Template-Directed Assembly and Characterization of Metallosalen

Dec 23, 2004 - Competitive assembly reactions demonstrated the template-directed nature of metallosalen−DNA formation. A single metallosalen−DNA ...
0 downloads 9 Views 182KB Size
Download PDF
Bioconjugate Chem. 2005, 16, 169−177

169

Template-Directed Assembly and Characterization of Metallosalen-DNA Conjugates Jennifer L. Czlapinski† and Terry L. Sheppard* Department of Chemistry and The Robert H. Lurie Comprehensive Cancer Center, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208-3113. Received October 17, 2004

Nucleic acid template-directed synthesis represents a powerful method for the encoded synthesis of new bioconjugates. Our laboratory previously reported a strategy for the synthesis of a new metalDNA hybrid, metallosalen-DNA, by the DNA or RNA template-directed cross-linking of two salicylaldehyde-modified DNA oligonucleotides. The current manuscript describes the optimal assembly requirements and biophysical characterization of metallosalen-DNA conjugates containing nickel and manganese ions. Competitive assembly reactions demonstrated the template-directed nature of metallosalen-DNA formation. A single metallosalen-DNA conjugate was assembled selectively in the presence of two pairs of salicylaldehyde precursor strands and a single DNA template. Assembly reactions were sensitive to base pair mismatches in the pairing arms. Single base mismatches resulted in a loss of metallosalen-DNA conjugate yield. Metallosalen-DNA assembly yields depended on the identity and length of the template spacer, the reaction pH, and the type of metal and diamine utilized in the assembly reaction. Metallosalen-DNA conjugates were stable to a variety of conditions, including extended incubation at 50 °C. Nickel metallosalen-DNA remained unchanged after incubation at 80 °C for 24 h, while decomposition of manganese metallosalen-DNA was observed under the same conditions. Circular dichroism (CD) spectroscopy indicated that DNA duplexes containing internal metallosalen moieties adopted B-form double helices. UV thermal denaturation analysis demonstrated that 32-nucleotide duplexes containing internal metallosalen modifications displayed melting temperatures ∼5 °C less than unmodified DNA duplexes.

INTRODUCTION

Metal ions are essential to nucleic acid structure and biochemistry. Nucleic acid duplexes and RNA tertiary structures are stabilized by metal ion binding (1, 2), and numerous catalytic nucleic acids require divalent metal ions for their function (3, 4). Despite a growing understanding of nucleic acid-metal interactions, the rational design or combinatorial discovery of metal-binding sites is still in its infancy. However, chemists have united the properties of metal ions with the coding abilities of nucleic acids by developing methods for the covalent conjugation of metal complexes to oligonucleotides. Early metal-DNA complexes involved the covalent attachment of EDTA to an oligonucleotide (5), for use in ironmediated DNA cleavage. Since this discovery, numerous metal-DNA hybrids have been developed for applications ranging from nucleic acid cleavage (6-12) to DNA electron transfer studies (13-19). Other metal-DNA conjugates have been utilized as DNA hybridization probes and sensors (20-23) and for mechanistic studies of the anticancer drug, cisplatin (24). Recently, metal complexes have been utilized as scaffolds for metalmediated DNA base pairs (25-31). Metal-DNA hybrids have been synthesized by several methods. One common approach involved the conjugation of the metal complex to a functional group “handle” on a DNA oligonucleotide (17-21). A second procedure was * To whom correspondence should be addressed. Tel: 847467-7636; Fax: 847-491-7713; E-mail: t-sheppard@northwestern. edu. † Current address: Department of Chemistry, University of California, Berkeley, CA 94720.

