Inactivation of transforming Bacillus subtilis deoxyribonucleic acid by monoadducts of 4,5',8-trimethylpsoralen.

4,5',8-Trimethylpsoralen (TMP) monoadducts inactive transforming deoxyribonucleic acid (DNA) in Bacillus subtilis. Contrary to TMP diadducts (TMP cross-links), which severely inhibit entry of donor DNA (G. Venema and U. Canosi, Mol. Gen. Genet. 179:1--11), TMP monoadducts have only a slight effect on entry. Since reextracted TMP-monoadduct-containing transforming DNA is a differentially repaired by Uvr- and Uvr+ recipients and cross-linkable to the recipient strand in the heteroduplex recombinant DNA molecules, the monoadducts can be integrated along with the donor DNA into the recipient chromosome.     

S1 nuclease does not cleave DNA at single-base mis-matches

Three assays have been designed to detect the cleavage of duplex phi X174 DNA at single-base mis-matches. Studies with S1 nuclease failed to detect cleavage at mis-matches. S1 nuclease digestion at 37 and 55 degrees C failed to produce a preferential degradation of a multiply mis-matched heteroduplex when compared to a mis-match-free homo-duplex as analyzed by sedimentation on sucrose gradients. Other heteroduplex templates were not cleaved by S1 nuclease at a defined single-base mis-match when assayed by gel electrophoresis or by marker rescue. In all cases, the amount of S1 nuclease employed was at least 10-times more than that required to render a single-stranded phi X174 DNA molecule completely acid soluble. The rate of hydrolysis of single-base mis-matches by S1 nuclease was estimated to be less than 0.016% of the rate at a base in single-strand phi X174 DNA. In no instance did we detect activity by S1 nuclease directed at mis-matched sites in our template molecules. Similarly, the single-strand specific endonuclease from Neurospora crassa does not cleave heteroduplex templates at a defined single-base mis-match when assayed by marker rescue.     

Deoxyribonucleic acid-protein and deoxyribonucleic acid interstrand cross-links induced in isolated chromatin by hydrogen peroxide and ferrous ethylenediaminetetraacetate chelates

DNA-protein and DNA interstrand cross-links were induced in isolated chromatin after treatment with H2O2 and ferrous ethylenediaminetetraacetate (EDTA). Retention of DNA on membrane filters after heating of chromatin in a dissociating solvent indicated the presence of a stable linkage between DNA and protein. Treatment of protein-free DNA with H2O2/Fe2+-EDTA did not result in enhanced filter retention. Incubation of cross-linked chromatin with proteinase K completely eliminated filter retention. Resistance to S1 nuclease after a denaturation-renaturation cycle was used to detect DNA interstrand cross-links. Heating the treated chromatin at 45 degrees C for 16 h and NaBH4 reduction enhanced the extent of interstrand cross-linking. The following data are consistent with, but do not totally prove, the hypothesis that cross-links are induced by hydroxyl radicals generated in Fenton-type reactions: (1) cross-linking was inhibited by hydroxyl radical scavengers; (2) the degree of inhibition of DNA interstrand cross-links correlated very closely with the rate constants of the scavengers for reaction with hydroxyl radicals; (3) cross-linking was eliminated or greatly reduced by catalase; (4) the extent of cross-linking was directly related to the concentration of Fe2+-EDTA. Partial inhibition of cross-linking by superoxide dismutase indicates that superoxide-driven Fenton chemistry is involved. The data indicate that DNA cross-linking may play a role in the manifestation of the biological activity of agents or systems that generate reactive hydroxyl radicals.     

Fate of homospecific transforming DNA bound to Streptococcus sanguis.

