The mechanisms of action of E. coli endonuclease III and T4 UV endonuclease (endonuclease V) at AP sites

Treatment of DNA containing AP sites with either T4 UV endonuclease or with E. coli endonuclease III followed by a human class II AP endonuclease releases a putative beta-elimination product. This result suggests that both the T4 endonuclease and E. coli endonuclease III class I AP endonucleases catalyze phosphodiester bond cleavage via a lyase- rather than a hydrolase mechanism. Indeed, we have not detected a class I AP endonuclease which hydrolytically catalyzes phosphodiester bond cleavage. Whereas these enzymes use a lyase-like rather than a hydrolytic mechanism, they nonetheless catalyze phosphodiester bond cleavage. We suggest that the term endonuclease can be properly applied to them.     

Chlamydia pneumoniae AP endonuclease IV could cleave AP sites of double- and single-stranded DNA

Endonuclease IV gene, the only putative AP endonuclease of C. pneumoniae genome, was cloned into pET28a. Recombinant C. pneumoniae endonuclease I V (CpEndoIV) was expressed in E. coli and purified to homogeneity. CpEndoIV has endonuclease activity against apurinic/apyrimidinic sites (AP sites) of double-stranded (ds) oligonucleotides. AP endonuclease activity of CpEndoIV was promoted by divalent metal ions Mg2+ and Zn2+, and inhibited by EDTA. The natural (A, T, C and G) and modified (U, I and 8-oxo-G (GO)) bases opposite AP site had little effect on the cleavage efficiency of AP site of ds oligonucleotides by CpEndoIV. However, the CpEndoIV-dependent cleavage of AP site opposite modified base GO was strongly inhibited by Chlamydia DNA glycosylase MutY. Interestingly, the AP site in single-stranded (ss) oligonucleotides was also the effective substrate of CpEndoIV. Similar to E. coli endonuclease IV, AP endonuclease activity of CpEndoIV was also heat-stable to some extent, with a half time of 5 min at 60 degrees C.     

Bleomycin-treated DNA is specifically cleaved only by endonuclease IV in E. coli.

We have isolated an endonuclease from E. coli active on bleomycin-treated DNA. Purification on DEAE-cellulose separated this activity in strains lacking endonuclease I, endonuclease III or exonuclease III. After DEAE chromatography, the enzyme was active in the absence of divalent cations and was not inhibited by tRNA or harmane. In addition, this enzyme was stable at 45 degrees C for 20 min. These properties are consistent with this activity being endonuclease IV. This was supported by our finding no activity in a strain lacking endonuclease IV.     

Thymine glycol-DNA glycosylase/AP endonuclease of CEM-C1 lymphoblasts: a human analog of Escherichia coli endonuclease III

A thymine glycol-DNA glycosylase/AP endonuclease has been identified in human CEM-C1 lymphoblasts. The enzyme is active in the absence of divalent cations and has an apparent molecular size of approximately 60,000 daltons. The enzyme releases thymine glycol from osmium tetroxide-damaged DNA via an N-glycosylase activity and is associated with an endonuclease activity that mediates phosphodiester bond cleavage at sites of thymine glycol and apurinic sites. We propose that this enzyme, which we call redoxyendonuclease, is the human analog of a bacterial enzyme, E. coli endonuclease III, that recognizes oxidative DNA damage.     

Endonuclease IV enhances base excision repair of endonuclease III from Methanobacterium thermoautotrophicum

Damaged DNA strands are repaired by base excision (BER) in organisms, a process initiated by repair enzymes, which include DNA glycosylases and endonucleases. We expressed and characterized two putative endonuclease genes from Methanobacterium thermoautotrophicum, Mt0764 and Mt1010, encoding homologues of endonuclease III (endo III) and endonuclease IV (endo IV) of Escherichia coli. The Mt0764 and Mt1010 proteins showed endo III activity by removing thymine glycol from DNA strand and AP endonuclease activity, respectively. The Mt0764 protein not only cleaved the oligonucleotide duplex, containing a thymine glycol/adenine pair efficiently, but also showed activity on the 8-oxoguanine-containing oligonucleotide duplex. In this study, we report upon the stimulation of endo III activity by endo IV using two recombinant proteins (Mt1010 and Mt0764) from M. thermoautotrophicum. Mt1010 stimulated the DNA glycosylase activity of Mt0764 for DNA substrates containing 8-oxoguanine residues and increasing the formation of the Mt0764 protein-DNA complex. The interaction between Mt1010 and Mt0764 was observed by using an in vitro binding assay. These results suggest that association between endo III and endo IV may occur in vivo, and this contributes to efficient base excision repair for the oxidative damage of DNA.     