based on the synthesis of oligonucleotides modified with metal ligands, followed by metal complexation (5-8, 22). Metal complexes also have been inserted into oligonucleotides directly by DNA automated synthesis (12-15, 23, 24). In a recent advance, a DNA polymerase was used to incorporate a metal-containing nucleoside triphosphate into an oligonucleotide (32). Our laboratory became interested in template-directed synthesis as a new method for the assembly of metal-DNA hybrids. In template-directed synthesis, substrate oligonucleotides bearing complementary reactive groups are brought into proximity by Watson-Crick base pairing with a nucleic acid template, which accelerates the selective formation of chemical bonds between the functionalized oligonucleotides. Initially, template-directed synthesis employed oligonucleotide strands containing chemically modified phosphate backbones (33-35). More recently, template-directed ligation reactions were expanded for the detection of DNA and RNA point mutations (36, 37). Reversible chemical ligation strategies also have been developed (33, 38) and adapted for the ligation of thymidine oligomers (39). Remarkably, DNA and RNA templates have directed the synthesis of a range of chemical structures across diverse reaction classes (4046). In addition, nucleic acid mimics, including hexitol nucleic acids (47) and peptide nucleic acids (48, 49), have demonstrated potential as templates in directed assembly reactions. More recently, template-directed synthesis has supported the construction of complex organic moleculeDNA conjugates (50-52), in some cases, with stereoselectivity (53) and in multistep transformations (54). We recently reported a new metal-DNA hybrid, metallosalen-DNA, which was assembled by DNA or RNA

10.1021/bc049753r CCC: $30.25 © 2005 American Chemical Society Published on Web 12/23/2004

170 Bioconjugate Chem., Vol. 16, No. 1, 2005 Scheme 1. Template-directed Conjugate Assembly.

Metallosalen-DNA

template-directed synthesis (55, 56). In our method, the metal complex was site-specifically incorporated into DNA by the diamine cross-linking of salicylaldehydemodified DNA on an external template (Scheme 1). Two complementary DNA oligonucleotides were modified with a salicylaldehyde moiety at either the 3′- or 5′-end (1 and 2, respectively). The substrates were aligned on the template strand, and the addition of a diamine and a divalent metal ion led to the formation of a metallosalenDNA hybrid. Herein we describe the variables that control optimal template-directed assembly of metallosalen-DNA conjugates containing nickel and manganese ions. Metallosalen-DNA assembly is controlled by the composition and length of template spacers that bridge the assembly site, reaction pH, the identities of the metal ion and diamine, and the presence of a complementary DNA template in the assembly reaction. MetallosalenDNA conjugates form B-form DNA duplexes, as judged by circular dichroism (CD) studies, and their presence destabilizes a 32-oligonucleotide duplex by ∼5 °C in melting temperature. EXPERIMENTAL PROCEDURES

General. All molecular biology reagents were purchased from Sigma. Metals for metallosalen-DNA conjugate formation were purchased from Aldrich as metal acetates, unless otherwise stated. T4 Polynucleotide kinase was obtained from US Biochemical, and [γ-32P]ATP (7000 Ci/mmol) was purchased from ICN. Radioactive bands in polyacrylamide gels were visualized using a Molecular Dynamics Storm Phosphorimager and quantitated using Molecular Dynamics ImageQuant software. 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 2-(4-morpholino)ethanesulfonic acid (MES), and 2-(cyclohexylamino)ethanesulfonic acid (CHES) were purchased as molecular biology grade reagents from Sigma. Matrix-assisted laser desorption ionization timeof-flight (MALDI-TOF) mass spectrometry was performed on a PerSeptive Biosystems, Inc. (Foster City, CA), Voyager-DE PRO Biospectrometry Workstation MALDITOF mass spectrometer as described (55). DNA Synthesis, Deprotection, and Purification. Standard unmodified DNA oligonucleotides were synthesized by Integrated DNA Technologies (Coralville, IA). Modified oligonucleotides were synthesized on a Pharmacia Gene Assembler Plus at the 1.3 µmol scale using phosphoramidites with labile base protecting groups for the natural nucleotides, as described (55). All oligonucleotides were purified by reverse-phase HPLC (RP-HPLC), lyophilized, and resuspended in 0.5× Tris-EDTA buffer (5 mM Tris and 0.5 mM EDTA, pH ) 7.5). Purified oligonucleotides were quantified by UV spectroscopy, as previously described (55).