The fate of [3H]DNA from Streptococcus sanguis str-r43 fus-s donors in [14C]S. sanguis str-s fus-r1 recipients was studied by examining the lysates prepared from such recipients at various times after 1 min of exposure to DNA. The lysates were analyzed in CsCl and 10 to 30% sucrose gradients; fractions from the gradients were tested for biological activity and sensitivity to nucleases, subjected to various treatments and retested for nuclease sensitivity, and run on 5 to 20% neutral and alkaline sucrose gradients. The results demonstrate that donor DNA bound to S. sanguis cells in a form resistant to exogenous deoxyribonuclease is initially single stranded and complexed to recipient material. Donor DNA can be removed from the complex upon treatment of the complex with Pronase, phenol, or isoamyl alcohol-chloroform. Within the complex, donor DNA is relatively insensitive to S1 endonuclease but can regain its sensitivity by treatment with phenol. With time the complex moves as a whole to associate physically with the recipient chromosome. After a noncovalent stage of synapsis, donor material is covalently bonded to and acquires the nuclease sensitivity of recipient DNA, while donor markers regain transforming activity and become linked to resident markers.     

Transformation in Bacillus subtilis: characterization of an intermediate in recombination observed during transformation.

Summary From recombination-proficient competent cells of Bacillus subtilis in which the donor DNA entered at 17°, and which were kept at the same temperature, a complex of donor DNA and the recipient chromosome can be obtained which has a relatively high buoyant density in CsCl gradients. Exposure of the isolated complex to nuclease S1 liberates donor radioactivity. The limited biological activity of DNA re-extracted from cells attempting to recombine at 17° is decreased upon incubation with nuclease S1. If recombination is allowed to proceed at 30°, the high buoyant density of the donor-recipient complex decreases to normal values and less radioactivity can be liberated from the complex by nuclease S1. Concomitantly the biological activity of re-extracted DNA becomes less vulnerable to nuclease S1 under these conditions. On the basis of these observations we assume that the intermediate complex partly consists of unpaired single-stranded donor DNA.Support for the correctness of this assumption is derived from experiments with a mutant, which is delayed in the processing of high buoyant density donor-recipient complex to normal buoyant density donor-recipient complex. This delay is reflected in the time of acquisition of resistance to nuclease S1 digestion of the isolated complex.     

Formation of strand breaks and interstrand cross-links in DNA by methylglyoxal

Methylglyoxal (MG), a dietary mutagen, is present in various frequently consumed beverages and foods and in cigarette smoke. A combination of S1 nuclease hydrolysis and alkaline unwinding assay was used to demonstrate the formation of single-strand breaks and interstrand cross-links in DNA upon treatment with MG. Calf thymus DNA, when treated with increasing concentrations of MG, showed an increasing degree of S1 nuclease hydrolysis. It also showed the formation of an increasing number of strand breaks per molecule as determined by an alkaline unwinding assay. Incubation of DNA with relatively higher concentrations of methylglyoxal or prolonged treatment gave increased thermal melting temperatures and an enhanced rate of reannealing after thermal denaturation. These results indicated the formation of interstrand cross-links. Upon treatment with MG, A-T base pair depleted DNA showed a reduced number of single-strand break formation. It also showed a significantly lower decrease in Tm as compared with MG-treated normal DNA. These results showed that under the conditions used, MG primarily reacts with A-T base pairs in duplex DNA.     

Destruction of low efficiency markers is a slow process occurring at a heteroduplex stage of transformation

Summary Direct evidence is presented that the mechanism which discriminates against low efficiency markers in transformation of Diplococcus pneumoniae of genotype hex ⁺ acts on them after the formation of donor-recipient heteroduplexes. This conclusion is based on assays of the transforming activity of donor markers in lysates made after various times of incubation of recipient cells following exposure to DNA. The activity of a low efficiency marker rises substantially, indicating formation of native-like heteroduplex structures, and then falls. At 37° C the process is essentially completed 10 minutes after entry, and the apparent half life of a susceptible heteroduplex is 1.5 to 2 minutes. Data from these and other experiments imply that about as many of the surviving low efficiency markers have simply escaped attack as have been inserted into both strands by the excision-repair process suggested by Ephrussi-Taylor.     

