Contacts between gamma delta resolvase and the gamma delta res site.

We have investigated the interaction between resolvase and the res site of the transposon gamma delta by methylation and ethylation interference experiments. We have examined the effect of these DNA modifications both on binding and resolution in vitro. Major groove methylations within a 9 bp sequence that borders each site inhibit binding of resolvase to that site. Ethylation of certain phosphates within, and adjacent to, this border sequence inhibits binding. Together, these interference points define a contact region, present at all three res sites. In vitro resolution is inhibited only by modifications within site I. Inhibition of resolution by methylation of adenines at the center of site I suggests that minor groove contacts near the crossover may be required for resolution activity.     

DNA-Triesters - The Synthesis and Absolute Configuration Assignments at P-Stereogenic Centres

Abstract

Decadeoxyribonucleotide GGGAATTCCC and nine diastereomeric pairs of its mono-O-ethyl ester analogues were synthesized via phosphoramidite approach using the combination of 5′-DMT-base protected (except T) nucleoside 3′-(2-cyanoethyl N,N-diisopropyl phosphoramidites) and 3′-(0-ethyl N,N-diisopropyl phosphoramidites). Under conditions of release from solid support and removal of base-protecting groups (25% NH₄OH, 25°C, 48 h) 2-cyanoethyl groups were removed while O-ethyl phosphate triester functions were practically intact. Isolation of products and separation of diastereomers were performed by means of RP-HPLC. Absolute configuration at P-stereogenic centres was established via degradation of decamers into corresponding dinucleoside O-ethyl phosphates and stereochemical correlation with dinucleoside phosphorothioates of known configuration at phosphorus. Decadeoxyribonucleotide mono-O-ethyl esters were used for mapping the contact points between DNA and Eco RI endonuclease - the restriction enzyme which recognizes canonical sequence. GAATTC and cleaves unmodified DNA strands giving G and p AATTC.     

A complete cleavage map of Neurospora crassa mtDNA obtained with endonucleases Eco RI and Bam HI.

A physical map of Neurospora crassa mitochondrial DNA has been constructed using specific fragments obtained with restriction endonucleases. The DNA has 5 cleavage sites for endonuclease Bam HI, 12 for endonuclease Eco RI and more than 30 for endonuclease Hind III. The sequence of the Eco RI and Bam HI fragments has been established by analysis of partial fragments. By digestion of the Eco RI fragments with Bam HI, a complete overlapping map has been constructed. The position of the largest Hind III fragment on this map has also been determined. The map is circular and the added molecular weight of the fragments is 40 - 10(6), which is in good agreement with earlier measurements on intact DNA, using the electron microscope.     

NMR studies of the exocyclic 1,N6-ethenodeoxyadenosine adduct (epsilon dA) opposite deoxyguanosine in a DNA duplex. Epsilon dA(syn).dG(anti) pairing at the lesion site

Proton NMR studies are reported on the complementary d(C-A-T-G-G-G-T-A-C).d(G-T-A-C-epsilon A-C-A-T-G) nonanucleotide duplex (designated epsilon dA.dG 9-mer duplex), which contains exocyclic adduct 1,N6-ethenodeoxyadenosine positioned opposite deoxyguanosine in the center of the helix. The present study focuses on the alignment of dG5 and epsilon dA14 at the lesion site in the epsilon dA.dG 9-mer duplex at neutral pH. This alignment has been characterized by monitoring the NOEs originating from the NH1 proton of dG5 and the H2, H5, and H7/H8 protons of epsilon dA14 in the central d(G4-G5-G6).d(C13-epsilon A14-C15) trinucleotide segment of the epsilon dA.dG 9-mer duplex. These NOE patterns establish that epsilon dA14 adopts a syn glycosidic torsion angle that positions the exocyclic ring toward the major groove edge while all the other bases including dG5 adopt anti glycosidic torsion angles. We detect a set of intra- and interstrand NOEs between protons (exchangeable and nonexchangeable) on adjacent residues in the d(G4-G5-G6).d(C13-epsilon A14-C15) trinucleotide segment which establish formation of right-handed helical conformations on both strands and stacking of the dG5(anti).epsilon dA14(syn) pair between stable dG4.dC15 and dG6.dC13 pairs. The energy-minimized conformation of the central d(G4-G5-G6).d(C13-epsilon A14-C15) segment establishes that the dG5(anti).epsilon dA14(syn) alignment is stabilized by two hydrogen bonds from the NH1 and NH2-2 of dG5(anti) to N9 and N1 of epsilon dA14(syn), respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Factor D is a selective single-stranded oligodeoxythymidine binding protein.

