HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 51, 40218-40225, December 22, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/51/40218    most recent M909658199v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Becker, N. A. Articles by Maher, L. J. Search for Related Content PubMed PubMed Citation Articles by Becker, N. A. Articles by Maher, L. J., III Social Bookmarking                      What's this?

In Vitro and in Vivo Ligation-mediated Polymerase Chain Reaction Analysis of a Polypurine/Polypyrimidine Sequence Upstream of the Mouse metallothionein-I Gene^*

Nicole A. Becker, Heather A. O'Neill, Jeff M. Zimmerman, and L. James Maher III

From the Department of Biochemistry and Molecular Biology, Mayo Foundation, Rochester, Minnesota 55905

Received for publication, November 30, 1999, and in revised form, May 9, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The mouse metallothionein-I homopurine/homopyrimidine (MT-I R/Y) sequence is a 128-base pair element located ~1.2 kilobase pairs upstream of the MT-I gene. Previous in vitro studies of this sequence in purified plasmids indicated the formation of a non-B DNA structure stabilized by acidic pH and negative supercoiling. We now present a detailed in vitro and in vivo analysis of the MT-I R/Y sequence using chemical probes of DNA structure and ligation-mediated polymerase chain reaction. In vivo analysis suggests neither profound base unpairing nor protein binding within the MT-I R/Y sequence before or after metal induction of MT-I. We conclude for this element that the propensity to adopt an unusual DNA structure in vitro does not imply the occurrence of such a structure in vivo. We were able to show both in purified genomic DNA and in vivo that only isolated thymines and the 3' terminal thymine in strings of consecutive thymines are modified significantly by KMnO₄, indicating an altered thymine accessibility pattern within the R/Y sequence. This KMnO₄ reactivity pattern is more consistent and predictable within the R/Y sequence when compared with flanking sequences. We propose a simple steric interference model to explain the observed pattern of KMnO₄ modification of thymines.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The familiar DNA double helix typically exists in B-form when studied as a purified molecule in solution. The extent to which DNA adopts other, more exotic structures in vivo is a fascinating question of continuing interest (reviewed in Ref. 1). For example, during transcription and replication, DNA exists in a partially unpaired configuration. These two enzymatic processes promote obvious non-B DNA structures that are transient and largely sequence-independent.

Under specific in vitro conditions, unusual non-B DNA structures have also been shown to exist at specific sequences (reviewed in Refs. 1 and 2). These structures include Z-DNA, cruciform DNA, and H-DNA. Z-DNA (2-5) is a left-handed DNA helix with a sequence preference for alternating purine-pyrimidine repeats. Cruciform DNA (2, 6) is characterized by hairpin extrusions from both DNA strands with the corresponding requirement for palindromic sequence symmetry, and a preference for A/T-rich DNA. H-DNA (2, 7, 8) is characterized by an intramolecular DNA triplex and accompanying regions of base unpairing with a sequence requirement for homopurine/homopyrimidine mirror repeats (7, 9). All three of the non-B structures described above are stabilized by negative supercoiling (which favors base unpairing) and typically require additional sequence-specific in vitro conditions. For example, H-DNA isomers involving the pyrimidine triple helix motif are strongly stabilized by protonation of cytosines at acidic pH (7), and Z-DNA structures are stabilized at high ionic strength (3). Z-DNA, cruciform DNA, and H-DNA have been detected in isolated plasmids under specific in vitro conditions, and in isolated nuclear preparations using antibodies thought to be specific for these structures (10-12). Other structural transitions involving base unpairing have been proposed in certain enhancer elements (13-15).

The major non-B DNA structures described above all contain regions of base unpairing. Detection of unpaired structures at single nucleotide resolution typically requires treatment of large amounts (5-10 µg) of purified plasmid with chemical probes that specifically modify either unpaired or unstacked bases (16). Here we apply a different approach involving the use of cell-permeable chemical probes together with LMPCR¹ to characterize the in vitro and in vivo chemical reactivity of a naturally occurring R/Y sequence. LMPCR has been shown to be an excellent tool for in vivo footprinting of protein-DNA interactions and methylation patterns (17, 18). We now apply LMPCR to examine the possibility of unusual DNA structures or protein binding within a long R/Y sequence in vivo. LMPCR permits the analysis of specific gene sequences at nucleotide resolution using as little as 10 pg of plasmid or 2 µg of total eukaryotic genomic DNA. LMPCR relies on exponential amplification of a population of gene-specific chemical degradation fragments to create a sequencing pattern.

The focus of our study is a peculiar R/Y sequence upstream of the mouse MT-I gene (19, 20). We refer to this element as the MT-I R/Y sequence (Fig. 1). The MT-I R/Y sequence is of particular interest for three reasons. First, it is upstream of the well characterized MT-I gene. Metallothioneins are thought to play an important role in metal homeostasis (21-23). MT-I gene expression is induced by heavy metals via transcription factor binding to several copies of a 15-bp metal responsive element (MRE) present in the promoter regions of MT genes (24-28). The mouse MT-I proximal promoter contains six such MREs mediating metal induction. Upon exposure to heavy metals, a rapid 5-20-fold increase in MT mRNA levels is typically detected in cultured cells (24, 25). Second, the MT-I R/Y sequence is of interest because of its unusual length and provocative location (centered 1184 bp upstream of the MT-I gene). The element is a 128-bp R/Y sequence containing a single pyrimidine interruption within the homopurine strand. The MT-I R/Y sequence contains no single plane of extensive mirror symmetry, but it contains several imperfect homopurine/homopyrimidine mirror repeats. Third, this sequence has been previously studied in purified plasmid in vitro and shown to adopt a non-B DNA structure reminiscent of H-DNA in vitro. Formation of this non-B DNA structure requires negative superhelical strain and additional stabilization is offered by low pH (19, 20). Our previous genetic studies have not detected a function for the MT-I R/Y sequence in the regulation of MT-I gene expression after transient transfection of reporter constructs into cultured cells (20).

