High-resolution NMR study of a GdAGA tetranucleotide loop that is an improved substrate for ricin, a cytotoxic plant protein.

Ricin is a cytotoxic plant protein that inactivates ribosomes by hydrolyzing the N-glycosidic bond at position A4324 in eukaryotic 28S rRNA. Recent studies showed that a four-nucleotide loop, GAGA, can function as a minimum substrate for ricin (the first adenosine corresponds to the site of depurination). We previously clarified the solution structure of this loop by NMR spectroscopy [Orita et al. (1993) Nucleic Acids Res. 21, 5670-5678]. To elucidate further details of the structural basis for recognition of its substrate by ricin, we studied the properties of a synthetic dodecanucleotide, r1C2U3C4A5G6dA7G8A9U10G11A12G (6dA12mer), which forms an RNA hairpin structure with a GdAGA loop and in which the site of depurination is changed from adenosine to 2'-deoxyadenosine. The N-glycosidase activity against the GdAGA loop of the A-chain of ricin was 26 times higher than that against the GAGA loop. NMR studies indicated that the overall structure of the GdAGA loop was similar to that of the GAGA loop with the exception of the sugar puckers of 6dA and 7G. Therefore, it appears that the 2'-hydroxyl group of adenosine at the depurination site (6A) does not participate in the recognition by ricin of the substrate. Since the 2'-hydroxyl group can potentially destabilize the developing positive charge of the putative transition state intermediate, an oxycarbonium ion, the electronic effect may explain, at least in part, the faster rate of depurination of the GdAGA loop compared to that of GAGA loop. We also show that the amino group of 7G is essential for substrate recognition the ricin A-chain.     

Ribosomal RNA identity elements for ricin A-chain recognition and catalysis. Analysis with tetraloop mutants.

Ricin is a cytotoxic protein that inactivates ribosomes by hydrolyzing the N-glycosidic bond between the base and the ribose of the adenosine at position 4324 in eukaryotic 28 S rRNA. Ricin A-chain will also catalyze depurination in naked prokaryotic 16 S rRNA; the adenosine is at position 1014 in a GAGA tetraloop. The rRNA identity elements for recognition by ricin A-chain and for the catalysis of cleavage were examined using synthetic GAGA tetraloop oligoribonucleotides. The RNA designated wild-type, an oligoribonucleotide (19-mer) that approximates the structure of the ricin-sensitive site in 16 S rRNA, and a number of mutants were transcribed in vitro from synthetic DNA templates with phage T7 RNA polymerase. With the wild-type tetraloop oligoribonucleotide the ricin A-chain-catalyzed reaction has a Km of 5.7 microM and a Kcat of 0.01 min-1. The toxin alpha-sarcin, which cleaves the phosphodiester bond on the 3' side of G4325 in 28 S rRNA, does not recognize the tetraloop RNA, although alpha-sarcin does affect a larger synthetic oligoribonucleotide that has a 17-nucleotide loop with a GAGA sequence; thus, there is a clear divergence in the identity elements for the two toxins. Mutants were constructed with all of the possible transitions and transversions of each nucleotide in the GAGA tetraloop; none was recognized by ricin A-chain. Thus, there is an absolute requirement for the integrity of the GAGA sequence in the tetraloop. The helical stem of the tetraloop oligoribonucleotide can be reduced to three base-pairs, indeed, to two base-pairs if the temperature is decreased, without affecting recognition; the nature of these base-pairs does not influence recognition or catalysis by ricin A-chain. If the tetraloop is opened so as to form a GAGA-containing hexaloop, recognition by ricin A-chain is lost. This suggests that during the elongation cycle, a GAGA tetraloop either exists or is formed in the putative 17-member single-stranded region of the ricin domain in 28 S rRNA and this bears on the mechanism of protein synthesis.     

Non-specific deadenylation and deguanylation of naked RNA catalyzed by ricin under acidic condition.