Factor D, a protein purified from rabbit liver that selectively enhances traversal of template oligodeoxythymidine tracts by diverse DNA polymerases, was examined for the sequence specificity of its binding to DNA. Terminally [32P]-labeled oligomers with the sequence 5'-d[AATTC(N)16G]-3', N being dT, dA, dG, or dC, were interacted with purified factor D and examined for the formation of protein-DNA complexes that exhibit retarded electrophoretic mobility under nondenaturing conditions. Whereas significant binding of factor D to 5'-d[AATTC(T)16G]-3' is detected, there is no discernable association between this protein and oligomers that contain 16 contiguous moieties of dG, dA, or dC. Furthermore, factor D does not form detectable complexes with the duplexes oligo(dA).oligo(dT) or poly(dA).poly(dT). The preferential interaction of factor D with single-stranded poly(dT) is confirmed by experiments in which the polymerase-enhancing activity of this protein is protected by poly(dT) against heat inactivation two- and four-fold more efficiently than by poly(dA) or poly(dA).poly(dT), respectively.     

Solution structure of the (+)-cis-anti-benzo[a]pyrene-dA ([BP]dA) adduct opposite dT in a DNA duplex.

Minor adducts, derived from the covalent binding of anti-benzo[a]pyrene-7,8-dihydroxy-9,10-epoxide to cellular DNA, may play an important role in generating mutations and initiating cancer. We have applied a combined NMR-computational approach including intensity based refinement to determine the solution structure of the minor (+)-cis-anti-[BP]dA adduct positioned opposite dT in the d(C1-T2-C3-T4-C5-[BP]A6-C7-T8-T9-C10-C11). (d(G12-G13-A14-A15-G16-T17-G18-A19-G20+ ++-A21-G22) 11-mer duplex. The BP ring system is intercalated toward the 5'-side of the [BP]dA6 lesion site without disrupting the flanking Watson-Crick dC5.dG18 and [BP]dA6.dT17 base pairs. This structure of the (+)-cis-anti-[BP]dA.dT 11-mer duplex, containing a bay region benzo[a]pyrenyl [BP]dA adduct, is compared with the corresponding structure of the (+)-trans-anti-[BPh]dA.dT 11-mer duplex (Cosman et al., Biochemistry 32, 12488-12497, 1993), which contains a fjord region benzo[c]phenanthrenyl [BPh]dA adduct with the same R stereochemistry at the linkage site. The carcinogen intercalates toward the 5'-direction of the modified strand in both duplexes (the adduct is embedded within the same sequence context) with the buckling of the Watson-Crick [BP]dA6.dT17 base pair more pronounced in the (+)-cis-anti-[BP]dA.dT 11-mer duplex compared to its Watson-Crick [BPh]dA.dT17 base pair in the (+)-trans-anti-[BPh]dA.dT 11-mer duplex. The available structural studies of covalent polycyclic aromatic hydrocarbon (PAH) carcinogen-DNA adducts point toward the emergence of a general theme where distinct alignments are adopted by PAH adducts covalently linked to the N(6) of adenine when compared to the N(2) of guanine in DNA duplexes. The [BPh]dA and [BP]dA N(6)-adenine adducts intercalate their polycyclic aromatic rings into the helix without disruption of their modified base pairs. This may reflect the potential flexibility associated with the positioning of the covalent tether and the benzylic ring of the carcinogen in the sterically spacious major groove. By contrast, such an intercalation without modified base pair disruption option appears not to be available to [BP]dG N(2)-guanine adducts where the covalent tether and the benzylic ring are positioned in the more sterically crowded minor groove. In the case of [BP]dG adducts, the benzopyrenyl ring is either positioned in the minor groove without base pair disruption, or if intercalated into the helix, requires disruption of the modified base pair and displacement of the bases out of the helix.     