We now use chemical probes of DNA structure together with LMPCR to determine if the MT-I R/Y sequence adopts a discernable non-B DNA structure before or after metal induction of cultured cells. We first show LMPCR data for plasmids studied in vitro. These data confirm and extend our published demonstration of a non-B DNA structure that depends upon negative supercoiling and acidic pH (20). In contrast, the results of in vivo probing with KMnO₄ and chloroacetaldehyde (CAA) do not indicate the formation of either a profoundly unpaired DNA structure or protein complexes within the MT-I R/Y sequence. Using KMnO₄ as a probe of thymine accessibility, we show both in vivo and with purified genomic DNA that thymine residues within the R/Y sequence react with KMnO₄ in a predictable pattern that is more consistent than for thymine residues outside the R/Y sequence. These data suggest that although the R/Y sequence is not forming an unpaired structure, its thymine bases have subtle distinguishing characteristics. Possible explanations for the uniform pattern of thymine reactivity within the MT-I R/Y sequence are discussed.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Construct-- Plasmid pMTCAT (6300 bp (29)) contains ~1775 bp of the mouse MT-I promoter including the upstream MT-I R/Y sequence. Plasmid pMTCAT at native superhelical density was extracted from E. coli DH5 cells and purified by CsCl equilibrium density gradient centrifugation in the presence of ethidium bromide (30).

Oligonucleotides-- All oligonucleotides were synthesized by standard phosphoramidite chemistry on an ABI 394 synthesizer. The oligonucleotides were purified on 20% denaturing polyacrylamide gels, eluted from gel slices, and desalted by Sep-Pak C₁₈ cartridge chromatography (Waters).

Cell Culture, CdCl₂ Induction, and Reverse Transcription Analysis of MT-I RNA-- Mouse NIH3T3 fibroblasts were maintained at 37 °C and 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Life Technologies, Inc.). For metal induction and RNA isolation, NIH3T3 cells at 90% confluency were shifted to Dulbecco's modified Eagle's medium with or without 2 µM CdCl₂ for 4 h. Total cellular RNA was isolated from cells using Trizol Reagent (Life Technologies, Inc.) according to the manufacturer's instructions. RNA concentration was determined spectrophotometrically at 260 nm. In each 10-µl reverse transcription reaction, 15 µg of total RNA was combined with 0.5 pmol of the MT-I mRNA reverse primer (5'-GCAG₂AG₂TGCACT₂GCAGT₂C) and 0.5 pmol of the glyceraldehyde-3-phosphate dehydrogenase mRNA reverse primer (5'-A₃TG₂ CAGC₃TG₂TGAC₂) in 200 mM KCl. The MT-I and glyceraldehyde-3-phosphate dehydrogenase primers yield reverse transcripts of 159 and 110 nucleotides, respectively. Reverse primers were annealed to total RNA by incubation at 80 °C for 5 min, 42 °C for 20 min, and 37 °C for 30 min. Each reaction was supplemented with 25 µl of a reaction mixture containing 20 mM Tris-HCl, pH 8.3, 10 mM MgCl₂, 5 mM dithiothreitol, 0.3 mM dNTPs, 10 pg/µl actinomycin D, and 25 units of Moloney murine leukemia virus-reverse transcriptase enzyme (Promega). Primer extension was carried out at 42 °C for 30 min. Each reaction was terminated by addition of 100 µl of room temperature stop buffer (0.3 M sodium acetate, pH 5.2, 10 mM MgCl₂, and 0.5% glycogen) followed by ethanol precipitation. Reverse transcripts were analyzed by electrophoresis through a denaturing 5% polyacrylamide gel (19:1, acrylamide:bisacrylamide).