Involvement of two endonuclease III homologs in the base excision repair pathway for the processing of DNA alkylation damage in Saccharomyces cerevisiae

DNA base excision repair (BER) is initiated by DNA glycosylases that recognize and remove damaged bases. The phosphate backbone adjacent to the resulting apurinic/apyrimidinic (AP) site is then cleaved by an AP endonuclease or glycosylase-associated AP lyase to invoke subsequent BER steps. We have used a genetic approach in Saccharomyces cerevisiae to determine whether or not AP sites are blocks to DNA replication and the biological consequences if AP sites persist in the genome. We previously reported that yeast cells deficient in the two AP endonucleases (apn1 apn2 double mutant) are extremely sensitive to killing by a model DNA alkylating agent methyl methanesulfonate (MMS) and that this sensitivity can be reduced by deleting the MAG1 3-methyladenine DNA glycosylase gene. Here we report that in the absence of the AP endonucleases, deletion of two Escherichia coli endonuclease III homologs, NTG1 and NTG2, partially suppresses MMS-induced killing, which indicates that the AP lyase products are deleterious unless they are further processed by an AP endonuclease. The severe MMS sensitivity seen in AP endonuclease deficient strains can also be rescued by treatment of cells with the AP lyase inhibitor methoxyamine, which suggests that the product of AP lyase action on an AP site is indeed an extremely toxic lesion. In addition to the AP endonuclease interactions, deletion of NTG1 and NTG2 enhances the mag1 mutant sensitivity to MMS, whereas overexpression of MAG1 in either the ntg1 or ntg2 mutant severely affects cell growth. These results help to delineate alkylation base lesion flow within the BER pathway.     

Mechanism of action of a mammalian DNA repair endonuclease

The mechanism of action of a DNA repair endonuclease isolated from calf thymus was determined. The calf thymus endonuclease possesses a substrate specificity nearly identical with that of Escherichia coli endonuclease III following DNA damage by high doses of UV light, osmium tetroxide, and other oxidizing agents. The calf thymus enzyme incises damaged DNA at sites of pyrimidines. A cytosine photoproduct was found to be the primary monobasic UV adduct. The calf thymus endonuclease and E. coli endonuclease III were found to possess similar, but not identical, DNA incision mechanisms. The mechanism of action of the calf thymus endonuclease was deduced by analysis of the 3' and 5' termini of the enzyme-generated DNA scission products with DNA sequencing methodologies and HPLC analysis of the material released by the enzyme following DNA damage. The calf thymus endonuclease removes UV light and osmium tetroxide damaged bases via an N-glycosylase activity followed by a 3' apurinic/apyrimidinic (AP) endonuclease activity. The calf thymus endonuclease also possesses a novel 5' AP endonuclease activity not possessed by endonuclease III. The product of this three-step mechanism is a nucleoside-free site flanked by 3'-and 5'-terminal phosphate groups. These results indicate the conservation of both substrate specificity and mechanism of action in the enzymatic removal of oxidative base damage between prokaryotes and eukaryotes. We propose the name redoxy endonucleases for this group of enzymes.     

Escherichia coli endonuclease III is not an endonuclease but a beta-elimination catalyst.