Czlapinski and Sheppard

5′-GCCGATACCACGCTCTTCACCGACGATTGCCT-3′ (7). Oligonucleotide 7 was synthesized by Integrated DNA Technologies and purified by RP-HPLC. Salicylaldehyde-Modified DNA Synthesis. Salicylaldehyde-modified DNA strands were synthesized by solid-phase DNA synthesis using standard methods for the natural nucleoside phosphoramidites and 2-(1,3dioxan-2-yl)-3-benzoyloxyphenyl-4-ethyl-O-(2-cyanoethyl)N,N′-diisopropylphosphoramidite (55) (27, 0.11 M solution in anhydrous CH3CN). After synthesis, oligonucleotides were deprotected as described (55) and purified by RP-HPLC. 5′-AGGCAATCGTCGGTG-SAL-3′ (5). Oligonucleotide 5 was synthesized in the 5′-to-3′ direction using a dA-5′-CPG 500 solid support (Glen Research, 33.4 mg, 31 µmol/g loading, 1.03 µmol scale), using standard 5′β-cyanoethyl nucleoside phosphoramidites and 27. MALDITOF (m/z): [M]- calcd for 5, 4860; found, 4860. 5′-SAL-GAGCGTGGTATCGGC-3′ (6). Oligonucleotide 6 was synthesized in the traditional 3′-to-5′ direction using standard 3′-nucleoside phosphoramidites (Glen Research, Ultramild) and 27 on a 1.3 µmol scale. MALDITOF (m/z): [M]- calcd for 6, 4877; found, 4877. 5′-Radiolabeling. Oligonucleotide 1 or 5 (24 pmol) was radiolabeled as described (55). The reaction was purified by 20% PAGE, eluted from the gel slice (10 mM Tris-HCl, 200 mM NaCl, 1 mM EDTA, pH 7.5), and recovered by precipitation with two volumes of 2-propanol, which had been supplemented to 0.3 M NaOAc, pH 5.2. The oligonucleotide was redissolved in ddH2O. Mn Metallosalen-DNA Conjugate Formation Reactions (Mn-4 or Mn-8). A mixture of 1-3 (or 5-7), each at 4 µM in 10 mM HEPES, pH 8.0 and 150 mM NaCl, were annealed by heating to 95 °C and slow cooling to RT over 2 h. Solutions of manganese acetate (2 mM) and ethylenediamine (EN, 500 µM) were added to give final concentrations of 400 µM Mn(OAc)2 and 100 µM EN. Each strand was present in a final concentration of 2 µM. The solution was incubated at 37 °C for 1 h. Preparative scale reactions were performed on a 4 nmol scale in 50 mM HEPES, pH 8.0 and 150 mM NaCl. Conjugates were analyzed and purified by 20% polyacrylamide gel electrophoresis (PAGE) using TB buffer (89 mM Tris, 89 mM borate, no EDTA). The metallosalen-DNA complex was visualized using a handheld UV lamp and recovered by electroelution with TB buffer in a Schleicher & Schuell Elutrap (30 min, 200 V at RT). Mn-4 or Mn-8 were concentrated in a Millipore Microcon YM-3 filter, washed twice with 200 µL of ddH2O, and eluted into 50 µL of ddH2O for storage. Ni Metallosalen-DNA Conjugate Formation Reactions (Ni-4 or Ni-8). Ni metallosalen-DNA formation reactions followed the same general procedure as for Mn; however, the Ni(OAc)2 and EN had final concentrations of 300 µM and 150 µM, respectively, in a pH 6.5 buffer solution (10 mM MES and 150 mM NaCl). Preparative scale reactions required no changes in the reaction buffer. Competitive Metallosalen-DNA Assembly. Two sets of reactions were prepared, one for formation of Ni-4 and the other for Mn-4, using the buffer conditions listed above. For each metal, three reactions were performed: (1) a mixture of strands 1-3 (2 µM) with 20 fmol of radiolabeled 1; (2) 2 µM each of 1-3 and 5-6 with 20 fmol of radiolabeled 1; and (3) 2 µM each of 1-3 and 5-6 with 20 fmol of radiolabeled 5. The appropriate amount of metal and diamine was added, and the reactions were incubated at 37 °C and analyzed by 20% PAGE using 1× TB as the running buffer.