Competence Mutant of Haemophilus influenzae with Abnormal Ratios of Marker Efficiencies in Transformation

In studies of competence-deficient mutants of Haemophilus influenzae which absorb deoxyribonucleic acid (DNA) but fail to produce transformants, it was observed that in some mutants the residual transforming activity for different markers varied widely, i.e., produced a ratio effect. One of these mutants, com⁻⁵⁶, was studied intensively to determine the cause of the residual efficiency of transformation and the reason for the ratio effect. The residual frequency of transformation was higher for markers considered single-site mutations (like naladixic acid resistance), whereas the least efficient markers tested were those conferring resistance to high levels of streptomycin or novobiocin which are more complex than single-site mutations. Measurement of frequencies of cotransformation indicated that overall genetic linkage was reduced. Transfection was fairly efficient with phage S2 DNA, but not prophage DNA. Donor marker activity could be detected in transformed cell lysates, but not linked to recipient markers in recombinant molecules. Sucrose gradient analysis of such lysates revealed that donor material was associated with recipient DNA in at least normal quantities, but lacked detectable genetic activity. Material from donor DNA labeled with heavy isotopes was incorporated into recipient chromosomal fragments having a density indistinguishable from normal density, unlike the hybrid density recombinant material found in normal cells. No excessive solubilization or nicking of unincorporated donor was detected. It is postulated that this strain contains a hyperactive nuclease, which reduces the effective size of the input DNA during the integration process.     

Fine-resolution mapping of spontaneous and double-strand break-induced gene conversion tracts in Saccharomyces cerevisiae reveals reversible mitotic conversion polarity.

Spontaneous and double-strand break (DSB)-induced gene conversion was examined in alleles of the Saccharomyces cerevisiae ura3 gene containing nine phenotypically silent markers and an HO nuclease recognition site. Conversions of these alleles, carried on ARS1/CEN4 plasmids, involved interactions with heteroalleles on chromosome V and were stimulated by DSBs created at HO sites. Crossovers that integrate plasmids into chromosomes were not detected since the resultant dicentric chromosomes would be lethal. Converted alleles in shuttle plasmids were easily transferred to Escherichia coli and analyzed for marker conversion, facilitating the characterization of more than 400 independent products from five crosses. This analysis revealed several new features of gene conversions. The average length of DSB-induced conversion tracts was 200 to 300 bp, although about 20% were very short (less than 53 bp). About 20% of spontaneous tracts also were also less than 53 bp, but spontaneous tracts were on average about 40% longer than DSB-induced tracts. Most tracts were continuous, but 3% had discontinuous conversion patterns, indicating that extensive heteroduplex DNA is formed during at least this fraction of events. Mismatches in heteroduplex DNA were repaired in both directions, and repair tracts as short as 44 bp were observed. Surprisingly, most DSB-induced gene conversion tracts were unidirectional and exhibited a reversible polarity that depended on the locations of DSBs and frameshift mutations in recipient and donor alleles.     

Unique cross-link and monoadduct repair characteristics of a xeroderma pigmentosum revertant cell line

Monoadducts and cross-links formed in DNA of human cells by a psoralen derivative, 4'-hydroxy-methyl-4,5',8-trimethylpsoralen (HMT), have been measured by a new, simple method, based on S1 nuclease digestion of 3H-labeled adducts in DNA, that provides rapid information on the repair of both classes of lesions. Normal human fibroblasts and cells from patients with dyskeratosis congenita and xeroderma pigmentosum (XP) group C were capable of removing both monoadducts and cross-links, whereas XP groups A and D failed to remove either. An XP revertant, isolated from a group A cell line on the basis of an acquired mutagen-induced resistance to ultraviolet light, has the unique property of being capable of removing cross-links but not monoadducts. Consistent with this property, the XP revertant was found to be resistant to cell killing by the cross-linking psoralen derivative, HMT, but as sensitive as its parental cell line to a monofunctional psoralen derivative, 5-methylisopsoralen.     

The effect of mismatch repair and heteroduplex formation on sexual isolation in Bacillus.

In Bacillus transformation, sexual isolation is known to be an exponential function of the sequence divergence between donor and recipient. Here, we have investigated the mechanism under which sequence divergence results in sexual isolation. We tested the effect of mismatch repair by comparing a wild-type strain and an isogenic mismatch-repair mutant for the relationship between sexual isolation and sequence divergence. Mismatch repair was shown to contribute to sexual isolation but was responsible for only a small fraction of the sexual isolation observed. Another possible mechanism of sexual isolation is that more divergent recipient and donor DNA strands have greater difficulty forming a heteroduplex because a region of perfect identity between donor and recipient is required for initiation of the heteroduplex. A mathematical model showed that this heteroduplex-resistance mechanism yields an exponential relationship between sexual isolation and sequence divergence. Moreover, this model yields an estimate of the size of the region of perfect identity that is comparable to independent estimates for Escherichia coli. For these reasons, and because all other mechanisms of sexual isolation may be ruled out, we conclude that resistance to heteroduplex formation is predominantly responsible for the exponential relationship between sexual isolation and sequence divergence in Bacillus transformation.     