The oligonucleotide [5'-32P]pdT8d(-)dTn, containing an apurinic/apyrimidinic (AP) site [d(-)], yields three radioactive products when incubated at alkaline pH: two of them, forming a doublet approximately at the level of pdT8dA when analysed by polyacrylamide-gel electrophoresis, are the result of the beta-elimination reaction, whereas the third is pdT8p resulting from beta delta-elimination. The incubation of [5'-32P]pdT8d(-)dTn, hybridized with poly(dA), with E. coli endonuclease III yields two radioactive products which have the same electrophoretic behaviour as the doublet obtained by alkaline beta-elimination. The oligonucleotide pdT8d(-) is degraded by the 3'-5' exonuclease activity of T4 DNA polymerase as well as pdT8dA, showing that a base-free deoxyribose at the 3' end is not an obstacle for this activity. The radioactive products from [5'-32P]pdT8d(-)dTn cleaved by alkaline beta-elimination or by E. coli endonuclease III are not degraded by the 3'-5' exonuclease activity of T4 DNA polymerase. When DNA containing AP sites labelled with 32P 5' to the base-free deoxyribose labelled with 3H in the 1' and 2' positions is degraded by E. coli endonuclease VI (exonuclease III) and snake venom phosphodiesterase, the two radionuclides are found exclusively in deoxyribose 5-phosphate and the 3H/32P ratio in this sugar phosphate is the same as in the substrate DNA. When DNA containing these doubly-labelled AP sites is degraded by alkaline treatment or with Lys-Trp-Lys, followed by E. coli endonuclease VI (exonuclease III), some 3H is found in a volatile compound (probably 3H2O) whereas the 3H/32P ratio is decreased in the resulting sugar phosphate which has a chromatographic behaviour different from that of deoxyribose 5-phosphate. Treatment of the DNA containing doubly-labelled AP sites with E. coli endonuclease III, then with E. coli endonuclease VI (exonuclease III), also results in the loss of 3H and the formation of a sugar phosphate with a lower 3H/32P ratio that behaves chromatographically as the beta-elimination product digested with E. coli endonuclease VI (exonuclease III). From these data, we conclude that E. coli endonuclease III cleaves the phosphodiester bond 3' to the AP site, but that the cleavage is not a hydrolysis leaving a base-free deoxyribose at the 3' end as it has been so far assumed. The cleavage might be the result of a beta-elimination analogous to the one produced by an alkaline pH or Lys-Trp-Lys. Thus it would seem that E. coli 'endonuclease III' is, after all, not an endonuclease.     

Determination of active site residues in Escherichia coli endonuclease VIII.

Endonuclease VIII from Escherichia coli is a DNA glycosylase/lyase that removes oxidatively damaged bases. EndoVIII is a functional homologue of endonuclease III, but a sequence homologue of formamidopyrimidine-DNA glycosylase (Fpg). Using multiple sequence alignments, we have identified six target residues in endoVIII that may be involved in the enzyme's glycosylase and/or lyase functions: the N-terminal proline, and five acidic residues that are completely conserved in the endoVIII-Fpg proteins. To investigate the contribution of these residues, site-directed mutagenesis was used to create seven mutants: P2T, E3D, E3Q, E6Q, D129N, D160N, and E174Q. Each mutant was assayed both for lyase activity on abasic (AP) sites and for glycosylase/lyase activity on 5-hydroxyuracil, thymine glycol, and gamma-irradiated DNA with multiple lesions. The P2T mutant did not have lyase or glycosylase/lyase activity but could efficiently form Schiff base intermediates on AP sites. E6Q, D129N, and D160N behaved essentially as endoVIII in all assays. E3D, E3Q, and E174Q retained significant AP lyase activity but had severely diminished or abolished glycosylase/lyase activities on the DNA lesions tested. These studies provide detailed predictions concerning the active site of endoVIII.     

Characterization of the Escherichia coli X-ray endonuclease, endonuclease III

The X-ray endonuclease endonuclease III of Escherichia coli has been purified to apparent homogeneity by using the criterion of sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The most purified fraction shows endonucleolytic activity against apurinic and apyrimidinic (AP) sites and a dose-dependent response to DNA that has been X irradiated, UV irradiated, or treated with OsO4. The endonuclease also nicks OsO4-treated DNA that has been subsequently treated with alkali to produce fragmented thymine residues and DNA treated with potassium permanganate. The enzyme does not incise the alkali-labile sites present in DNA X irradiated in vitro in the presence of hydroxyl radical scavengers. The most purified fractions exhibit two distinct activities, an AP endonuclease that cleaves on the 3' side of the damage leaving a 3'-OH and a 5'-PO4 and a DNA N-glycosylase that recognizes at least two substrates, thymine glycol residues and urea residues. The glycosylase activity is sensitive to N-ethylmaleimide while the AP endonuclease is not.     