Mutational analysis of two putative catalytic motifs of the type IV restriction endonuclease Eco57I

The role of two sequence motifs (SM) as putative cleavage catalytic centers (77)PDX(13)EAK (SM I) and (811)PDX(20)DQK (SM II) of type IV restriction endonuclease Eco57I was studied by site-directed mutational analysis. Substitutions within SM I; D78N, D78A, D78K, and E92Q reduced cleavage activity of Eco57I to a level undetectable both in vivo and in vitro. Residual endonucleolytic activity of the E92Q mutant was detected only when the Mg(2+) in the standard reaction mixture was replaced with Mn(2+). The mutants D78N and E92Q retained the ability to interact with DNA specifically. The mutants also retained DNA methylation activity of Eco57I. The properties of the SM I mutants indicate that Asp(78) and Glu(92) residues are essential for cleavage activity of the Eco57I, suggesting that the sequence motif (77)PDX(13)EAK represents the cleavage active site of this endonuclease. Eco57I mutants containing single amino acid substitutions within SM II (D812A, D833N, D833A) revealed only a small or moderate decrease of cleavage activity as compared with wild-type Eco57I, indicating that the SM II motif does not represent the catalytic center of Eco57I. The results, taken together, allow us to conclude that the Eco57I restriction endonuclease has one catalytic center for cleavage of DNA.     

Effect of temperature and 4,6'-diamidine-2-phenylindole on restriction of supercoiled Col E1 DNA by Eco RI endonuclease.

Supercoiled Col E1 DNA is split by Eco RI endonuclease at 37 degrees C without intermediate formation of open circular DNA. Accumulation of this restriction product is observed at low temperature. The fluorescent dye, 4,6'-diamidine-2-phenylindole (DAPI) inhibits restriction by Eco RI endonuclease. This effect is due to the DAPI:DNA rather than to the DAPI:Eco RI interactions.     

Cleavage by restriction endonucleases MvaI and EcoRII of substrates modified in amino groups of heterocyclic bases

14-membered DNA-duplexes containing modified nucleoside residues, viz 4-N-methyldeoxycytidine (m4dC), 6-N-methyldeoxyadenosine (m6dA) or deoxyinosine (dI), in only one strand of the recognition site (CCA/TGG) of MvaI and EcoRII endonucleases were synthesized. It was shown that MvaI and EcoRII endonucleases interact with the exocyclic amino groups of the external dC residues and of the central dA residue of the recognition site exposed into the DNA major groove. These endonucleases which are isochizomers were found to possess different mechanisms of substrate cleavage. The ability of MvaI endonuclease to hydrolyze only unmodified strand of methylated duplexes allows one to make site-directed single-strand nicks in double-stranded DNA. Elimination of the 2-NH2-group located in the minor groove of DNA by substituting dI for dG had little, if any, effect on the hydrolytic activity of EcoRII and MvaI endonucleases.     

Solution structures of aminofluorene [AF]-stacked conformers of the syn [AF]-C8-dG adduct positioned opposite dC or dA at a template-primer junction.

