller, M., Genenger, G., von Wilcken-Bergmann, B., and Muller-Hill, B. (1987) EMBO J. 6, 3145-3153[Medline] [Order article via Infotrieve]
  39. Wu, H.-Y., Tan, J., and Fang, M. (1995) Cell 82, 445-451