Ricin A-chain catalyzes the hydrolysis of the N-glycosidic bond of a conserved adenosine residue at position 4324 in the sarcin/ricin domain of 28S RNA of rat ribosome. The GAGA tetraloop closed by C-G pairs is required for recognition of the cleavage site on 28S ribosomal RNA by ricin A-chain. In this study, ricin A-chain (reduced ricin) exhibits specific depurination on a synthetic oligoribonucleotide (named SRD RNA) mimic of the sarcin/ricin domain of rat 28S ribosomal RNA under neutral and weak acidic conditions. Furthermore, the activity of intact ricin is also similar to that of ricin A-chain. However, under more acidic conditions, both enzymes lose their site specificity. The alteration in specificity of depurination is not dependent on the GAGA tetraloop of SRD RNA. A higher concentration of KCl inhibits the non-specific N-glycosidase activity much more than the specific activity of ricin A-chain. In addition, characterization of depurination sites by RNA sequencing reveals that under acidic conditions ricin A-chain can release not only adenines, but also guanines from SRD RNA or 5S ribosomal RNA. This is the first report of the non-specific deadenylation and deguanylation activity of ricin A-chain to the naked RNA under acidic conditions.     

Ricin A-chain substrate specificity in RNA, DNA, and hybrid stem-loop structures

Ricin toxin A-chain (RTA) is the catalytic subunit of ricin, a heterodimeric toxin from castor beans. Its ribosomal inactivating activity arises from depurination of a single adenine from position A(4324) in a GAGA tetraloop from 28S ribosomal RNA. Minimal substrate requirements are the GAGA tetraloop and stem of two or more base pairs. Depurination activity also occurs on stem-loop DNA with the same sequence, but with the k(cat) reduced 200-fold. Systematic variation of RNA 5'-G(1)C(2)G(3)C(4)[G(5)A(6)G(7)A(8)]G(9)C(10)G(11)C(12)-3' 12mers via replacement of each nucleotide in the tetraloop with a deoxynucleotide showed a 16-fold increase in k(cat) for A(6) --> dA(6) but reduced k(cat) up to 300-fold for the other sites. Methylation of individual 2'-hydroxyls in a similar experiment reduced k(cat) by as much as 3 x 10(-3)-fold. In stem-loop DNA, replacement of d[G(5)A(6)G(7)A(8)] with individual ribonucleotides resulted in small kinetic changes, except for the dA(6) --> A(6) replacement for which k(cat) decreased 6-fold. Insertion of d[G(5)A(6)G(7)A(8)] into an RNA stem-loop or G(5)A(6)G(7)A(8) into a DNA stem-loop reduced k(cat) by 30- and 5-fold, respectively. Multiple substitutions of deoxyribonucleotides into RNA stem-loops in one case (dG(5),dG(7)) decreased k(cat)/K(m) by 10(5)-fold, while a second change (dG(5),dA(8)) decreased k(cat) by 100-fold. Mapping these interactions on the structure of GAGA stem-loop RNA suggests that all the loop 2'-hydroxyl groups play a significant role in the action of ricin A-chain. Improved binding of RNA-DNA stem-loop hybrids provides a scaffold for inhibitor design. Replacing the adenosine of the RTA depurination site with deoxyadenosine in a small RNA stem-loop increased k(cat) 20-fold to 1660 min(-1), a value similar to RTA's k(cat) on intact ribosomes.     

Ribosomal RNA identity elements for ricin A-chain recognition and catalysis

Ricin is a cytotoxic protein that inactivates ribosomes by hydrolyzing the N-glycosidic bond between the base and the ribose at position A4324 in eukaryotic 28 S rRNA. The requirements for the recognition by ricin A-chain of this nucleotide and for the catalysis of cleavage were examined using a synthetic oligoribonucleotide that reproduces the sequence and the secondary structure of the RNA domain (a helical stem, a bulged nucleotide, and a 17-member single-stranded loop). The wild-type RNA (35mer) and a number of mutants were transcribed in vitro from synthetic DNA templates with phage T7 RNA polymerase. With the wild-type oligoribonucleotide the ricin A-chain catalyzed reaction has a Km of 13.55 microM and a Kcat of 0.023 min-1. Recognition and catalysis by ricin A-chain has an absolute requirement for A at the position that corresponds to 4324. The helical stem is also essential; however, the number of base-pairs can be reduced from the seven found in 28 S rRNA to three without loss of identity. The nature of these base-pairs can affect catalysis. A change of the second set from one canonical (G.C) to another (U.A) reduces sensitivity to ricin A-chain; whereas, a change of the third pair (U.A----G.C) produces supersensitivity. The bulged nucleotide does not contribute to identification. Hydrolysis is affected by altering the nucleotides in the universal sequence surrounding A4324 or by changing the position in the loop of the tetranucleotide GA(ricin)GA: all of these mutants have a null phenotype. If ribosomes are treated first with alpha-sarcin to cleave the phosphodiester bond at G4325 ricin can still catalyze depurination at A4324. This implies that cleavage by alpha-sarcin at the center of what has been presumed to be a 17 nucleotide single-stranded loop in 28 S rRNA produces ends that are constrained in some way. On the other hand, hydrolysis by alpha-sarcin of the corresponding position in the synthetic oligoribonucleotide prevents recognition by ricin A-chain. The results suggest that the loop has a complex structure, affected by ribosomal proteins, and this bears on the function in protein synthesis of the alpha-sarcin/ricin rRNA domain.     