Investigation of the tertiary folding of Escherichia coli ribosomal RNA by intra-RNA cross-linking in vivo

"In vivo" cross-links were introduced into ribosomal RNA by direct ultraviolet irradiation of intact Escherichia coli cells, during growth in a 32P-labelled medium. Ribosomes were isolated from the irradiated cultures, dissociated into subunits and subjected to partial digestion with cobra venom nuclease. The intra-RNA cross-linked fragments were separated by two-dimensional gel electrophoresis and the sites of cross-linking determined, using our published methodology. A comparison with the data previously obtained by this procedure, after irradiation of isolated 30 S and 50 S subunits, showed that in the case of the 50 S subunit nine out of the ten previous cross-links in the 23 S RNA could be identified in the "in vivo" experiments, and correspondingly in the 30 S subunit five out of the six previous cross-links in the 16 S RNA were identified. Some new cross-links were found, as well as two cross-links in the 16 S RNA, which had hitherto only been observed after partial digestion of irradiated 30 S subunits with ribonuclease T1. The relevance of these data to the tertiary folding of the rRNA in situ is discussed, with particular reference to the work of other authors, in which "naked" RNA was used as the substrate for cross-linking and model-building studies.     

Fate of transforming DNA following uptake by competent Bacillus subtilis. I. Formation and properties of the donor-recipient complex

Following uptake by competent Bacillus subtilis, transforming DNA is converted to two distinct slowly sedimenting molecular forms which possess little transforming activity (eclipse). A few minutes after uptake is initiated, a physical complex of donor and recipient DNA begins to form. The recovery of donor transforming activity following eclipse, and the appearance of recombinant activity, previously reported by Venema, Pritchard &; Venema-Schröder (1965), is shown to be due to changes occurring in the donor—recipient complex. This complex exists transiently in a form with low recombinant-type transforming activity. This transient form may be one in which the donor and recipient components are joined non-covalently. The donor-recipient complex is shown to be a heteroduplex structure in which the donor moiety has an approximate molecular weight of 750,000.     

Biochemical events essential to the recombination activity of Escherichia coli RecA protein. II. Co-dominant effects of RecA142 protein on wild-type RecA protein function

In the accompanying paper, RecA142 protein was found to be completely defective in DNA heteroduplex formation. Here, we show that RecA142 protein not only is defective in this activity but also is inhibitory for certain activities of wild-type RecA protein. Under appropriate conditions, RecA142 protein substantially inhibits the DNA strand exchange reaction catalyzed by wild-type RecA protein; at equimolar concentrations of each protein, formation of full-length gapped duplex DNA product molecules is less than 7% of the amount produced by wild-type protein alone. Inhibition by RecA142 protein is also evident in S1 nuclease assays of DNA heteroduplex formation, although the extent of inhibition is less than is observed for the complete DNA strand exchange process; at equimolar concentrations of wild-type and mutant proteins, the extent of DNA heteroduplex formation is 36% of the wild-type protein level. This difference implies that RecA142 protein prevents, at minimum, the branch migration normally observed during DNA strand exchange. RecA142 protein does not inhibit either the single-strand (ss) DNA-dependent ATPase activity or the coaggregation activities of wild-type RecA protein. This suggests that these reactions are not responsible for the inhibition of wild-type protein DNA strand exchange activity by RecA142 protein. However, under conditions where RecA142 protein inhibits DNA strand exchange activity, RecA142 protein renders the M13 ssDNA-dependent ATPase activity of wild-type protein sensitive to inhibition by single-strand DNA-binding protein, and it inhibits the double-strand DNA-dependent ATPase activity of wild-type RecA protein. These results imply that these two activities are important components of the overall DNA strand exchange process. These experiments also demonstrate the applicability of using defective mutant RecA proteins as specific codominant inhibitors of wild-type protein activities in vitro and should be of general utility for mechanistic analysis of RecA protein function both in vitro and in vivo.     