In Vitro Chemical Treatment of Plasmid and Genomic DNA Samples-- Anhydrous hydrazine, dimethyl sulfate (DMS), 2,2'-bipyridyl, formic acid, osmium tetroxide (OT: 4% weight solution in H₂O), KMnO₄, and piperidine were obtained from Aldrich. CAA (45% weight solution in H₂O) was obtained from Fluka. All chemical reagents were used without further purification. The chemical reactivities of pMTCAT DNA and naked genomic DNA were analyzed using modifications of published procedures (31). Briefly, supercoiled or HpaI-linearized pMTCAT DNA samples (1 µg) were dissolved in 100 µl of buffer containing 15 µg of sheared calf-thymus DNA as a carrier. For plasmid experiments at pH 4.5, the buffer contained 20 mM sodium acetate, pH 4.5, 4 mM MgCl₂, and 100 mM NaCl. For plasmid experiments at pH 7.1, the buffer contained 25 mM MOPS, pH 7.1, 4 mM MgCl₂, and 100 mM NaCl. For in vitro analysis of genomic DNA, 40 µg of purified NIH3T3 genomic DNA was dissolved in 300 µl of buffer containing 25 mM MOPS, pH 7.1, 4 mM MgCl₂, and 100 mM NaCl. To monitor OT reactivity, samples were treated for 15 min at 37 °C by adding 7.6 µl (plasmid samples) or 15.2 µl (genomic samples) of a solution made by mixing 14.2 µl of 0.5% 2,2'-bipyridyl and 3 µl of OT solution (final concentration of both OT and 2,2'-bipyridyl was 2 mM). OT reactions were terminated by adding 0.1 volume of sodium acetate, pH 5.2, followed by two ethanol precipitations. To monitor KMnO₄ reactivity, samples were treated with 5 mM (final concentration) KMnO₄ for 2 min at 24 °C. KMnO₄ reactions were terminated by adding 0.1 volume of sodium acetate, pH 5.2, containing 1 M -mercaptoethanol, followed by two ethanol precipitations. To monitor DMS reactivity, samples were treated with a 5% aqueous DMS solution in 50 mM sodium cacodylate and 1 mM EDTA (0.5% final DMS concentration) for 2 min at 24 °C. DMS reactions were terminated by adding 0.1 volume of sodium acetate, pH 5.2, containing 1 M -mercaptoethanol, followed by two ethanol precipitations. To monitor CAA reactivity, samples were treated with 2% (final concentration) CAA for 30 min at 37 °C. CAA reactions were terminated by adding 50 mM (final concentration) NaCl followed by two ethanol precipitations. CAA-treated samples were resuspended in H₂O and then incubated with either formic acid or hydrazine in high salt followed by two ethanol precipitations to superimpose the CAA data on Maxam and Gilbert reference ladders (32). Dried DNA samples were then resuspended in 1 M piperidine, and base modifications were cleaved by incubation at 90 °C for 30 min. The samples were lyophilized extensively to remove all traces of piperidine and resuspended in H₂O. DNA samples were then analyzed by LMPCR.

In Vivo Chemical Treatment-- Mouse NIH3T3 fibroblasts were seeded into 10-cm culture plates and maintained at 37 °C and 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. For CdCl₂ induction, NIH3T3 cells at 80-90% confluency were grown in Dulbecco's modified Eagle's medium for 4 h with or without 2 µM CdCl₂. For chemical treatment, cells at 80-90% confluency were washed twice with phosphate-buffered saline (PBS) at ambient temperature just prior to treatment. For KMnO₄ treatment, cells were exposed to 10 mM KMnO₄ in PBS for 2 min at 24 °C. For DMS treatment, cells were exposed to 0.1% DMS in PBS for 2 min at 24 °C. For CAA treatment, cells were exposed to 3% CAA in PBS for 20 min at 37 °C. All reactions were terminated by washing cells twice with ice-cold PBS. Cell lysis and DNA purification were by standard protocols (33). CAA-treated DNA was then processed as described above for in vitro-treated DNA to superimpose either G + A or C > T chemistry. Dried genomic DNA was cleaved with piperidine and prepared for LMPCR as described for in vitro genomic samples. The genomic DNA concentration was then determined spectrophotometrically at 260 nm.

LMPCR Analysis-- Three gene-specific primer sets were designed with the assistance of Oligo 4.0 software for LMPCR analysis of the MT-I promoter (34). Primer set A was used to analyze the pyrimidine strand of the MT-I R/Y sequence and included gene-specific primers A1 (5'-C₂A₃TGA₂G₂T₉G₂ATC), A2 (5'-G₂ATCT₂C₃ACGCAGAG₃TAT₂GC), and A3 (5'-G₂TAT₂GCTGTGT₂GTCTC₂TC₂A₂GAG). Primer set B was used to analyze the TATA-proximal promoter non-template strand and included gene-specific primers B1 (5'-C₂AG₃AGCTCTGCACTC), B2 (5'-TCTGCACTC₂GC₃GA₄GTG), and B3 (5'-GTGCGCTCG₂CTCTGC₂A₂G). Primer set C was used to analyze the TATA-proximal promoter template strand and included gene-specific primers C1 (5'-AGCAGT₂G₄T-C₂AT₂C), C2 (5'-TC₂AT₂C₂GAGATCTG₂TGA₂GCTG), and C3 (5'-G₂TGA₂GCTG₂A-GCTACG₂AGA₂G). The chemical reactivities of genomic and plasmid samples were analyzed using standard LMPCR methods (17, 18). Briefly, chemically treated and purified DNA samples (2 µg of genomic DNA or 10 pg of plasmid DNA) were used at the start of LMPCR. Primer extension was performed using Sequenase 2.0 (Amersham Pharmacia Biotech) with the first gene-specific primer. After ligation of a blunt unidirectional linker duplex (17, 18), PCR was performed in the presence of Taq DNA polymerase (Fisher Scientific), the second gene-specific primer, and the unidirectional linker primer (5'-GCG₂TGAC₃G₃AGATCTGA₂T₂C) for 20 cycles. The PCR products were separated on either a 6 or 8% denaturing polyacrylamide sequencing gel (19:1, acrylamide:bisacrylamide). The sequencing gel was electroblotted to a nylon membrane (PerkinElmer Life Sciences), prehybridized, and hybridized with a radioactive probe created by single-sided PCR using the third gene-specific primer. Sequencing markers were created by standard Maxam and Gilbert (32) chemical modifications of genomic DNA in vitro, together with standard LMPCR detection. Radioactive signals were analyzed using a Molecular Dynamics Storm 840 PhosphorImager.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Strategy-- Chemical and enzymatic probes of non-B DNA structures have previously been employed in vitro to characterize the 128-bp MT-I R/Y sequence (Fig. 1) present in purified plasmids (19, 20). Using both low and high resolution probing techniques, these studies indicated that the MT-I R/Y sequence forms a non-B DNA structure under conditions of negative superhelical strain and low pH (19, 20). The pattern of chemical reactivity suggested the possible formation of multiple, concurrent H-DNA-like structures. Transient transfection assays detected no effect of the R/Y sequence on promoter strength (20). Thus, although the MT-I R/Y sequence is able to adopt a non-B DNA structure under specific in vitro conditions, it plays no demonstrable role in transcriptional regulation. These previous data did not rule out the possibility that the R/Y sequence adopts a non-B DNA structure in vivo, perhaps stabilized by increased negative superhelical strain that might accompany metal-induced transcription of the MT-I gene (35). This possibility is emphasized by a recent report of transcription-induced cruciform extrusion within living Escherichia coli cells (36). Assessment of chemical reactivity in the MT-I R/Y sequence before and after induction of MT-I transcription by exogenous metal thus provides an excellent eukaryotic opportunity to explore the possibility of similar phenomena.