A solution structural study has been undertaken on the aminofluorene-C8-dG ([AF]dG) adduct located at a single-strand-double-strand d(A1-A2-C3-[AF]G4-C5-T6-A7-C8-C9-A10-T11-C12-C13). d(G14-G15-A16-T17-G18-G19-T20- A21-G22-N23) 13/10-mer junction (N = C or A) using proton-proton distance restraints derived from NMR data in combination with intensity-based relaxation matrix refinement computations. This single-strand-double-strand junction models one arm of a replication fork composed of a 13-mer template strand which contains the [AF]dG modification site and a 10-mer primer strand which has been elongated up to the modified guanine with either its complementary dC partner or a dA mismatch. The solution structures establish that the duplex segment retains a minimally perturbed B-DNA conformation with Watson-Crick hydrogen-bonding retained up to the dC5.dG22 base pair. The guanine ring of the [AF]dG4 adduct adopts a syn glycosidic torsion angle and is displaced into the major groove when positioned opposite dC or dA residues. This base displacement of the modified guanine is accompanied by stacking of one face of the aminofluorene ring of [AF]dG4 with the dC5.dG22 base pair, while the other face of the aminofluorene ring is stacked with the purine ring of the nonadjacent dA2 residue. By contrast, the dC and dA residues opposite the junctional [AF]dG4 adduct site adopt distinctly different alignments. The dC23 residue positioned opposite the adduct site is looped out into the minor groove by the aminofluorene ring. The syn displaced orientation of the modified dG with stacking of the aminofluorene and the looped out position of the partner dC could be envisioned to cause polymerase stalling associated with subsequent misalignment leading to frameshift mutations in appropriate sequences. The dA23 residue positioned opposite the adduct site is positioned in the major groove with its purine ring aligned face down over the van der Waals surface of the major groove and its amino group directed toward the T6.A21 base pair. The Hoogsteen edge of the modified guanine of [AF]dG4 and the Watson-Crick edge of dA23 positioned opposite it are approximately coplanar and directed toward each other but are separated by twice the hydrogen-bonding distance required for pairing. This structure of [AF]dG opposite dA at a model template-primer junctional site can be compared with a previous structure of [AF]dG opposite dA within a fully paired duplex [Norman, D., Abuaf, P., Hingerty, B. E., Live, D. , Grunberger, D., Broyde, S., and Patel, D. J. (1989) Biochemistry 28, 7462-7476]. The alignment of the Hoogsteen edge of [AF]dG (syn) positioned opposite the Watson-Crick edge of dA (anti) has been observed for both systems with the separation greater in the case of the junctional alignment in the model template-primer system. However, the aminofluorene ring is positioned in the minor groove in the fully paired duplex while it stacks over the junctional base pair in the template-primer system. This suggests that the syn [AF]dG opposite dA junctional alignment can be readily incorporated within a duplex by a translation of this entity toward the minor groove.     

S-adenosyl methionine alters the DNA contacts of the EcoKI methyltransferase.

The EcoKI methyltransferase methylates two adenines on opposite strands of its bipartite DNA recognition sequence AAC(N6)GTGC. The enzyme has a strong preference for hemimethylated DNA substrates, but the methylation state of the DNA does not influence its binding affinity. Methylation interference was used to compare the contacts made by the EcoKI methyltransferase with unmodified, hemimethylated or fully modified DNAs. Contacts were seen at or near the N7 position of guanine, in the major groove, for all of the guanines in the EcoKI recognition sequence, and at two guanines on the edge of the intervening spacer sequence. The presence of the cofactor and methyl donor S-adenosyl methionine had a striking effect on the interference pattern for unmodified DNA which could not be mimicked by the presence of the cofactor analogue S-adenosyl homocysteine. In contrast, S-adenosyl methionine had no effect on the interference patterns for either kind of hemimethylated DNA, or for fully modified DNA. Differences between the interference patterns for the unmodified DNA and any of the three forms of methylated DNA provide evidence that methylation of the target sequence influences the conformation of the protein-DNA interface, and illustrate the importance of S-adenosyl methionine in the distinction between unmodified and methylated DNA by the methyltransferase.     