Template-Directed Metallosalen−DNA Assembly

Nonspecific Metallosalen-DNA Assembly. To assay for nonspecific metallosalen-DNA formation, a series of assembly reactions was performed for each metal ion. Two reactions were assembled for each time point: one that contained 1-3 (each at 2 µM) with 20 fmol radiolabeled 1, and the other with 2 µM of 1 and 2 with 20 fmol radiolabeled 1. Since assembly could not be quenched, the reactions were initiated in order of decreasing time. At the end of the time course, products were analyzed by 20% PAGE with TB as the running buffer. Mismatch Metallosalen-DNA Formation. Reactions containing 2 µM each of 1, 2, template 3 or one of the mismatched templates 9-13, and 20 fmol of radiolabeled 1, were initiated under standard conditions. Reactions were performed with both Mn(OAc)2 and Ni(OAc)2 for each template and were assayed by 20% PAGE electrophoresis with TB buffer. Metallosalen-DNA Formation with Diverse Metal Ions. Metal ion assays were performed using 2 µM each of 1-3 with 20 fmol of radiolabeled 1. Each reaction was incubated for 1 h at 37 °C with 500 µM of metal salt and 250 µM of EN in 10 mM buffer at the required pH with 150 mM NaCl, as indicated by metal speciation data (57). Cu(OAc)2 at pH 9.0 (CHES), VOSO4 at pH 7.0 (HEPES), and RhCl3 and ZnSO4 at pH 7.5 (HEPES). Reaction products were analyzed by PAGE using 1× TB as the running buffer. Metallosalen-DNA Thermal Stability and EDTA Decomposition. Radiolabeled Mn-4 and Ni-4 conjugates were prepared as described above. For thermal stability, solutions containing radiolabeled conjugate 4 were incubated at various temperatures (RT, 37 °C, 50 °C, and 80 °C), and reaction aliquots were taken over 48 h. The time points were analyzed by 20% PAGE with TB running buffer. For EDTA decomposition analysis, the radiolabeled conjugates were prepared in smaller quantities (2 µM strand concentrations) with 20 fmol of radiolabeled 1. Mn-4 and Ni-4 formation reactions were treated with EDTA, pH 8.0 (final concentration: 0-40 mM) for 15 min at RT. Product decomposition was followed by 20% PAGE with TB as the running buffer. Circular Dichroism Spectroscopy. CD spectra were recorded using a Jasco J-715 CD spectropolarimeter. Oligonucleotide duplex samples were prepared by mixing solutions of 1 µM of each strand in 10 mM sodium phosphate, 150 mM NaCl, pH 7.0. Spectra were recorded from 200 to 350 nm at 25 °C in a quartz cuvette (1 cm path length). The sample compartment was purged with inert atmosphere (N2) during data acquisition. Each spectrum represents a baseline corrected average of four accumulations per scan, and all data were corrected by subtracting the buffer reference signal. UV Thermal Denaturation Analysis. UV thermal denaturation data were acquired on a Cary 500 spectrophotometer equipped with a multicuvette thermoelectric controller in 1 cm quartz cuvettes. Solutions of 1 µM duplex oligonucleotides in 10 mM sodium phosphate, 150 mM NaCl, pH ) 7.0 were annealed by heating to 95 °C for 5 min, then cooled to 25 °C over 2 h. The samples were degassed under vacuum for 3 min prior to melting analysis. The absorbance was measured at 260 nm as the samples were heated from 30 °C to 90 °C at 0.5 °C/ min. Data sets for the downward transition from 90 °C to 30 °C also were collected at 0.5 °C/min; all melting profiles were fully reversible to within 1.0 °C. Melting temperatures (Tm) were determined from plots of dA260/ dT vs T as described (58) and are derived from replicate melting experiments using at least two independently prepared samples.