View larger version (56K): [in this window] [in a new window]   Fig. 1.   Mouse MT-I promoter regions studied by in vivo footprinting. Upper sequence includes the 128-bp MT-I R/Y sequence and flanking region. The pyrimidine strand (shaded) is studied in this paper. The MT-I R/Y sequence is centered at position 1184 relative to the major MT-I transcription start point (tsp). A single pyrimidine interruption (·) occurs within the purine strand at position 1176. Lower sequence is the TATA proximal promoter region of the MT-I gene. Shown are three of the six metal responsive elements (MRE) and the TATA box (shaded). The major MT-I tsp is indicated by +1.

In the present study we sought to answer the following questions. Can chemical probing and LMPCR be used to corroborate the presence of a non-B DNA structure in plasmid DNA containing the MT-I R/Y sequence? Is there evidence of protein binding to this element in cultured cells either before or after metal induction? Is there evidence for a non-B DNA structure within the R/Y sequence in vivo either before or after metal induction?

Before LMPCR analysis of genomic DNA in cultured NIH3T3 cells, we wished to confirm the metal responsiveness of the endogenous MT-I gene in these cells. Reverse transcription analysis was employed to quantitate MT-I mRNA levels before and after CdCl₂ induction (Fig. 2). NIH3T3 cells induced with 2 µM CdCl₂ for 4 h showed a ~10-fold increase in MT-I mRNA levels compared with uninduced cells (Fig. 2, compare lanes 2 and 3). Glyceraldehyde-3-phosphate dehydrogenase mRNA was monitored as an internal control. This result confirmed that our experimental metal treatment significantly induced MT-I gene expression, and suggested that metals might increase negative superhelical strain in the vicinity of the MT-I R/Y sequence.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Reverse transcription analysis of MT-I mRNA levels following metal induction. NIH3T3 cells at 90% confluency were grown in media supplemented with or without 2 µM CdCl2 for 4 h. Reverse transcriptase analysis was performed as described under "Experimental Procedures." The reverse transcriptase products were analyzed on a 5% denaturing polyacrylamide gel. Marker (lane 1) is radiolabeled 100-nucleotide ladder.

Chemical probes of DNA structure were used in combination with LMPCR to analyze two regions of the MT-I promoter (Fig. 1). Chemical reagents that react preferentially with unusual DNA structures (CAA, OT, and KMnO₄) or serve as footprinting agents (DMS) can be valuable for detecting unusual non-B DNA structures and indicating regions of protection resulting from the binding of proteins or nucleic acid strands (16, 31, 37). DMS modifies the N-7 position of guanine within a DNA duplex. In contrast, guanine is protected from DMS modification when a protein is bound to the DNA or when guanine is involved in Hoogsteen (or reverse-Hoogsteen) hydrogen bonding (e.g. within a C + G·C triplet) because the N-7 position of guanine is made inaccessible to DMS. KMnO₄ is thought to react primarily with unstacked or unpaired thymine bases, and to a much lower degree with unstacked or unpaired cytosine and guanine residues (38). CAA reacts with the Watson-Crick base pairing face of unpaired adenine and cytosine residues to form their respective etheno derivatives. OT, in the presence of 2,2'-bipyridyl reacts primarily with unpaired or unstacked thymine bases (38).

Chemical Probing of the Proximal MT-I Promoter in Vitro and in Vivo-- To validate our chemical probing and LMPCR assay, we first analyzed the TATA-proximal region of the MT-I promoter in isolated genomic DNA and in cultured cells. Previous in vivo studies of the MT-I promoter used DMS and a genomic primer extension analysis to reveal subtle differences between the chemical reactivity of promoter DNA sequences of uninduced and metal-induced cells (24). Mueller et al. (24) noted differences between DMS reactivities of samples treated in vitro versus in vivo as well as differences between samples treated in vivo before or after CdCl₂ induction. We analyzed NIH3T3 cells before or after CdCl₂ induction using either DMS or KMnO₄ followed by DNA extraction and LMPCR analysis. Reactivities were compared with those observed for isolated genomic DNA treated with the same reagents.