A vaccinia virus DNase preparation which cross-links superhelical DNA.

Multiple DNA-dependent enzyme activities have been detected in highly purified preparations of a single-strand-specific nuclease from vaccinia virus. These enzyme preparations were extensively purified and characterized by using superhelical DNAs as substrates. In particular, the nuclease activity was monitored by the extent of conversion of supercoiled closed duplex DNA (DNA I) to nicked circular DNA (DNA II), which could subsequently be converted to duplex linear DNA (DNA III) by prolonged incubation with the enzyme. DNA species which were not substrates for the enzyme included relaxed closed duplex DNA, DNA II which had been prepared by nuclease S1 treatment or by photochemical nicking of DNA I, and DNA III. With plasmid pSM1 DNA as substrate, the extent of cleavage of DNA I to DNA II was found to increase with superhelix density above a threshold value of about -0.06. The linear reaction products were examined by gel electrophoresis after restriction enzyme digestion of the DNAs from plasmids pSM1 and pBR322 and of the viral DNAs from bacteriophage phi X174 (replicative form) and simian virus 40, and the map coordinate locations of the scissions were determined. These products were further examined by electron microscopy and by gel electrophoresis under denaturing conditions. Electron micrographs taken under partially denaturing conditions revealed molecules with terminal loops or hairpins such as would result from the introduction of cross-links at the cutting sites. These species exhibited snapback renaturation. The denaturing gel electrophoresis experiments revealed the appearance of new bands at locations consistent with terminal cross-linking. With pSM1 and pBR322 DNAs, this band was shown to contain DNA that was approximately twice the length of a linear single strand. The terminal regions of the cross-linked linear duplex reaction products were sensitive to nuclease S1 but insensitive to proteinase K, suggesting that the structure is a hairpin loop not maintained by a protein linker. A similar structure is found in mature vaccinia virus DNA.     

Chromatographically Fractionated Complementary Strands of Bacillus subtilis Deoxyribonucleic Acid: Transformation of Hybrids

The annealing properties as measured by the restoration of transforming activity and hypochromicity of methylated albumin-kieselguhr (MAK)-fractionated complementary strands of Bacillus subtilis deoxyribonucleic acid (DNA) are presented. Temperature-absorbance measurements performed on annealed mixtures of various L and H strand fractions indicated the existence of a complementarity gradient between the two MAK peaks. The markers purA16, leu-8, metB(5), thr-5, and the linked marker hisB(2)-try-2 exhibited different bimodal distributions on MAK columns. The transforming efficiency of heteroduplex mixtures, prepared by cross-annealing resolved complementary strands of wild-type and recipient DNA, was compared. The transforming efficiency of the wild-type L and H strands was equal in one preparation and unequal in a second preparation. It was found that in the second strand preparation the heteroduplex DNA containing the H strand from wild type was more efficient for all of the markers tested. The variations in transforming efficiencies of the complementary strands in heteroduplex molecules reported here and by others are due in part to strands of unequal length and probably to the self-annealing property of the H strands. At present, no conclusion could be made regarding the existence of strand selection bias during integration of donor DNA in competent B. subtilis cells.     

Precursor-product relationship of larger to smaller molecular forms of the BAL 31 nuclease from Alteromonas espejiana: preferential removal of duplex exonuclease relative to endonuclease activity by proteolysis