Selective inhibition by methoxyamine of the apurinic/apyrimidinic endonuclease activity associated with pyrimidine dimer-DNA glycosylases from Micrococcus luteus and bacteriophage T4

The UV endonucleases [endodeoxyribonuclease (pyrimidine dimer), EC 3.1.25.1] from Micrococcus luteus and bacteriophage T4 possess two catalytic activities specific for the site of cyclobutane pyrimidine dimers in UV-irradiated DNA: a DNA glycosylase that cleaves the 5'-glycosyl bond of the dimerized pyrimidines and an apurinic/apyrimidinic (AP) endonuclease that thereupon incises the phosphodiester bond 3' to the resulting apyrimidinic site. We have explored the potential use of methoxyamine, a chemical that reacts at neutral pH with AP sites in DNA, as a selective inhibitor of the AP endonuclease activities residing in the M. luteus and T4 enzymes. The presence of 50 mM methoxyamine during incubation of UV- (4 kJ/m2, 254 nm) treated, [3H]thymine-labeled poly(dA).poly(dT) with either enzyme preparation was found to protect completely the irradiated copolymer from endonucleolytic attack at dimer sites, as assayed by yield of acid-soluble radioactivity. In contrast, the dimer-DNA glycosylase activity of each enzyme remained fully functional, as monitored retrospectively by release of free thymine after either photochemical- (5 kJ/m2, 254 nm) or photoenzymic- (Escherichia coli photolyase plus visible light) induced reversal of pyrimidine dimers in the UV-damaged substrate. Our data demonstrate that the inhibition of the strand-incision reaction arises because of chemical modification of the AP sites and is not due to inactivation of the enzyme by methoxyamine. Our results, combined with earlier findings for 5'-acting AP endonucleases, strongly suggest that methoxyamine is a highly specific inhibitor of virtually all AP endonucleases, irrespective of their modes of action, and may therefore prove useful in a wide variety of DNA repair studies.     

Isolation of cDNA clones encoding a human apurinic/apyrimidinic endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III) mutants.

Apurinic/apyrimidinic (AP) sites in cellular DNA are considered to be both cytotoxic and mutagenic, and can arise spontaneously or following exposure to DNA damaging agents. We have isolated cDNA clones which encode an endonuclease, designated HAP1 (human AP endonuclease 1), that catalyses the initial step in AP site repair in human cells. The predicted HAP1 protein has an Mr of 35,500 and shows striking sequence similarity (93% identity) to BAP 1, a bovine AP endonuclease enzyme. Significant sequence homology to two bacterial DNA repair enzymes, E. coli exonuclease III and S. pneumoniae ExoA proteins, and to Drosophila Rrp1 protein is also apparent. We have expressed the HAP1 cDNA in E. coli mutants lacking exonuclease III (xth), endonuclease IV (nfo), or both AP endonucleases. The HAP1 protein can substitute for exonuclease III, but not for endonuclease IV, in respect of some, but not all, DNA repair and mutagenesis functions. Moreover, a dut xth (ts) double mutant, which is nonviable at 42 degrees C due to an accumulation of unrepaired AP sites following excision of uracil from DNA, was rescued by expression of the HAP1 cDNA. These results indicate that AP endonucleases show remarkable conservation of both primary sequence and function. We would predict that the HAP1 protein is important in human cells for protection against the toxic and mutagenic effects of DNA damaging agents.     

Nucleotide excision repair 3' endonuclease XPG stimulates the activity of base excision repairenzyme thymine glycol DNA glycosylase.

An ionizing radiation-induced DNA lesion, thymine glycol, is removed from DNA by a thymine glycol DNA glycosylase with an apurinic/apyrimidinic (AP) lyase activity encoded by the Escherichia coli endonuclease III ( nth ) gene and its homolog in humans. Cells from Cockayne syndrome patients with mutations in the XPG gene show approximately 2-fold reduced global repair of thymine glycol. Hence, I decided to investigate the molecular mechanism of the effect of XPG protein observed in vivo on thymine glycol removal by studying the interactions of XPG protein and human endonuclease III (HsNTH) protein in vitro and the effect of XPG protein on the activity of HsNTH protein on a substrate containing thymine glycol. The XPG protein stimulates the binding of HsNTH protein to its substrate and increases its glycosylase/AP lyase activity by a factor of approximately 2 through direct interaction between the two proteins. These results provide in vitro evidence for a second function of XPG protein in DNA repair and a mechanistic basis for its stimulatory activity on HsNTH protein.     