KMnO₄ has not previously been used to analyze the MT-I promoter. Differences in KMnO₄ reactivity on both DNA strands of the promoter were noted when in vitro and in vivo samples were compared (Fig. 3, A and B, compare lanes 4 and 5). KMnO₄ footprints were evident within the MREs on both non-template and template strands (Fig. 3, A and B, red bars and circles). Curiously, only subtle increases in reactivity were detected near the transcriptional start point (Fig. 3A, compare lanes 1-3, red circle at +3), but these differences indicated base unpairing in vivo. Metal induction also resulted in subtle differences between uninduced and CdCl₂-induced samples (Fig. 3, A and B, compare lanes 5 and 6, blue circles). It was interesting and unexpected that many of the MRE footprints detected in vivo by KMnO₄ were present prior to metal induction. We also noted that in strings of two or more consecutive thymine residues, KMnO₄ reacts only with the 3' terminal thymine residue of the string. This curious phenomenon has been previously reported and will be further discussed below (38, 39).

View larger version (66K): [in this window] [in a new window]   Fig. 3.   LMPCR analysis of the non-template and template strands of the TATA region in the MT-I promoter. The same DNA samples were analyzed using different primer sets to examine either the non-template (panel A) or template (panel B) strands. The MREs and TATA box are indicated (left of each panel). The positions relative to the major MT-I tsp are labeled (right of each panel). Lanes 4-6 indicate relative reactivities to KMnO4. Lanes 7-9 indicate relative reactivities to DMS. Lanes 4 and 7 display the results of in vitro treatment of purified genomic DNA with the indicated chemicals. Lanes 5 and 8 display the results of in vivo treatment of uninduced NIH3T3 cells with the indicated chemicals. Lanes 6 and 9 display the results of in vivo treatment of CdCl2-induced NIH3T3 cells with the indicated chemicals. Reference markers for both panels are Maxam and Gilbert G + A (lane 1) and C > T (lane 2) ladders. Lane 3 in both panels indicates results for purified genomic DNA samples treated with piperidine alone. Red bars and circles to the left of lanes 4 and 7 in each panel indicate regions or specific bases, respectively, that show variation between the in vitro and in vivo treated samples. Blue circles highlight differences between the in vivo CdCl2-induced and uninduced samples. Samples were analyzed on a 8% denaturing polyacrylamide gel.

Differences in DMS reactivity on both DNA strands of the MT-I promoter were also noted when in vitro and in vivo samples were compared (Fig. 3, A and B, compare lanes 7 and 8). Both hyper-reactivities and protections were observed (Fig. 3, A and B, red bars and circles). Metal induction again resulted in subtle differences between samples probed before or after CdCl₂ induction (Fig. 3, A and B, compare lanes 8 and 9, blue circles). Importantly, the DMS reactivity patterns were consistent with previous in vivo footprinting data for this region (24). Taken together the data obtained using KMnO₄ and DMS probes reassured us that our in vivo footprinting protocol was sensitive and reproducible.

Chemical Probing of the MT-I R/Y Sequence-- For analysis of the pyrimidine strand of the MT-I R/Y sequence we first analyzed plasmid pMTCAT with OT or KMnO₄ in vitro (Fig. 4A). In these studies, supercoiled or HpaI-linearized pMTCAT was chemically treated at either acidic or neutral pH. We first wished to determine if LMPCR analysis would confirm the presence of an unusual DNA structure at low pH in supercoiled DNA as has been previously detected by chemical treatment followed by radiolabeling of fragments (20). OT hyper-reactivity was indeed observed when supercoiled pMTCAT DNA was treated at pH 4.5 (Fig. 4A, lane 3, red bar). The OT hyper-reactivity was greatly reduced at pH 7.1 (Fig. 4A, lane 5), and OT reactivity was lost when the plasmid was linearized (Fig. 4A, lanes 4 and 6). Changes in KMnO₄ reactivity were also observed when pMTCAT was treated under these different conditions. A region of hyper-reactivity was observed between bases 1247 and 1205 when supercoiled pMTCAT was treated at pH 4.5 (Fig. 4A, lane 7, blue bar). The KMnO₄ hyper-reactivity was reduced when the plasmid was linearized and probed at pH 4.5 or when supercoiled plasmid was treated at pH 7.1 (Fig. 4A, lanes 8 and 9). When linearized pMTCAT DNA was treated with KMnO₄ at pH 7.1 a more uniform pattern of thymine modification was observed over the length of the R/Y sequence (Fig. 4A, lane 10).