The ribosomal RNA identity elements for ricin and for alpha-sarcin: mutations in the putative CG pair that closes a GAGA tetraloop.

alpha-Sarcin is a ribonuclease that cleaves the phosphodiester bond on the 3' side of G4325 in 28S rRNA; ricin A-chain is a RNA N-glycosidase that depurinates the 5' adjacent A4324. These single covalent modifications inactivate the ribosome. An oligoribonucleotide that reproduces the structure of the sarcin/ricin domain in 28S rRNA was synthesized and mutations were constructed in the 5' C and the 3' G that surround a GAGA tetrad that has the sites of toxin action. Covalent modification of the RNA by ricin, but not by alpha-sarcin, requires a Watson-Crick pair to shut off a putative GAGA tetraloop. Either the recognition elements for the two toxins are different despite their catalyzing covalent modification of adjacent nucleotides in 28S rRNA or there are transitions in the conformation of the alpha-sarcin/ricin domain in 28S rRNA and one conformer is recognized by alpha-sarcin and the other by ricin A-chain.     

The RNA N-glycosidase activity of ricin A-chain

The RNA N-glycosidase activity of ricin A-chain has been characterized. When rat liver ribosomes were used as substrates, the A-chain cleaved the N-glycosidic bond at A-4324 in 28S rRNA. An apparent Michaelis constant (Km) for the reaction was determined to be 2.6 microM and the turnover number (Kcat) was 1777 min-1. When naked rRNA was the substrate, the A-chain cleaved the same bond in 28S rRNA but at a greatly reduced rate. The Km value was 5.8 microM. The results suggest that the A-chain has a similar affinity for 28S rRNA in both ribosomes and the naked states. When the deproteinized Escherichia coli rRNA was the substrates, ricin A-chain cleaved a N-glycosidic bond at A-2600 in 23S rRNA which corresponds to the ricin-site in 28S rRNA of rat liver ribosomes, while the A-chain has little activity on 23S rRNA in the ribosomes. The results suggest that ricin A-chain acts directly on RNA by recognizing a certain structure in the molecules. Using the secondary structure models for each species of rRNA, we have deduced a loop and stem structure having GAGA in the loop to be a minimum requirement for the substrate of ricin A-chain.     

Ricin A-chain activity on stem-loop and unstructured DNA substrates

Ricin toxin A-chain (RTA) depurinates a single adenylate on a GAGA stem-loop region of eukaryotic 28S RNA, making it a potent toxin. Steady state rate analysis is used to establish the kinetic parameters for depurination of short RNA, DNA, and RNA-DNA hybrids of GAGA linear segments and stem-loop regions as substrates for RTA. Both stem and tetraloop structures are essential for action on RNA. For DNA stem-loop substrates, stem stability plays a small role in enhancing catalytic turnover but can enhance binding by up to 3 orders of magnitude. DNA sequences of d[GAGA] without stem-loop structures are found to be slow substrates for RTA. In contrast, equivalent RNA sequences exhibit no activity with RTA. Introduction of a deoxyadenosine at the depurination site of short RNA oligonucleotides restores catalytic function. NMR analysis indicates that the short, nonsubstrate GAGA is converted to substrate in GdAGA by the presence of a more flexible ribosyl group at the deoxyadenosine site. Conversion between C2'-endo and C2'-exo conformations at the deoxyadenosine site moves the 3'- and 5'-phosphorus atoms by 1.1 A, and the former is proposed to place them in a catalytically favorable configuration. The ability to use short RNA-DNA hybrids as substrates for RTA permits exploration of related structures to function as substrates and inhibitors.     