Synthesis of a dA-dT base pair analogue and its effects on DNA-ligand binding

Two nucleoside derivatives containing the base analogues 3-deazaadenine and 3-methyl-2-pyridone have been prepared as analogues of dA and dT, respectively. After conversion into the appropriately protected phosphoramidites, DNA sequences were prepared with site-specifically placed analogues. When present in a duplex DNA sequence, the analogues result in the deletion of one or both of the hydrogen bonding functional groups (the N3-nitrogen of dA and the O2-carbonyl of dT) present in the minor groove. Binding by two ligands, 4',6-diamidine-2-phenyl indole (DAPI) and Hoechst 33258 in the minor groove has been probed using a variety of DNA sequences. These sequences contain a d(GAATTC)2 core with analogue nucleosides substituted for one or more of the dA and dT residues. DAPI bound strongly to any sequence that contained both O2-carbonyls of the central two dT residues. The presence of a dc3A residue did in some cases enhance binding. With one of the central O2-carbonyls deleted, the binding was noticeably reduced, and with both absent, no significant binding could be detected. Similar although less dramatic results were observed with Hoechst 33258 binding to analogue sequences.     

Solution structure of the N-(deoxyguanosin-8-yl)-1-aminopyrene ([AP]dG) adduct opposite dA in a DNA duplex.

Solution structural studies have been undertaken on the aminopyrene-C(8)-dG ([AP]dG) adduct in the d(C5-[AP]G6-C7). d(G16-A17-G18) sequence context in an 11-mer duplex with dA opposite [AP]dG, using proton-proton distance and intensity restraints derived from NMR data in combination with distance-restrained molecular mechanics and intensity-restrained relaxation matrix refinement calculations. The exchangeable and nonexchangeable protons of the aminopyrene and the nucleic acid were assigned following analysis of two-dimensional NMR data sets on the [AP]dG.dA 11-mer duplex in H2O and D2O solution. The broadening of several resonances within the d(G16-A17-G18) segment positioned opposite the [AP]dG6 lesion site resulted in weaker NOEs, involving these protons in the adduct duplex. Both proton and carbon NMR data are consistent with a syn glycosidic torsion angle for the [AP]dG6 residue in the adduct duplex. The aminopyrene ring of [AP]dG6 is intercalated into the DNA helix between intact Watson-Crick dC5.dG18 and dC7.dG16 base pairs and is in contact with dC5, dC7, dG16, dA17, and dG18 residues that form a hydrophobic pocket around it. The intercalated AP ring of [AP]dG6 stacks over the purine ring of dG16 and, to a lesser extent dG18, while the looped out deoxyguanosine ring of [AP]dG6 stacks over dC5 in the solution structure of the adduct duplex. The dA17 base opposite the adduct site is not looped out of the helix but rather participates in an in-plane platform with adjacent dG18 in some of the refined structures of the adduct duplex. The solution structures are quite different for the [AP]dG.dA 11-mer duplex containing the larger aminopyrene ring (reported in this study) relative to the previously published [AF]dG.dA 11-mer duplex containing the smaller aminofluorene ring (Norman et al., Biochemistry 28, 7462-7476, 1989) in the same sequence context. Both the modified syn guanine and the dA positioned opposite it are stacked into the helix with the aminofluorene chromophore displaced into the minor groove in the latter adduct duplex. By contrast, the aminopyrenyl ring participates in an intercalated base-displaced structure in the present study of the [AP]dG.dA 11-mer duplex and in a previously published study of the [AP]dG.dC 11-mer duplex (Mao et al., Biochemistry 35, 12659-12670, 1996). Such intercalated base-displaced structures without hydrogen bonding between the [AP]dG adduct and dC or mismatched dA residues positioned opposite it, if present at a replication fork, may cause polymerase stalling and formation of a slipped intermediate that could produce frameshift mutations, the most dominant mutagenic consequence of the [AP]dG lesion.     