Bioconjugate Chem., Vol. 16, No. 1, 2005 171

Figure 1. Metallosalen-DNA formation reactions for Mn-4 and Ni-4. (A) SAL-DNA sequences 5 and 6 align on complementary template 7 to support assembly of Mn-8 and Ni-8. (B) Gel electrophoresis assay of metallosalen-DNA competitive assembly. Lane 1: 10 bp marker. Lanes 2-4: Mn assembly reactions at pH 8.0 for 1 h. Lane 2: All components: 400 µM Mn(OAc)2, 1, 2, and template 3, 100 µM EN, radiolabeled 1. Lane 3: +SAL strands 5 and 6. Lane 4: +5 and 6, -radiolabeled 1 and + radiolabeled 5. Lanes 5-7: Analogous reactions to Lanes 2-4, but with 300 µM Ni(OAc)2 and 150 µM EN at pH 6.5 for 24 h. Lane 8: Radiolabeled 1. RESULTS AND DISCUSSION

Generality of Template-Directed MetallosalenDNA Synthesis. Our approach to metallosalen-DNA by template-directed synthesis was reported previously for the formation of Mn-4 and Ni-4, as shown in Scheme 1 (55). We provided evidence that complementary template 3, a diamine, and either manganese or nickel ions were required for the assembly of metallosalen-DNA 4 from 1 and 2. To demonstrate the generality of templatedirected synthesis of metallosalen-DNA, a second set of SAL-DNA precursors were prepared that varied the DNA sequences flanking the metallosalen-DNA. SALDNA sequences (5 and 6, Figure 1A) were synthesized, purified, and characterized by MALDI-TOF MS as described previously for 1 and 2 (55). Template-directed assembly reactions of metallosalen-DNA conjugates Mn-8 and Ni-8 from 5 and 6 and the complementary template 7 were analyzed by denaturing polyacrylamide gel electrophoresis (PAGE). When DNA 5 and 6 were hybridized to template 7 and incubated at 37 °C in the presence of 100 µM ethylenediamine (EN) and 400 µM Mn(OAc)2, a new DNA product, corresponding to Mn-8, was produced in 50% yield. Similarly, Ni-8 was prepared in 70% yield after 24 h, when 150 µM EN and 300 µM Ni(OAc)2 were incubated with 5 and 6 in the presence of 7. Metallosalen-DNA assembly required EN, divalent metal ion, and template 7 for optimal assembly, and Mn-8 and Ni-8 were produced in yields comparable to those of conjugates 4 (55). To confirm the identities of these new metallosalen-DNA hybrids, preparative syntheses (4 nmol) of Mn-8 and Ni-8 were performed with an RNA template. Purification of the conjugates was attempted by selective RNase H digestion of the RNA template as previously described (55). Although this strategy was effective for the purification of Mn-4 and Ni-4, it was not reliable when applied to metallosalen-