Two molecularly and kinetically distinct major species of the extracellular nuclease BAL 31 from Alteromonas espejiana, previously characterized as the "fast" (F) and "slow" (S) BAL 31 nucleases, have been evidenced to derive from proteolysis starting from a still larger (approximately 120 kDa) precursor nuclease. The expected protease activity in the culture fluid has been confirmed and is strongly dependent on the cell growth phase. The disappearance of the largest nuclease species with the concomitant sequential appearance of first the F and then the S species has been demonstrated for nuclease obtained from culture supernatants as a function of cell growth phase. Nuclease from periplasmic extracts displayed very little of the F and S nucleases. Treatment of purified F nuclease with Pronase or subtilisin readily converted it to species with only a few percent of the native exonuclease activity against duplex DNA but retaining much of the initial activity against single-stranded DNA. Electrophoresis in nuclease-detecting gels demonstrated a parallel conversion of the larger species to one indistinguishable in molecular weight from the S species. The observed loss of exonuclease activity could correspond to the conversion of the F to the S nuclease. However, treatment of S nuclease with subtilisin resulted in a drastic reduction of exonuclease activity of this enzyme on duplex DNA with retention of most of the activity against single-stranded and nicked circular duplex DNA substrates. Evidence of internal proteolysis of the S nuclease could be seen after electrophoresis in denaturing gels but only after the denaturation buffer was adjusted to 6 M in urea. The preferential removal of the exonuclease activity may enhance the usefulness of the BAL 31 nuclease in such applications as heteroduplex mapping.     

DNA binding, coprotease, and strand exchange activities of mycobacterial RecA proteins: implications for functional diversity among RecA nucleoprotein filaments

One of the fundamental questions concerning homologous recombination is how RecA or its homologues recognize several DNA sequences with high affinity and catalyze all the diverse biological activities. In this study, we show that the extent of single-stranded DNA binding and strand exchange (SE) promoted by mycobacterial RecA proteins with DNA substrates having various degrees of GC content was comparable with that observed for Escherichia coli RecA. However, the rate and extent of SE promoted by these recombinases showed a strong negative correlation with increasing amounts of sequence divergence embedded at random across the length of the donor strand. Conversely, a positive correlation was seen between SE efficiency and the degree of sequence divergence in the recipient duplex DNA. The extent of heteroduplex formation was not significantly affected when both the pairing partners contained various degrees of sequence divergence, although there was a moderate decrease in the case of mycobacterial RecA proteins with substrates containing larger amounts of sequence divergence. Whereas a high GC content had no discernible effect on E. coli RecA coprotease activity, a negative correlation was apparent between mycobacterial RecA proteins and GC content. We further show clear differences in the extent of SE promoted by E. coli and mycobacterial RecA proteins in the presence of a wide range of ATP:ADP ratios. Taken together, our findings disclose the existence of functional diversity among E. coli and mycobacterial RecA nucleoprotein filaments, and the milieu of sequence divergence (i.e., in the donor or recipient) exerts differential effects on heteroduplex formation, which has implications for the emergence of new genetic variants.     

Genetic transformation of Streptococcus pneumoniae by heterologous plasmid deoxyribonucleic acid.

A number of heterologous plasmid deoxyribonucleic acids (DNAs) coding for erythromycin, tylosin, lincomycin, tetracycline, or chloramphenicol resistance have been introduced into Streptococcus pneumoniae via genetic transformation with frequencies that varied between 10(-5) to as high as 5 x 10(-1) per colony-forming unit. Transformation with plasmid DNA required pneumococcal competence, was competed by chromosomal DNA, and showed a saturation at about 0.5 micrograms/ml (with a recipient population of 3 x 10(7) colony-forming units of competent cells per ml). Plasmid transformation did not occur with a recipient strain, 410, defective in endonuclease I activity and in chromosomal genetic transformation. All erythromycin-resistant transformants examined contained covalently closed circular DNA with the same electrophoretic mobility on agarose gels as the donor DNAs, and when examined in detail the plasmid reisolated from the transformants had the same restriction patterns and the same specific transforming activity as the donor DNA. In the cases of two plasmids examined in detail--pAM77 and pSA5700 Lc9--most of the transforming activity was associated with DNA monomers; DNA multimers present in pSA5700 Lc9 also had biological activity. An unexpected finding was the demonstration of transformation (2 x 10(-5) per colony-forming unit) with plasmid DNAs linearized by treatment with S1 nuclease or with restriction endonucleases.     

Double-strand cleavage at a two-base deletion mismatch in a DNA heteroduplex by nuclease S1

A two-base deletion mismatch was generated in a DNA heteroduplex by hybridization of two linear plasmid DNA molecules differing only by the presence of a two-base deletion in one of them. The heteroduplex was shown to be sensitive to double-strand cleavage by nuclease S1, thus demonstrating the potential value of single-stranded probes for the detection of polymorphisms in genomic DNA due to very small deletions.