The consensus sequence for self-catalyzed site-specific G residue depurination in DNA

The sequence variation tolerated within the stem-loop-forming genomic consensus sequence for self-catalyzed site-specific depurination of G residues is explored. The variation in self-depurination kinetics with sequence changes in the loop residues and stem base pairs, as well as with pH, provides insights into the self-catalytic mechanism. The observations suggest that self-catalyzed depurination of the 5' G residue of the loop consensus sequence 5'-G(T/A)GG-3' probably involves formation of some intraloop hydrogen-bonded base pair with the 3'-terminal G residue; although the electronic structure of both these G residues is retained, their 2-amino substituents are not critical for that interaction. The strong dependence of the self-depurination kinetics on stem stability suggests that the lifetime of some strained form of the loop is controlled by the integrity of the stem. In addition to the effects of length and base pair sequence on stem stability, there is a base pair requirement at the base of the loop: self-depurination is suppressed by 5'-C·G-3', 5'-A·T-3', or a mismatch but is most favored by 5'T·A3' and less so by 5'-G·C-3'. The occurrence in T and G of a similarly located carbonyl capable of hydrogen-bonding to the water molecule required for glycosyl bond hydrolysis may explain this sequence requirement. In toto, the more complete definition of the consensus sequence provided by this investigation enables a more accurate estimation of their number in the human genome and their distribution among different genes.     

Vinyldeoxyadenosine in a sarcin-ricin RNA loop and its binding to ricin toxin a-chain

8-Vinyl-2'-deoxyadenosine (8vdA) is a fluorophore with a quantum yield comparable to that of 2-aminopurine nucleoside. 8vdA was incorporated into a 10-mer stem-tetraloop RNA (8vdA-10) structure for characterization of the properties of the base, 8-vinyladenine (8-vA), with respect to adenine as a substrate or inhibitor for ribosome-inactivating proteins. Ricin toxin A-chain (RTA) and pokeweed antiviral protein (PAP) catalyze the release of adenine from a specific adenosine on a stem-tetraloop (GAGA) sequence at the elongation factor (eEF2) binding site of the 28S subunit of eukaryotic ribosomes, thereby arresting translation. RTA does not catalyze the release of 8-vinyladenine from 8vdA-10. Molecular dynamics simulations implicate a role for Arg180 in oxacarbenium ion destabilization and the lack of catalysis. However, 8vdA-10 is an active site analogue and inhibits RTA with a Ki value of 2.4 microM. Adenine is also released from the second adenosine in the modified tetraloop, demonstrating an alternative mode for the binding of this motif in the RTA active site. The 8vdA analogue defines the specificities of RTA for the two adenylate depurination sites in a RNA substrate with a GAGA tetraloop. The rate of nonenzymatic acid-catalyzed solvolysis of 8-vinyladenine from the stem-loop RNA is described. Unlike RTA, PAP catalyzes the slow release of 8-vinyladenine from 8vdA-10. The isolation of 8-vA and its physicochemical characterization is described.     

The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins

Ricin is a potent cytotoxic protein derived from the higher plant Ricinus communis that inactivates eukaryotic ribosomes. In this paper we have studied the mechanism of action of ricin A-chain on rat liver ribosomes in vitro. Our findings indicate that the toxin inactivates the ribosomes by modifying both or either of two nucleoside residues, G4323 and A4324, in 28 S rRNA. These nucleotides are located close to the alpha-sarcin cleavage site and become resistant to all ribonucleases tested. The examination of the lability of phosphodiester bonds of these nucleotides to both mild alkaline digestion and aniline treatment at acidic pH suggests that the base of A4324 is removed by the toxin. This unique activity of ricin A-chain was also observed when naked 28 S rRNA is used as a substrate, indicating that the toxin directly acts on the RNA. Similar activity on 28 S rRNA is also exhibited by abrin and modeccin, ricin-related toxins, suggesting a general mechanistic pathway for ribosome inactivation by lectin toxins.     