172 Bioconjugate Chem., Vol. 16, No. 1, 2005

DNA conjugates Mn-8 and Ni-8. Thus, a more general method for the purification of metallosalen-DNA conjugates was developed, which utilized complementary DNA templates that were four nucleotides shorter than 7 (two nucleotides were removed from both the 5′- and 3′-ends). The shorter templates facilitated the separation of the desired metallosalen-DNA from the template strand by gel electrophoresis. Using this approach, Mn-8 and Ni-8 were purified by 20% PAGE, and their identities were confirmed by MALDI-TOF MS and base composition analysis (55) (see Supporting Information). Template Specificity of Metallosalen-DNA Assembly. Given the sequence generality of metallosalenDNA formation, further studies were performed to evaluate the effects of template specificity on the assembly reactions. To assay the template dependence of metallosalen-DNA conjugate assembly, a series of competitive assembly reactions were performed. Four noncomplementary salicylaldehyde-modified (SAL) oligonucleotides were combined with a template that was complementary to only one pair of SAL strands. Upon addition of a diamine and a divalent metal ion, only one metallosalenDNA should be produced, dependent on the identity of the template strand present in the reaction. A series of reactions using the four SAL strands (1-2 and 5-6) and complementary DNA templates, 3 and 7, respectively, were undertaken to assay metallosalen-DNA formation under competitive conditions. Three independent metallosalen-DNA reactions were performed and assayed by PAGE using radiolabeled SAL strands as tracers for the assembly of two different metallosalens, 4 and 8. The first reaction was a typical assembly of 4 (Scheme 1) from SAL strands 1 and 2, and template 3, where radiolabeled 1 was used as a tracer. The second and third reactions contained four SAL strands, 1/2 and 5/6, and the complementary template to 1 and 2 (oligonucleotide 3). The second reaction, which included the radioactive tracer strand 1 (to track assembly of 4), was designed to demonstrate that the correct templated conjugate, 4, was formed even in the presence of other SAL oligonucleotides. The presence of labeled strand 5 in the third reaction was used to follow the assembly of 8 and show that strands 5 and 6 do not form metallosalen-DNA conjugate 8 in the presences of the “wrong” template (3). Gel electrophoretic analysis of the competition reactions is shown in Figure 1B for metallosalen-DNA assembly with manganese and nickel ions. When 1 and 2 were hybridized to their complementary template 3 and incubated for 1 h at 37 °C in the presence of 100 µM EN and 400 µM Mn(OAc)2, Mn-4 was produced in 57% yield (lane 2). The addition of SAL strands 5 and 6, which were noncomplementary to 3, did not disrupt Mn-4 formation (58%, lane 3). In the absence of labeled 1 (lane 4) and the presence of labeled 5, no formation of Mn-8 was observed. The assembly reactions performed with nickel provided similar results. When 1 and 2 were hybridized to complementary 3 and incubated for 24 h at 37 °C in the presence of 150 µM EN and 300 µM Ni(OAc)2, Ni-4 was produced in 73% yield (lane 5). The addition of unlabeled noncomplementary SAL oligonucleotides (5 and 6) did not alter the yield of Ni-4 (72%, lane 6). However, when noncomplementary 1 was present as the labeled strand, no assembly of metallosalen-DNA Ni-4 was observed (lane 7). Taken together, the results shown in Figure 1B demonstrated that metallosalen-DNA formation was template-directed, and that a single DNA template was capable of directing the synthesis of a unique metallosalen-DNA conjugate from a mixture of SAL-DNA strands.

Czlapinski and Sheppard Table 1. Mismatched Template Sequences and Yields for Mn-4 Formation Fidelitya % yield Mn-4

sequence 5′-GCTACCGAATACGCTTTTGCCTACGAACCGCT-3′ 5′-GCTACCGAATACGCTTTGGCCTACGAACCGCT-3′ 5′-GCTACCGAATACGCTTTCGCCTACGAACCGCT-3′ 5′-GCTACCGACTACGCTTTCTGCCTACGAACCGCT-3′ 5′-GCTACCGAATACGCTTTTGCCTACGCAACCGCT-3′ 5′-GCTACC_AATACGCTTTTGCCTACGAACCGCT-3′

(3) (9) (10) (11) (12) (13)

64 58 60 55 51 40

a The bolded/underlined letter indicates the mismatched nucleotide; the nucleotide in bold italics is an additional base; and the underscore shows a missing nucleotide.