View larger version (82K): [in this window] [in a new window]   Fig. 4.   LMPCR analysis of the pyrimidine strand of the MT-I R/Y sequence. Samples were analyzed on a 6% denaturing polyacrylamide gels. MT-I R/Y sequence boundaries and positions relative to the major MT-I tsp are labeled (left of panels A and B). Reference markers are Maxam and Gilbert G + A and C > T ladders. The single interruption in the R/Y sequence is indicated by a "G" (left of panels A and B). A, DNA reactivity to OT and KMnO4 in vitro and in vivo. Lanes 3-6 indicate relative reactivities to OT. Lanes 7-13 indicate relative reactivities to KMnO4. Plasmid samples were treated as either supercoiled (S) or HpaI-linearized (L) molecules at pH 4.5 or 7.1. Lanes 11-13 contain genomic DNA samples. Lane 11 displays the in vitro treatment of purified genomic DNA with KMnO4. Lanes 12 and 13 display the in vivo KMnO4 treatment of uninduced or CdCl2-induced cells, respectively. The red bar to the right of lane 3 indicates thymines hyper-reactive to OT modification in supercoiled plasmid DNA at low pH. The blue bar to the right of lane 7 indicates thymines hyper-reactive to KMnO4 modification under the same conditions. Thymines within the R/Y sequence (horizontal red ticks) and thymines upstream of the R/Y sequence (horizontal blue ticks) are indicated (right of lane 13). B, CAA reactivities superimposed on Maxam and Gilbert G + A ladders. Lanes 3-6 are in vitro treated pMTCAT plasmid samples. Plasmid samples were treated as either supercoiled (S) or HpaI-linearized (L) molecules at pH 4.5 or 7.1. Lanes 7-9 contain genomic DNA samples. Lane 7 displays the in vitro CAA treatment of purified genomic DNA. Lanes 8 and 9 display the in vivo CAA treatment of uninduced or CdCl2-induced cells, respectively. The black bar to the right of lane 3 indicates cytosines slightly hyper-reactive to CAA modification in supercoiled plasmid at low pH.

When genomic DNA samples were treated with KMnO₄ in vitro, a similar pattern of uniform thymine modification was observed. Thus for KMnO₄ treatment at pH 7.1, genomic DNA treated in vitro and linearized pMTCAT plasmid samples treated in vitro yielded similar patterns of reactivity (Fig. 4A, lanes 10 and 11). KMnO₄ probing of the MT-I R/Y sequence in cultured NIH3T3 cells showed no changes upon metal induction (Fig. 4A, compare lanes 12 and 13) and yielded a similar pattern of thymine reactivity to genomic DNA samples treated in vitro (Fig. 4A, lanes 11-13). In all cases, the KMnO₄ reactivity pattern for thymine residues is quite consistent: KMnO₄ reactivity is limited to isolated thymines and to the 3' terminal thymine of a string of consecutive thymines. This uniform pattern of thymine reactivity persisted throughout the R/Y sequence, whereas thymine residues upstream and downstream from the R/Y sequence reacted with KMnO₄ in a less uniform manner (Fig. 4A, lane 13, compare horizontal red ticks with horizontal blue ticks).

Plasmid pMTCAT and genomic DNA samples were also analyzed by CAA treatment followed by Maxam and Gilbert G+A chemistry (Fig. 4B, lanes 3-9). When pMTCAT was analyzed after treatment with CAA at pH 4.5 or 7.1, differences in reactivities were detected (Fig. 4B, compare lanes 3-6). CAA modifications were reduced when pMTCAT was linearized (Fig. 4B, lanes 4 and 6). Hyper-reactivities (vertical bar near lane 3 of Fig. 4B) include cytosines adjacent to reactive thymines that had been identified by OT (Fig. 4A, lane 3). These CAA data from purified supercoiled plasmid DNA treated at acidic pH again confirm the presence of an unpaired, non-B DNA structure within the R/Y sequence. It is evident that OT (Fig. 4A, lane 3) is a better probe of base unpairing within the element in vitro than CAA (Fig. 4B, lane 3), confirming our previous observations (20). CAA in vivo analysis suggested no regions of hyper-reactivity before (Fig. 4B, lane 8) or after (Fig. 4B, lane 9) metal induction, further confirming that this sequence did not have an unpaired configuration in living cells.

The chemical reactivities of thymines within the MT-I R/Y sequence are compared quantitatively in Fig. 5. Regions of hyper-reactivity were noted within the 3' half of the R/Y sequence when pMTCAT was treated with OT or KMnO₄ at low pH (Fig. 5, red and blue bars). In contrast, when NIH3T3 cells were treated with KMnO₄, a more uniform thymine reactivity was observed across the R/Y sequence with no asymmetry (Fig. 5, bottom).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Comparison of MT-I R/Y sequence reactivity. Storage phosphor analysis of data in Fig. 4A, lane 3 (top scan), lane 7 (middle scan), and lane 12 (bottom scan). Relative reactivity is indicated on the y axis. The MT-I R/Y sequence boundary is indicated below the x axis. The top scan depicts supercoiled plasmid pMTCAT reactivity with OT at pH 4.5. The middle scan depicts supercoiled plasmid pMTCAT reactivity with KMnO4 at pH 4.5. The bottom scan depicts genomic DNA reactivity with KMnO4 under uninduced in vivo conditions. The red bar (top) corresponds to the bar in Fig. 4A (lane 3). The blue bar corresponds to the bar in Fig. 4A (lane 7).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characteristics of the MT-I R/Y Sequence-- There are several examples of H-DNA-like structures in eukaryotic and prokaryotic cells (40-44). In one case, E. coli cells containing plasmid DNA with an inserted homopurine/homopyrimidine mirror repeats were grown under acidic conditions and treated with cell-permeable chemicals that react with distorted DNA. Extracted plasmid DNA indicated the formation of a pyrimidine motif H-DNA structure (42). Kohwi and Panchenko (44) showed that a G·G·C (purine motif) triplex formed in living E. coli cells following transcriptional activation of a downstream gene. In eukaryotic cells, isolated nuclei have been shown to bind monoclonal antibodies thought to be specific for triplex DNA, thus indicating the possible presence of such structures (40, 41). While intriguing, these data also raise the question of whether such examples are a true measure for the occurrence of such structures within the genomic DNA of living cells. Studies with E. coli often involve plasmid DNA under unusual conditions, and studies with monoclonal antibodies involve isolated nuclei. The present study sought to determine if a non-B DNA structure could be detected in living cells at base pair resolution in an element previously shown to adopt a pyrimidine motif H-DNA structure in vitro (19, 20). Although the MT-I R/Y sequence showed a pH dependence in vitro, we felt in vivo analysis was justified because additional factors in vivo (e.g. high negative supercoiling) might aid in the formation and stabilization of such a structure. Such an effect has recently been reported for cruciform extrusion (36).