Heteroduplexes of phiX174 and G4 DNAs: orientation to genetic map and comparison with predictions from nucleotide sequences.

Heteroduplexes between the viral DNA of phiX174 and DNA from the replicative form (RF) of phage G4 were examined by electron microscopy. The single Eco RI site of G4-RF was utilized as a physical marker by preparing the heteroduplexes from the denatured, linear DNA obtained by restricting G4-RF with Eco RI endonuclease. Restriction fragments of phiX were used in a separate series of heteroduplexes to align the heteroduplex map and the G4 Eco RI site with the similar genetic maps of the two phages. The positions of the branch migrating junctions of recombinant phiX-G4 figure-8s, previously located only with respect to the G4-Eco RI site, have now been located with high proability within the gene A region of the two genomes. The degree of mismatch between the known nucleotide sequences of phi X and G4 accounts for positions of all of the regions of single-strandedness in the observed heteroduplexes, but unexplained discrepancies were also found.     

Studies on nucleic acid interactions. I. Stabilities of mini-duplexes (dG2A4XA4G2-dC2T4YT4C2) and self-complementary d(GGGAAXYTTCCC) containing deoxyinosine and other mismatched bases.

The thermal stability of DNA duplexes containing deoxyinosine in a pairing position in turn with each of the four major deoxynucleotides has been investigated by measuring ultraviolet-absorbance at different temperatures. d(G2A4 X A4G2) and d(C2T4YT4C2) were prepared by the solid-phase phosphotriester method. When X is deoxyinosine, the Tm values of the duplexes are in the order Y = dC greater than dA greater than dG greater than dT greater than dU. The Tm of other duplexes containing dG, dA and dT at X were also measured. Self-complementary duplexes d(GGGAAINTTCCC) showed the same order of stability with N being dC, dA, dG and dT. Thermal stabilities of duplexes containing dG instead of dI were compared with other matched and mismatched duplexes. The Tm values of sequence isomers containing purine-pyrimidine combinations were compared. Self-complementary duplexes containing G-C and A-T in the central positions showed Tm values ca. 10 degrees higher than those containing C-G and T-A in the same positions. Thermodynamic parameters and circular dichroism spectra of these oligonucleotides were compared.     

Evidence that populations of Dictyostelium single-copy mRNA transcripts carry common repeat sequences.

Two recombinant plasmids, M4 and KH10, carrying Dictyostelium DNA inserted into the Eco RI restriction endonuclease site of pMB9 by poly(dA)-poly(dT) tailing, were selected for study because they are complementary to abundant mRNA populations from Dictyostelium. Both plasmids have been shown to hybridize a heterogeneous size class of mRNAs which, in the case of KH10, comprise 5-10% of the pulse-labeled poly(A)+ RNA from vegetative cells. Analysis of the sequence organization of the two pieces of Dictyostelium DNA shows that they consist mostly of single-copy sequences with a short DNA sequence which is repeated in the genome and interspersed with single-copy DNA. These and other results suggest that the majority of the hybridization of pulse-labeled mRNA to M4 and KH10 is to the short "repeated" DNA sequences. In the genome, members of these repeat families appear to be transcribed onto a population of different single-copy mRNAs. Additional results show that M4 DNA contains a sequence which is entirely complementary to a discrete mRNA.     