Nonspecific Metallosalen-DNA Assembly and Effects of Mismatched Base Pairs. To better assess how DNA templates enhance the rates of metallosalenDNA assembly reactions, a measurement of the background rate of nonspecific metallosalen assembly was required. Nonspecific metallosalen-DNA assembly results from the cross-linking of two SAL-DNA strands in the absence of the corresponding DNA template. To monitor the nonspecific formation of metallosalen-DNA in reactions that lacked a template strand, the background rate of nonspecific metallosalen-DNA assembly was measured by following the formation of metallosalen-DNA product in a reaction containing only the two noncomplementary SAL-DNA strands in the presence of a diamine and a divalent metal ion, but in the absence of a template DNA. SAL strands 1 and 2 (including radiolabeled 1 as a tracer) were incubated under standard assembly conditions. For reactions with Mn(II), the nonspecific yield of material that comigrated with Mn-4 was found to be 1% at 1 h. Nonspecific product formation increased gradually to 34% at 48 h of reaction. Although the yields of nonspecific conjugate formation increased at longer incubation times, template-directed Mn-4 reactions typically were completed in 1 h (70% yield), which guaranteed that the major product of template-directed synthesis was the desired Mn-4 under these conditions. As seen earlier (56), nonspecific assembly reactions with Ni(II) were considerably slower than those with manganese ions. No detectable nonspecific products were formed within 48 h, which was 24 h after templated Ni-4 and Ni-8 formations were generally complete. Thus, both manganese and nickel ion template-directed metallosalen-DNA assembly reactions were completed before significant amounts of nonspecific products were formed. Numerous template-directed chemical ligation reactions have been shown to be sensitive to single base pair mismatches at or near the ligation site (36, 43, 45, 48). As a result, the effect of base pair mismatches on metallosalen-DNA conjugate formation was assayed. A series of DNA templates (Table 1) for synthesis of metallosalen-DNA 4 (template a) or 8 (template b) were prepared that contained a range of mismatch types and sites: adjacent to the metallosalen assembly site (9 and 10), at the center of one metallosalen-DNA arm (11), an insertion of a dC nucleotide (12) or deletion of a single dG nucleotide (13) in the template strands. Utilizing the conditions for Mn-4 formation, the four mismatched templates 9-13 were hybridized with SAL strands 1 and 2 and incubated in the presence of 100 µM EN and 400 µM Mn(OAc)2 at 45 °C for 1 h. Under these conditions, the mismatched templates led to lower yields of Mn-4 (Table 1). The Mn-4 yields with mismatched templates 9-13 were between 4 and 24% lower than with complementary template 3 (64%). When the mismatches were near the metallosalen formation site, the yields did not decrease significantly: 9 (58%, G‚A

Bioconjugate Chem., Vol. 16, No. 1, 2005 173

Template-Directed Metallosalen−DNA Assembly

Table 2. Assembly Yields for Mn-4 and Ni-4 Metallosalen-DNA Conjugates with Different Nucleotide Spacer Templates % yield

Figure 2. Template sequences and spacers used to evaluate the effects of spacer length and composition on metallosalenDNA conjugate (4 and 8) formation. The spacer identities include a three-carbon linker (14), a nine-atom glycol linker (15), an eighteen-atom glycol linker (16), two D-spacers (17), and various dinucleotides: AT (18), GC (19), GG (20), AA (21), and CC (22).

mismatch) and 10 (60%, C‚A mismatch). A more substantial decrease in yield occurred when the mismatch was centrally located within a DNA arm of the metallosalen (11, C‚T mismatch, 55%) or with a single base insertion (12, 51%). The best discrimination was observed with a single base deletion (13, 40%). The results in Table 1 indicate that metallosalen-DNA formation may not be sensitive to mismatches near central assembly site. Due to the lack of standard base pairing of the salicylaldehyde groups with bases in the templates and the likely requirement for flexibility at the assembly site, the loss of an additional base pair interaction by mismatch pairing may not perturb the assembly process. Nonetheless, there is a measurable decrease in yield for the other mismatched templates. Effects of Template Spacers, Reaction pH, and Metal Ion Identity on Metallosalen-DNA Assembly Reactions. We previously established that efficient template-directed metallosalen-DNA assembly was dependent on several variables, including the identity of the template spacer nucleotides, the metal ion selected for the assembly reaction, and pH that was optimal for the selected metal ion (55). We now elaborate on the variables that affect the efficiency of template-directed metallosalen-DNA synthesis. Other studies have revealed that the rate of bond formation in a template-directed reaction is relatively independent of the distance between the reactive partners (51). Thus, we evaluated the effect of the chemical identity and length of the central template spacer (See Scheme 1, Figure 1A) on the efficiency of metallosalenDNA assembly reactions. Our initial report of metallosalen-DNA assembly (55) employed a DNA template with a central spacer consisting of two thymidine residues (TT, Scheme 1). We prepared a panel of template spacers to determine the effect of the chemical nature and the distance of the spacer groups on template-directed synthesis yields (Figure 2). Specifically, we examined spacers with flexible backbones (14-16), D-spacer abasic site analogues (17), and dinucleotides (18-22), which are shown in Figure 2. The optimal template for each metallosalen-DNA conjugate depended on the metal ion used in the assembly reaction and, in some cases, the oligonucleotide sequence. For example, significant template spacer effects were observed for assembly of manganese-containing metallosalen-DNA. For Mn-4 assembly, optimal yields were achieved with template 3 (66%,

