A spectroscopic and electron microscopic examination of the highly condensed DNA structures formed by denaturation in Mg(ClO4)2.

1. Thermal denaturation in 1.5 M Mg(ClO4)2 of the DNA from bacteriophage lambda results in four well-separated subtransitions, as monitored by the accompanying increase in absorbance. The midpoint of the hyperchromic spectrum is significantly lowered compared to either 1.5 M MgCl2 or 3.0 M NaClO4. 2. The first two subtransitions are associated with the melting of the A . T-richest regions of the lambda DNA, as revealed by electron micrographs following fixation with formaldehyde. 3. Commencing with the third subtransition, an unusual DNA structure is observed in electron micrographs. In this structure the A . T-rich half of the molecule appears completely condensed, whereas the G . C-rich half remains native. 4. During the fourth subtransition DNA molecules condense completely and eventually aggregate to form extremely high molecular weight particles containing centers of electron density. Tendrils of DNA, primarily duplex, radiate outward from these centers. 5. The aggregation may be reversed by the removal of magnesium. The intramolecular condensation may be at least partly reversed by increasing the Mg(ClO4)2 concentrations to saturating levels.

Molecular cloning of the replication terminus of the plasmid R6K.

Analyses of the intermediates of DNA replication of the R6K plasmid derivatives, RSF1040, RJHC12 and RJHC26 demonstrate the transient accumulation of open circular DNA molecules with a discontinuity in either the plus or the minus strand of the DNA. The location of this discontinuity is nonrandom and is near the terminus of DNA replication. The discontinuity is not due to the activation of a relaxation complex since neither RJHC12 nor RJHC26 are relaxable. The replication terminus is functional in a clone containing approx. a 2000 bp sub-fragment of the HindIII-2 fragment of R6K. The replication terminus temporarily arrests the progression of replication fork of the unidirectionally replicating plasmid pBR313 at a region approx. 800 bp from HindIII site located nearest to the BamHI site of the vector. Subcloning experiments reveal that the upper limit of the replication termination sequence is 216 bp in length.

The nucleotide sequence surrounding the origin of DNA replication of Col E1.

The DNA of Col E1 replicates from a unique origin located at a distance of 17-19% of the genome length from the single Eco RI clevage site. The nucleotide sequence about this site has been determined by a combination of RNA and DNA sequencing techniques. The principal features of the sequence are two palindromes, one of which resembles a palindrome located in the intercistronic region of 0X174. The sequence also contains stretches of purine and pyrimidine clusters of the following compositions: pAT5G, pC2T5G, pGT5G. The origin sequence demonstrates that initiation of DNA replication takes place in an intercistronic region of Col E1DNA, although the possibility that this region makes small polypeptides 30-40 residues long cannot be strictly eliminated at this time.

A replication origin is turned off by an origin-"silencer" sequence.

The chromosome of R6K contains multiple origins of replication. The origin gamma is infrequently used in the original plasmid and remains "silent" in certain miniplasmid derivatives. The inactivation of the origin is caused by a natural origin silencer located adjacent to the minimal ori gamma sequence. The silencer functions in cis and has no trans activity. It has functional polarity and works only in one orientation when present immediately downstream from ori gamma. The silencer apparently initiates an RNA that invades ori gamma and turns it off either by competing with a primer RNA or by disrupting ori gamma structure. As predicted, removal of the silencer blocks the synthesis of silencer RNA and derepresses the origin.

Rapid purification of a cloned gene product by genetic fusion and site-specific proteolysis.

We have developed a rapid and general technique for purification of a protein encoded by a cistron contained in a recombinant DNA clone. The technique consists of fusing the target cistron DNA in the correct reading frame to a marker cistron via a piece of DNA that codes for a linker peptide. The target cistron in the example presented here is the replication initiator cistron of the plasmid R6K. The linker is a DNA fragment encoding 60 amino acids from the triple helical region of chicken pro alpha-2 collagen, and the marker cistron encodes the beta-galactosidase protein of Escherichia coli. The tripartite hybrid protein was rapidly purified by selective binding to and elution from a beta-galactosidase specific-affinity column. The hybrid protein was then digested with a purified microbial collagenase to cleave the linker, and high-pressure liquid chromatography allowed the rapid isolation of the target protein from the marker protein. Using this technique, we have purified the highly labile R6K replication initiator to homogeneity, and we have resolved the protein into NH2-terminal and COOH-terminal segments. We have further shown, by in vitro binding, that the COOH-terminal segment has at least one DNA-binding domain. The domain binds to the same restriction fragments of the R6K chromosome as the intact or beta-galactosidase-tagged initiator protein.

DNA-protein interaction at the origin of DNA replication of the plasmid pSC101.