Identification of a novel functional domain of ricin responsible for its potent toxicity

Ribosome-inactivating proteins (RIPs) are toxic N-glycosidases that depurinate the universally conserved α-sarcin loop of large rRNAs. They have received attention in biological and biomedical research because of their unique biological activities toward animals and human cells as cell-killing agents. A better understanding of the depurination mechanism of RIPs could allow us to develop potent neutralizing antibodies and to design efficient immunotoxins for clinical use. Among these RIPs, ricin exhibited remarkable efficacy in depurination activity and highly conserved tertiary structure with other RIPs. It can be considered as a prototype to investigate the depurination mechanism of RIPs. In the present study, we successfully identified a novel functional domain responsible for controlling the depurination activity of ricin, which is located far from the enzymatic active site reported previously. Our study indicated that ricin A-chain mAbs binding to this domain (an α-helix comprising the residues 99-106) exhibited an unusual potent neutralizing ability against ricin in vivo. To further investigate the potential role of the α-helix in regulating the catalytic activity of ricin, ricin A-chain variants with different flexibility of the α-helix were rationally designed. Our data clearly demonstrated that the flexibility of the α-helix is responsible for controlling the depurination activity of ricin and determining the extent of protein synthesis inhibition, suggesting that the conserved α-helix might be considered as a potential target for the prevention and treatment of RIP poisoning.     

Depurination of yeast 26S ribosomal RNA by recombinant ricin A chain

The nature of the modification of yeast ribosomes by the recombinant form of the ricin A chain has been examined. Evidence is presented that the 26S rRNA molecule is depurinated at a specific site and that the activity is inhibited by antibody raised to ricin A chain. It thus appears that the recombinant form of this toxin retains the depurination activity of the native molecule. These results are consistent with the model that the site of depurination is in a highly conserved sequence forming a loop on the surface of the ribosome, a domain involved in elongation factor-dependent binding of aminoacyl-tRNA.     

Inhibition of ricin A-chain with pyrrolidine mimics of the oxacarbenium ion transition state

Ricin A-chain (RTA) catalyzes the hydrolytic depurination of a specific adenosine at position 4324 of 28S rRNA. Kinetic isotope effects on the hydrolysis of a small 10mer stem-tetraloop oligonucleotide substrate established the mechanism of the reaction as D(N)*A(N), involving an oxacarbenium ion intermediate in a highly dissociative transition state. An inhibitor with a protonated 1,4-dideoxy-1,4-imino-D-ribitol moiety, a 4-azasugar mimic, at the depurination site in the tetraloop of a 14mer oligonucleotide with a 5 bp duplex stem structure had previously been shown to bind to RTA with a K(d) of 480 nM, which improved to 12 nM upon addition of adenine. Second-generation stem-tetraloop inhibitors have been synthesized that incorporate a methylene bridge between the nitrogen of a 1-azasugar mimic, namely, (3S,4R)-3-hydroxy-4-(hydroxymethyl)pyrrolidine, and substituents, including phenyl, 8-aza-9-deazaadenyl, and 9-deazaadenyl groups, that mimic the activated leaving group at the transition state. The values for the dissociation constants (K(i)) for these were 99 nM for the phenyl 10mer, 163 and 94 nM for the 8-aza-9-deazaadenyl 10- and 14mers, respectively, and 280 nM for the 9-deazaadenyl 14mer. All of these compounds are among the tightest binding molecules known for RTA. A related phenyl-substituted inhibitor with a deoxyguanosine on the 5'-side of the depurination site was also synthesized on the basis of stem-loop substrate specificity studies. This molecule binds with a K(i) of 26 nM and is the tightest binding "one-piece" inhibitor. 8-Aza-9-deaza- and 9-deazaadenyl substituents provide an increased pK(a) at N7, a protonation site en route to the transition state. The binding of these inhibitors is not improved relative to the binding of their phenyl counterpart, however, suggesting that RTA might also employ protonation at N1 and N3 of the adenine moiety to activate the substrate during catalysis. Studies with methylated adenines support this argument. That the various stem-loop inhibitors have similar potencies suggests that an optimal one-piece inhibitor remains to be identified. The second-generation inhibitors described here incorporate ribose mimics missing the 2-hydroxy group. On the basis of inhibition data and substrate specificity studies, the 2'-hydroxyl group at the depurination site seems to be critical for recruitment as well as catalysis by RTA.     

Perturbation of DNA hairpins containing the EcoRI recognition site by hairpin loops of varying size and composition: physical (NMR and UV) and enzymatic (EcoRI) studies.