Mechanism of action of Escherichia coli endonuclease III

Endonuclease III isolated from Escherichia coli has been shown to have both N-glycosylase and apurinic/apyrimidinic (AP) endonuclease activities. A nicking assay was used to show that the enzyme exhibited a preference for form I DNA when DNA containing thymine glycol was used as a substrate. This preference was reduced or eliminated either when the DNA was relaxed or when the type of damage was altered to urea residues or AP sites. The combined N-glycosylase/AP endonuclease activity was at least 10-fold higher than the AP endonuclease activity alone when urea-containing DNA was used as a substrate as compared to AP DNA. When DNA containing thymine glycol was used as a substrate, the combined N-glycosylase/AP endonuclease activity was about 2-fold higher than the AP endonuclease activity. Yet, when DNA containing thymine glycol or urea was used as substrate, no apurinic sites remained. Furthermore, magnesium selectively inhibited endonuclease III activity when AP DNA was used as a substrate but had no effect when DNA containing either urea or thymine glycol was used as substrate. These data suggest that both the N-glycosylase and AP endonuclease activities of endonuclease III reside on the same molecule or are in very tight association and that these activities act in concert, with the N-glycosylase reaction preceding the AP endonuclease reaction.     

Excision of 5'-terminal deoxyribose phosphate from damaged DNA is catalyzed by the Fpg protein of Escherichia coli.

Homogeneous Fpg protein of Escherichia coli has DNA glycosylase activity which excises some purine bases with damaged imidazole rings, and an activity excising deoxyribose (dR) from DNA at abasic (AP) sites leaving a gap bordered by 5'- and 3'-phosphoryl groups. In addition to these two reported activities, we show that the Fpg protein also catalyzes the excision of 5'-terminal deoxyribose phosphate (dRp) from DNA, which is the principal product formed by the incision of AP endonucleases at abasic sites. Moreover, the rate of the Fpg protein catalysis for the 2,6-diamino-4-hydroxy-5-formamidopyrimidine-DNA glycosylase activity is slower than the activities excising dR from abasic sites and dRp from abasic sites preincised by endonucleases. The product released by the Fpg protein in the excision of 5'-terminal dRp from an abasic site preincised by an AP endonuclease is a single base-free unsaturated dRp, suggesting that the excision results from beta-elimination. The release of 5'-terminal dRp by crude extracts of E. coli from wild type and fpg-mutant strains shows that the Fpg protein is one of the major EDTA-resistant activities catalyzing this reaction.     

Drosophila ribosomal protein PO contains apurinic/apyrimidinic endonuclease activity.

Drosophila ribosomal protein PO was overexpressed in Escherichia coli to allow for its purification, biochemical characterization and to generate polyclonal antibodies for Western analysis. Biochemical tests were originally performed to see if overexpressed PO contained DNase activity similar to that recently reported for the apurinic/apyrimidinic (AP) lyase activity associated with Drosophila ribosomal protein S3. The overexpressed ribosomal protein was subsequently found to act on AP DNA, producing scissions that were in this case 5' of a baseless site instead of 3', as has been observed for S3. As a means of confirming that the source of AP endonuclease activity was in fact due to PO, glutathione S-transferase (GST) fusions containing a Factor Xa cleavage site between GST and PO were constructed, overexpressed in an E.coli strain defective for the major 5'-acting AP endonucleases and the fusions purified using glutathione-agarose affinity column chromatography. Isolated fractions containing purified GST-PO fusion proteins were subsequently found to have authentic AP endonuclease activity. Moreover, glutathione-agarose was able to deplete AP endonuclease activity from GST-PO fusion protein preparations, whereas the resin was ineffective in lowering DNA repair activity for PO that had been liberated from the fusion construct by Factor Xa cleavage. These results suggested that PO was a multifunctional protein with possible roles in DNA repair beyond its known participation in protein translation. In support of this notion, tests were performed that show that GST-PO, but not GST, was able to rescue an E.coli mutant lacking the major 5'-acting AP endonucleases from sensitivity to an alkylating agent. We furthermore show that GST-PO can be located in both the nucleus and ribosomes. Its nuclear location can be further traced to the nuclear matrix, thus placing PO in a subcellular location where it could act as a DNA repair protein. Other roles beyond DNA repair seem possible, however, since GST-PO also exhibited significant nuclease activity for both single- and double-stranded DNA.     