The initiation of DNA replication of the low copy number plasmid pSC101 is dependent on the dnaA initiator protein encoded by Escherichia coli. We have previously reported that the minimum essential replicon of the plasmid encodes a approximately 37.5 kd protein and that the protein is necessary along with host-encoded proteins for the replication of the plasmid chromosome. In this communication we show that the plasmid-encoded protein has sequence-specific DNA-binding activity. The protein binds cooperatively to the replication origin of pSC101. Using chemical and enzymatic probes we have determined the contact points of the protein with the DNA and the precise domain of the replication origin recognized by the 37.5 kd protein. The specificity of the DNA-protein interaction would suggest that the 37.5 kd protein may possibly function by guiding the replisome to the correct DNA sequence on the chromosome of pSC101.

Replication initiator protein of plasmid R6K autoregulates its own synthesis at the transcriptional step.

The replication initiator protein of plasmid R6K preferentially repressed transcription initiated in vitro from the promoter of the initiator protein cistron. DNase I protection experiments revealed that the sequences in the region of the promoter recognized by the initiator protein partially overlapped the sequences of the same promoter recognized by RNA polymerase of Escherichia coli. Competitive DNase I protection experiments revealed that the initiator not only prevented the RNA polymerase from binding to the promoter sequence but also displaced RNA polymerase from preformed enzyme-promoter binary complexes. Thus, the initiator protein acts as a transcriptional repressor of its own cistron by either preventing RNA polymerase from binding to the promoter or by displacing RNA polymerase from promoter-enzyme complexes.

The replication initiator protein of plasmid pSC101 is a transcriptional repressor of its own cistron.

The plasmid-encoded replication initiator protein of pSC101 specifically repressed initiation of transcription of its own cistron from its natural promoter. Addition of the purified initiator had little or no visible effect on transcription initiated from a heterologous promoter. DNase protection experiments revealed that the RNA polymerase recognition sequence was overlapped by the initiator protein recognition sequences, which are vicinal to the replication origin. Using the labeled promoter sequence, we have performed competitive DNase protection experiments in two ways: by adding RNA polymerase and initiator protein simultaneously or by sequentially adding first RNA polymerase and then initiator protein to the DNase reaction mixture. The RNA polymerase protection pattern was recessive to that of the initiator regardless of whether the two proteins were added simultaneously or sequentially. This observation suggests that the mechanism of autoregulation is due to competition of the two proteins for the sequences in and around the promoter region. Furthermore, the sequential addition experiments raise the possibility of displacement of RNA polymerase from the promoter by the initiator protein.

DNA bending is induced in an enhancer by the DNA-binding domain of the bovine papillomavirus E2 protein.

The E2 gene of bovine papillomavirus type 1 has been shown to encode a DNA-binding protein and to trans-activate the viral enhancer. We have localized the DNA-binding domain of the E2 protein to the carboxyl-terminal 126 amino acids of the E2 open reading frame. The DNA-binding domain has been expressed in Escherichia coli and partially purified. Gel retardation and DNase I "footprinting" on the bovine papillomavirus type 1 enhancer identify the sequence motif ACCN6GGT (in which N = any nucleotide) as the E2 binding site. Using electrophoretic methods we have shown that the DNA-binding domain changes conformation of the enhancer by inducing significant DNA bending.

A replication terminus located at or near a replication checkpoint of Bacillus subtilis functions independently of stringent control.

We have examined a replication terminus (psiL1) located on the left arm of the chromosome of Bacillus subtilis and within the yxcC gene and at or near the left replication checkpoint that is activated under stringent conditions. The psiL1 sequence appears to bind to two dimers of the replication terminator protein (RTP) rather weakly and seems to possess overlapping core and auxiliary sites that have some sequence similarities with normal Ter sites. Surprisingly, the asymmetrical, isolated psiL1 site arrested replication forks in vivo in both orientations and independent of stringent control. In vitro, the sequence arrested DnaB helicase in both orientations, albeit more weakly than the normal Ter1 terminus. The key points of mechanistic interest that emerge from the present work are: (i) strong binding of a Ter (psiL1) sequence to RTP did not appear to be essential for fork arrest and (ii) polarity of fork arrest could not be correlated in this case with just symmetrical protein-DNA interaction at the core and auxiliary sites of psiL1. On the basis of the result it would appear that the weak RTP-L1Ter interaction cannot by itself account for fork arrest, thus suggesting a role for DnaB-RTP interaction.

Primary structure of the replication initiation protein of plasmid R6K.

The cistron of the replication initiation protein of plasmid R6K has been cloned into the single-strand DNA vectors M13mp8 and M13mp9 and its complete nucleotide sequence has been determined. The amino acid sequence of the initiator protein as predicted from its nucleotide sequence shows that the protein is lysine rich and weakly basic and has a molecular weight of 35,000, which is in close agreement with that estimated from the mobility in NaDodSO4/acrylamide gels. The secondary structure of the protein, approximately by the probabilistic methods of Chou and Fasman [Chou, P. & Fasman, G. (1978) Adv. Enzymol. 47, 45-148], suggests an NH2-terminal domain of primarily positively charged alpha-helical structure, a core region of interspersed short stretches of random coils and beta-sheets and -turns, and a COOH-terminal domain of alpha-helix.