A demonstration of the intrinsic importance of stabilizing hydrophobic binding and non-covalent van der Waals contacts dominant in the non-covalent CC-1065/B-DNA binding

The comparative DNA binding properties and cytotoxic activity of CDPIn methyl esters (n = 1-5) vs. PDE-In methyl esters (n = 1-3) are detailed in studies which provide experimental evidence for the intrinsic importance of stabilizing hydrophobic binding and non-covalent van der Waals contacts dominant in the CC-1065/B-DNA minor groove binding. High affinity minor groove binding to DNA was established through: (1) the observation of CDPI3 binding (UV) but not unwinding of supercoiled DNA (phi 174 RFI DNA) thus excluding intercalative binding; (2) the observation of CDPI3 binding to T4 phage DNA (UV, delta Tm) in which the major groove is occluded by glycosylation thus excluding major groove binding; (3) the observation of salt (Na+) concentration independent high affinity CDPI3 binding to poly(dA . poly(dT) thus excluding simple electrostatic binding to the DNA phosphate backbone; and further inferred through (4) the observation of an intense induced dichroism (ICD, poly(dA) . poly(dT) and poly(dG) . poly(dC) [phi]23(358) = 24,000 and 23,500). This high affinity minor groove binding is sufficient to produce a potent cytotoxic effect.(ABSTRACT TRUNCATED AT 400 WORDS)     

Probing the structure of gal operator-repressor complexes. Conformation change in DNA

The gal operon is regulated by binding of Gal repressor to two operator loci, OE and OI, which are separated by 114 base pairs (bp). We have probed the actual operator DNA segments with and without Gal repressor occupation by characterizing the regions protected by repressor from DNase I digestion and dimethyl sulfate methylation. The segments which are protected from DNase I digestion in both OE and OI are about 22 bp long and seem to include 2-3 extra bp on either side of a 16-bp similar sequence containing an approximate dyad symmetry, with a consensus half-symmetry sequence GTG(G/T)AA-C. Repressor occupation hinders the reactivity of the consensus guanines in the four half-symmetry sequences, as shown by retardation of methylation at the N-7 positions by dimethyl sulfate owing to repressor binding. The protected guanines are symmetrically located. Since a dimeric Gal repressor affects symmetrically located bases, it is consistent with the notion that each half-operator is occupied by a repressor subunit. Because the N-7 positions of methylation of guanines lie in the major grooves and the protected guanines are located at positions 1, 3, 8 and the rotational 1', 3', and 8' in the 16-bp dyad symmetry, we suggest that Gal repressor establishes direct contacts with bases at 1, 3, 1', and 3' through two major grooves lying on one face of an operator helix and prevents reactivity of the guanines at 8 and 8' of a third major groove on the opposite face by changing the DNA helical structure at this position. Contacts at other positions are also discussed.     

Tet repressor-tet operator contacts probed by operator DNA-modification interference studies

Contacts between tet operator DNA and Tet repressor protein are characterized by modification interference studies. The modified DNA fragments are separated into fractions with high, intermediate and low affinities for Tet repressor by polyacrylamide gel electrophoresis. Ethylation of the phosphates with N-ethylnitrosourea reveals 12 contacts of a repressor dimer to tet operator. Eight of these contacts appear to be important for Tet repressor binding, as judged by the strong interference at these positions, while four contacts are probably less important. All of the phosphate contacts are located on the same side of the B-DNA structure. The sequences of tet operators proposed to interact with the recognition alpha-helix of Tet repressor are TCTATC in three cases and CCTATC in one case. After methylation of N-7 with dimethylsulfate, strong interference is observed at the guanine residues at positions +/- 2. None of the N-7 functions of other guanine residues seems to be involved in tight contacts to Tet repressor. Tet repressor subunits form identical phosphate and guanine N-7 contacts with each half side of the two tet operators indicating twofold dyad symmetry of the complexes. Attempts to analyze the methylation interference at the adenine N-3 sites reveal different results for the operators. Modification of DNA fragments with diethylpyrocarbonate yields hypersensitive sites in the tet operators, indicating different local DNA structures. Carbethoxylation interference studies confirm the contacts at the purines found by methylation interference. All of the sequence-specific protein-DNA contacts detected in this study are centered at the inside four base-pairs in each tet operator half side. The contacts are discussed with respect to the structure of the repressor-operator complex.