Repair of oxidized abasic sites by exonuclease III, endonuclease IV, and endonuclease III

2-Deoxyribonolactone (L) and the C4'-oxidized abasic site (C4-AP) are produced by a variety of DNA-damaging agents. If not repaired, these lesions can be mutagenic. Exonuclease III and endonuclease IV are the major enzymes in E. coli responsible for 5'-incision of abasic sites (APs), the first steps in AP repair. Endonuclease III efficiently excises AP lesions via intermediate Schiff-base formation. Incision of L and C4-AP lesions by exonuclease III and endonuclease IV was determined under steady-state conditions using oligonucleotide duplexes containing the lesions at defined sites. An abasic lesion (AP) in an otherwise identical DNA sequence was incised by exonuclease III or endonuclease IV approximately 6-fold more efficiently than either of the oxidized abasic sites (L, C4-AP). Endonuclease IV incision efficiency of 2-deoxyribonolactone or C4-AP was independent of whether the lesion was opposite dA or dG. 2-Deoxyribonolactone is known to cross-link to endonuclease III (Hashimoto, M. (2001) J. Am. Chem. Soc. 123, 3161.). However, the C4-AP lesion is efficiently excised by endonuclease III. Oxidized abasic site repair by endonuclease IV and endonuclease III (C4-AP only) is approximately 100-fold less efficient than repair by exonuclease III. These results suggest that the first step of C4-AP and L oxidized abasic site repair will be the same as that of regular AP lesions in E. coli.     

Recognition of oxidized abasic sites by repair endonucleases.

The recognition of 'regular' and 'oxidized' sites of base loss (AP sites) in DNA by various AP endonucleases was compared. Model substrates with regular AP sites (resulting from mere hydrolysis of the glycosylic bond) were produced by damaging bacteriophage PM2 DNA by exposure to low pH; those with AP sites oxidized at the C-4'- and C-1'-position of the sugar moiety by exposure to Fe(III)-bleomycin in the presence of H2O2 and to Cu(II)-phenanthroline in the presence of H2O2 and ethanol, respectively. The results confirmed that AP sites-together with single-strand breaks-are indeed the predominant type of DNA modification in all three cases. For the recognition of 4'-oxidized AP sites, a 400-fold higher concentration of Escherichia coli exonuclease III and between 5-fold and 50-fold higher concentrations of bacteriophage T4 endonuclease V, E. coli endonuclease III and E. coli FPG protein were required than for the recognition of regular AP sites. In contrast, the recognition of 4'-oxidized AP sites by E. coli endonuclease IV was effected by 4-fold lower concentrations than needed for regular AP sites. 1'-oxidized AP sites (generated by activated Cu(II)-phenanthroline) were recognized by endonuclease IV and exonuclease III only slightly (3-fold and 13-fold, respectively) less efficiently than regular AP sites. In contrast, there was virtually no recognition of 1'-oxidized AP sites by the enzymes which cleave at the 3' side of AP sites (T4 endonuclease V, endonuclease III and FPG protein). The described differences were exploited for the analysis of the DNA damage induced by hydroxyl radicals, generated by ionizing radiation or Fe(III)-nitrilotriacetate in the presence of H2O2. The results indicate that both regular and 1'-oxidized AP sites represent only minor fractions of the AP sites induced by hydroxyl radicals.     

Processing in vitro of an abasic site reacted with methoxyamine: a new assay for the detection of abasic sites formed in vivo.

In this study we demonstrate that the different substrate recognition properties of bacterial and human AP endonucleases might be used to quantify and localize apurinic (AP) sites formed in DNA in vivo. By using a model oligonucleotide containing a single AP site modified with methoxyamine (MX), we show that endonuclease III and IV of E. coli are able to cleave the alkoxyamine-adducted site whereas a partially purified HeLa AP endonuclease and crude cell-free extracts from HeLa cells are inhibited by this modification. In addition MX-modified AP sites in a DNA template retain their ability to block DNA synthesis in vitro. Since MX can efficiently react with AP sites formed in mammalian cells in vivo we propose that the MX modified abasic sites thus formed can be quantitated and localized at the level of the individual gene by subsequent site specific cleavage by either E. coli endonuclease III or IV in vitro.