The best studied examples of R/Y DNA sequences capable of adopting H-DNA structures in vitro possess perfect mirror symmetry and are often repetitive in nature (8, 9, 45-50). In contrast, the MT-I R/Y sequence is a 128-bp R/Y sequence composed of multiple imperfect mirror repeats, but lacking any substantial region of perfect mirror symmetry. The extreme length and interesting location of the R/Y sequence raise the possibility that the element did not originate by a random process, and that the sequence may have some biological function. When present in a supercoiled plasmid at acidic pH, this R/Y sequence has been shown to adopt a non-B DNA structure with unpaired bases (19, 20). Previous experiments have also shown that the R/Y sequence displays no detectable activity as a cis-acting transcriptional regulatory element (20). We examined whether base unpairing suggestive of a stable, non-B DNA structure could be detected within the endogenous mouse MT-I R/Y sequence in cultured fibroblasts before or after metal induction.

Structural Analysis of the MT-I R/Y Sequence-- We began by using chemical probing and LMPCR to characterize the in vitro structure of purified plasmid containing the R/Y sequence. The pattern of OT reactivity for purified supercoiled plasmid agrees with previous results indicating the presence of base unpairing within this sequence at acidic pH (Fig. 4). In vitro analysis with KMnO₄ also indicated a non-B DNA structure. However, comparison of the OT and KMnO₄ reactivities of supercoiled DNA at low pH reveal differences. Although the regions of hyper-reactivity overlap, KMnO₄ reacts with a greater number of thymines than OT (Figs. 4 and 5, compare red and blue bars). The broader pattern of KMnO₄ reactivity relative to OT presumably reflects different reaction mechanisms and/or preferences for slightly different structural distortions (40, 51, 52). OT chelated by 2,2'-bipyridyl creates a bulkier complex than the permanganate anion, but both reagents are reportedly specific for single-stranded and/or unstacked thymines (38, 40, 51, 52). As previously reported (20), CAA reacts only slightly with unpaired cytosines adjacent to unpaired thymines that are strongly modified by OT and KMnO₄ in supercoiled plasmids at low pH (Fig. 4B, black bar to the right of lane 3).

In vivo analysis of the MT-I R/Y sequence with KMnO₄ and CAA provided no evidence of base unpairing as had been observed with purified plasmids (Fig. 4). When the KMnO₄ and CAA reactivities of isolated genomic DNA samples were compared with the reactivities observed in vivo, no suppressions suggestive of protein footprints were detected, nor were hyper-reactivities observed, arguing against base unpairing or H-DNA formation. The KMnO₄ and CAA reactivities of the MT-I R/Y sequence did not change after metal induction. Because our initial LMPCR analysis of the TATA region of the MT-I promoter confirmed the activity and specificity of the relevant chemical probes, and because analysis of MT-I mRNA levels indicated that the metal induction protocol was effective, we believe that the failure to detect footprints or base unpairing within the R/Y sequence was not the result of technical limitations.

KMnO₄ Reactivity within an R/Y Sequence-- Although stable unpairing of thymines in the R/Y sequence was not detected in vivo, we were intrigued by the pattern of KMnO₄ reactivity of thymines in this element. When we analyzed linearized plasmid DNA with KMnO₄ at pH 7.1 we expected the sequence to be resistant to KMnO₄ modification because the DNA was neither supercoiled nor protonated. In contrast, we observed a consistent pattern of thymine reactivity across the entire R/Y sequence (Fig. 4A, lane 10). This pattern did not suggest base unpairing since it occurred in relaxed DNA and at neutral pH. Furthermore, probing with the single-strand-specific reagent CAA did not show evidence of overt base unpairing. The observed KMnO₄ reactivity pattern involves oxidation of isolated thymines, and of only the 3' thymine residue in strings of consecutive thymines. This peculiar pattern was also observed for genomic DNA samples treated with KMnO₄ in vitro or in vivo, and suggests that the long R/Y sequence confers an unusually uniform structure upon thymine residues. The observed pattern of thymine reactivity to KMnO₄ within the R/Y sequence presumably reflects subtle differences in thymine accessibility. The greatest KMnO₄ reactivity occurs when thymines are immediately upstream from cytosines. In contrast, thymines upstream from thymines are not modified by KMnO₄. This reactivity pattern is consistent throughout the R/Y sequence both in vitro and in vivo. In contrast, the KMnO₄ reactivities of thymines upstream of the R/Y sequence are more variable and unpredictable: not all isolated thymines or 3' terminal thymines in strings of consecutive thymines show uniform reactivity. Moreover, differences in KMnO₄ reactivity can be observed in vitro versus in vivo upstream of the R/Y sequence (Fig. 4A, compare lanes 11 and 13, horizontal blue ticks). These data imply that the local structure of thymines in long R/Y sequences is more uniform than in mixed sequences.