Interaction of the plasmid R6K-encoded replication initiator protein with its binding sites on DNA.

Initiation of DNA replication in plasmid R6K is potentiated by the plasmid-encoded 35 kd replication initiator protein. We had previously reported that the initiator bound to two regions of R6K DNA called Site I and Site II. Using DNAase I footprinting technique we have demonstrated that the initiator bound to seven tandem repeats of a 22 bp long sequence in Site I. In Site II, the initiator bound to a single repeat having the same consensus sequence and to two partial repeats that most likely overlap the promoter of the initiation protein cistron. Using dimethyl sulfate as a chemical probe, we have determined the purine residues of Site I and Site II that make contact with the initiator protein. The results show that eight out of nine contact points per repeat in Site I were located on one of the two strands of the DNA. The binding of the initiator to the Site II sequence could explain the observed autoregulation of the synthesis of the initiator protein by promoter occlusion.

The replication initiator protein of plasmid R6K tagged with beta-galactosidase shows sequence-specific DNA-binding.

We have tagged the replication initiator protein of the plasmid R6K near the C-terminal end by fusion, in the correct reading frame, with the 89 amino acid long N-terminal alpha-donor polypeptide of beta-galactosidase of E. coli. This fusion was carried out with recombinant DNA methods. The protein chimera thus generated retained the activities of both initiation of DNA replication in vivo at the replication origin gamma of R6K and hydrolysis of beta-galactopyranoside when complemented in vivo with the alpha-acceptor polypeptide coded by the lac Z gene containing the M15 deletion. Using the simple and convenient assay for detecting beta-galactosidase, we have partially purified the tagged replication initiator, and have demonstrated that the protein binds to specific DNA sequences of the R6K chromosome. The protein bound to DNA sequences located at two places in the 5' untranslated leader region of the initiator protein cistron.

Primary structure of the essential replicon of the plasmid pSC101.

The replicon of the low copy number plasmid pSC101 has an obligatory requirement for the dnaA initiator protein of Escherichia coli as well as a plasmid-encoded initiator protein. We have identified the cistron of the plasmid-encoded initiator by DNA sequence analysis. Fusion of the initiator cistron with the lacZ gene of E. coli yielded a fusion protein of approximately equal to 150 kilodaltons, thus confirming that the open reading frame detected by DNA sequence analysis actually encoded a 37.5-kilodalton protein. Deletion of 26 amino acid residues from the COOH terminus of the plasmid initiator abolished autonomous replication from pSC101 origin. By in vitro deletion analysis we have shown that, although sequences downstream from the initiator cistron are dispensable, a maximum of 400 base pairs immediately upstream from the NH2-terminal region of the initiator is necessary for plasmid replication. These upstream sequences contain an A + T-rich region and three tandem repeats of a 21-base pair sequence; these features are characteristics of other replication origins.

Termination of DNA replication in vitro at a sequence-specific replication terminus.

The replication terminus of the drug resistance factor R6K has been cloned into the plasmid vectors pBR313 and pBR322. When the exogenously added DNA is replicated in vitro using cell extracts prepared from Escherichia coli, the plasmid replication terminus temporarily arrests the progression of the unidirectionally moving replication fork at or near the cloned terminator sequence. When the relative location of the terminator sequence is changed with respect to the replication origin, the point of arrest of the replication fork shifts correspondingly to the new location of the terminator. Termination of replication takes place in vitro regardless of whether the cell extracts used in the in vitro reaction are prepared from E. coli with a resident terminus sequence containing plasmid. From these observations we conclude that the termination of replication in vitro is identical or very similar to that observed in vivo, membrane association is not necessary for the activity of the replication terminus and the terminus sequence does not code for a transacting factor necessary for termination of replication. Therefore, any transacting factor which may be needed for the termination of replication must be coded by the host chromosome.

The E2 "gene" of bovine papillomavirus encodes an enhancer-binding protein.

The E2 early open reading frame (presumably gene) of bovine papillomavirus-1 was fused in frame with the collagen-beta-galactosidase-encoding region of the vector pJG200 and was expressed in and partially purified from Escherichia coli. The hybrid protein specifically bound to the enhancer region of bovine papillomavirus at several sites. DNase I-cleavage protection analysis of one such site revealed the protected sequence. A comparison of the protected sequence with the remainder of the DNA sequences that also have affinity for the protein revealed a consensus sequence having the motif AATTGGCGGNNCG, in which N is any nucleotide. The protected region also includes a sequence with 2-fold rotational symmetry--ATCGGTG/CACCGAT.