We have investigated loop-induced structural perturbation of the stem structure in hairpins d(GAATTCXnGAATTC) (X = A, T and n = 3, 4, 5 and 6) that contain an EcoRI restriction site in close proximity to the hairpin loop. Oligonucleotides containing either a T3 or a A3 loop were not hydrolyzed by the restriction enzyme and also showed only weak binding to EcoRI in the absence of the cofactor Mg2+. In contrast, hairpins with larger loops are hydrolyzed by the enzyme at the scission site next to the loop although the substrate with a A4 loop is significantly more resistant than the oligonucleotide containing a T4 loop. The hairpin structures with 3 loop residues were found to be thermally most stable while larger hairpin loops resulted in structures with lower melting temperatures. The T-loop hairpins are thermally more stable than the hairpins containing the same number of A residues in the loop. As judged from proton NMR spectroscopy and the thermodynamic data, the base pair closest to the hairpin loop did form in all cases studied. The hairpin loops did, however, affect the conformation of the stem structure of the hairpins. From 31P and 1H NMR spectroscopy we conclude that the perturbation of the stem structure is stronger for smaller hairpin loops and that the extent of the perturbation is limited to 2-3 base pairs for hairpins with T3 or A4 loops. Our results demonstrate that hairpin loops modulate the conformation of the stem residues close to the loop and that this in turn reduces the substrate activity for DNA sequence specific proteins.     

The cytotoxins alpha-sarcin and ricin retain their specificity when tested on a synthetic oligoribonucleotide (35-mer) that mimics a region of 28 S ribosomal ribonucleic acid

An oligoribonucleotide (35-mer) that mimics the alpha-sarcin and the ricin region of eukaryotic 28 S rRNA was transcribed in vitro from a synthetic template with T7 RNA polymerase and was used to test whether the specificity of the hydrolysis by the toxins was retained. alpha-Sarcin, at a low concentration, cleaved a single phosphodiester bond on the 3' side of a guanosine residue in the synthetic oligomer that corresponds to G-4325 in 28 S rRNA, the site of action of the toxin in intact ribosomes. At a high concentration of alpha-sarcin, the substrate (35-mer) was hydrolyzed after each of its purines. alpha-Sarcin was without an effect on a synthetic RNA (20-mer) that reproduces the near universal sequence of nucleotides in the loop, but lacks the stem, of the toxin's domain. Thus, the specificity of the attack of alpha-sarcin on a precise region of 28 S rRNA appears to be contingent on the sequence of the nucleotides and the structure of the domain. Ricin depurinated a nucleotide in the synthetic oligomer (35-mer), and in the presence of aniline the phosphoribose backbone was cleaved at a position that conforms to A-4324 in 28 S rRNA, the site of action of the toxin in vivo.     

Self-catalyzed site-specific depurination of G residues mediated by cruciform extrusion in closed circular DNA plasmids

A major variety of "spontaneous" genomic damage is endogenous generation of apurinic sites. Depurination rates vary widely across genomes, occurring with higher frequency at "depurination hot spots." Recently, we discovered a site-specific self-catalyzed depurinating activity in short (14-18 nucleotides) DNA stem-loop-forming sequences with a 5'-G(T/A)GG-3' loop and T·A or G·C as the first base pair at the base of the loop; the 5'-G residue of the loop self-depurinates at least 10(5)-fold faster than random "spontaneous" depurination at pH 5. Formation of the catalytic intermediate for self-depurination in double-stranded DNA requires a stem-loop to extrude as part of a cruciform. In this study, evidence is presented for self-catalyzed depurination mediated by cruciform formation in plasmid DNA in vitro. Cruciform extrusion was confirmed, and its extent was quantitated by digestion of the plasmid with single strand-specific mung bean endonuclease, followed by restriction digestion and sequencing of resulting mung bean-generated fragments. Appearance of the apurinic site in the self-depurinating stem-loop was confirmed by digestion of plasmid DNA with apurinic endonuclease IV, followed by primer extension and/or PCR amplification to detect the endonuclease-generated strand break and identify its location. Self-catalyzed depurination was contingent on the plasmid being supercoiled and was not observed in linearized plasmids, consistent with the presence of the extruded cruciform in the supercoiled plasmid and not in the linear one. These results indicate that self-catalyzed depurination is not unique to single-stranded DNA; rather, it can occur in stem-loop structures extruding from double-stranded DNA and therefore could, in principle, occur in vivo.