KMnO₄ is typically considered a chemical probe of unpaired DNA, selectively oxidizing the C5-C6 double bond of thymines (52). Attack on the C5-C6 double bond is thought to require approach from either above or below the plane of the base, and is therefore suppressed by base-stacking interactions (38, 51, 52). Although reactivity of the permanganate anion with DNA has been shown to be dependent on the ionic strength, similar band intensities were observed in vitro and in vivo (e.g. Fig. 4, compares lanes 10-13). What then is the structural basis for the very consistent context specificity of KMnO₄ modification of thymines within the R/Y sequence? Two possible explanations have previously been suggested. McCarthy et al. (39) and Nejedly et al. (38) observed irregular KMnO₄ reactivity within strings of consecutive thymines, such that only the 3' terminal thymine was attacked (39). These authors concluded that A tracts within mixed DNA sequences induce a unique B' helical structure and that the 3' junction of such an A tract is accompanied by a distortion that promotes KMnO₄ attack (39). Although a plausible explanation for KMnO₄ reactivity in thymine strings, this reasoning fails to account for our observation that isolated thymines are as reactive to KMnO₄ as the 3' terminal thymines in strings of consecutive thymines. Moreover, even in TT dinucleotide sequences, KMnO₄ modifies only the 3' thymine despite the fact that one or two isolated thymines are not thought to adopt B' helix structures.

Hunter and Lu (53) used theoretical calculations and x-ray crystal structures to examine the role of base stacking interactions in local DNA structure. These authors considered dinucleotide steps in terms of step parameters (twist, rise, and tilt). They concluded that AA steps are most like B-DNA and are very inflexible, AG and GG steps are slightly untwisted relative to B-DNA, and GC steps are slightly overtwisted compared with B-DNA (53). Because our data shows that thymines in the GA step are most reactive to KMnO₄, it might be proposed that the slight overtwisting of GA base pair steps promotes accessibility of the 5' thymine to KMnO₄ modification. Suzuki et al. (54) also analyzed structural data bases to understand additional sequence-dependent conformational aspects of DNA. These authors again noted that the GA dinucleotide step is slightly overtwisted and suggested that the overtwisting is a result of steric hindrance introduced by the thymine methyl. This work concluded that an AA step is "locked" into B-form due to repulsion between the methyl groups and the phosphate backbone (54).

Besides explanations involving A tract distortion or base pair twist, we propose an alternative simple steric model to account for the pattern of thymine reactivity to KMnO₄ observed so consistently in the MT-I R/Y sequence. Because attack on a thymine C5C6 double bond requires approach out of the plane of the base, an incoming trajectory within the major groove from 3' side of the thymine seems most favorable. The structural features of the base 3' to the target thymine thus may be critical in determining the exposure of the C5C6 double bond. In particular, inspection of molecular models shows that the methyl group at the C5 position of a 3' thymine substantially occludes access to the C5C6 double bond of a 5' thymine (Fig. 6, white arrow). In contrast, the degree of occlusion due to the other possible 3' bases is reduced (Fig. 6). These considerations suggest that the hyper-reactivity of isolated thymines and 3' terminal thymines in strings of consecutive thymines are due to C5 methyl footprinting of 3' thymines on adjacent 5' thymines.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Base stacks and occlusion of thymines. Consecutive stacked bases of dinucleotide steps involving a 5' thymine in B-DNA (green), viewed from the 3' side showing the overlap of the various possible 3' bases (magenta) with the C5C6 double bond of the 5' thymine residue (orange). White arrow indicates occlusion of the C-6 atom of a target thymine by the C-5 methyl group of a 3' thymine. Models were generated using Insight II software (MSI) and standard B-DNA parameters.

We conclude that despite the presence of a consistent pattern of KMnO₄ reactivity with the thymines in the MT-I R/Y sequence in vitro and in vivo, there is no evidence that the sequence adopts a stable, unpaired structure or interacts with proteins in vivo. Together with our previous genetic data (20), this result emphasizes the fact that although long R/Y sequences may be statistically over-represented and provocatively located in genomes, it cannot be assumed that they form unusual structures in vivo. Indeed, their function, if any, remains elusive.

	ACKNOWLEDGEMENTS

We thank Dr. Gerd Pfeifer for generous expert technical training in LMPCR and the staff of the Mayo Molecular Biology Core Facility for assistance.

	FOOTNOTES

^* This work was supported by the Mayo Foundation and National Institutes of Health Grants GM 47814 and GM 54411.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Mayo Foundation, 200 First St. Southwest, Rochester, MN 55905. Tel.: 507-284-9041; Fax: 507-284 -2053; E-mail: maher@mayo.edu.

Published, JBC Papers in Press, September 13, 2000, DOI 10.1074/jbc.M909658199

	ABBREVIATIONS

The abbreviations used are: LMPCR, ligation-mediated polymerase chain reaction; CAA, chloroacetaldehyde; DMS, dimethyl sulfate; KMnO₄, potassium permanganate; MRE, metal responsive element; MT-I, metallothionein-I; R/Y, homopurine/homopyrimidine; bp, base pair(s); OT, osmium tetroxide; PBS, phosphate-buffered saline; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This Article Abstract Full Text (PDF) All Versions of this Article: 275/51/40218    most recent M909658199v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Becker, N. A. Articles by Maher, L. J. Search for Related Content PubMed PubMed Citation Articles by Becker, N. A. Articles by Maher, L. J., III Social Bookmarking                      What's this?