template

spacer

Mn-4

Ni-4

3 18 19 20 21 22

TT AT GC GG AA CC

77 54 66 68 71 71

86 78 72 73 76 74

TT spacer), followed by 14a (51%, two three carbon spacers). The two glycol spacer templates, 15a and 16a, gave the next highest yields, at 49% each. The best template for Mn-8 formation was the nine-atom glycol spacer in 15b (61%), followed by 16b (55%, eighteen atom glycol spacer), and 7 (50%, TT spacer). The rest of the templates gave similar yields of Mn-8: 17b (two Dspacers, 25%) and 14b (21%). In contrast, more consistent results were observed for the assembly of the two different nickel-containing metallosalens. Ni-4 and Ni-8 formation was optimal with template spacers of similar type. For Ni-8, the original template 7 (TT spacer) exhibited the highest yield (70%). The next highest yielding template was 16b (67%), while the rest of the templates yielded 30-40% less Ni-8 than the standard thymidine dinucleotide template (7). The results for Ni-4 were similar: the original template 3, containing two thymidines, gave the optimal yield (69%). Other templates yielded between 12 and 62% less of the desired Ni-4, with 15a yielding the least Ni-4 (6%). In general, the formation of conjugates 4 and 8 was optimal when the templates contained natural nucleotides (T spacers, 3 or 7) compared with the flexible organic spacers (14-16) and the abasic “D-spacers” (17). The organic spacers were chosen to provide flexibility and low steric demand at the assembly site. Likewise, templates with D-spacers were designed to maintain the natural helical pitch of the DNA backbone while eliminating steric hindrance at the assembly site by nucleotide bases. The results shown here were unexpected, given that templates that contained non-nucleotidyl spacers yielded 12 to 62% less metallosalen-DNA conjugate than the templates containing two natural thymidine residues. With the successful formation of metallosalen-DNA conjugates using templates containing two thymidine spacers, other dinucleotide template spacers were evaluated: AT (18a), GC (19a), GG (20a), AA (21a), and CC (22a, Figure 2). Using 3 and the new templates 18a22a, the assembly reactions for Mn-4 and Ni-4 were examined, and the results are shown in Table 2. The yields for the metallosalen-DNA conjugates (4) with nickel and manganese ions showed comparable trends. Overall, template 3, with its TT spacer, provided the maximal yield for the metallosalen-DNA conjugate 4 (77% for Mn-4 and 86% for Ni-4). For templates 19a22a, which varied the identity of the internal dinucleotide spacer, results for Mn-4 and Ni-4 assembly reactions followed a similar trend: TT (3) . AA (21a) g CC (22a) > GG (20a) > GC (19a). Template 18a (AT spacer) was unusual in that conjugate yields were dependent on the metal used. Template 18a yielded 54% of Mn-4, which was lower than with the GC spacer (19a, 66%). However, yields of Ni-4 with 18a (78%) were similar to AA (76%, 21a) and CC (74%, 22a). We suggest that the differences may be accounted for by differential metal ion binding abilities of the spacer sites or by unique hydrogen bonding interactions with the salicylaldehyde moiety of

174 Bioconjugate Chem., Vol. 16, No. 1, 2005

the SAL-DNA precursors. Evaluation of these hypotheses will require further structural studies of duplexes containing metallosalen-DNA. Nevertheless, Table 2 clearly shows that, of the templates screened, templates with two thymidine nucleotide spacers (3 or 7) are the optimal complementary strands for formation of the corresponding metallosalen-DNA derivatives. A major reason for the initial selection of metallosalens as a target for template-directed synthesis was their ability to accommodate a variety of metal ions and diamines (59). To determine the range of metal ions and diamines that were able to support metallosalen-DNA assembly, a series of assays was performed using SAL strands 1 and 2 with template 3. Metal ion assays were performed using concentrations of 500 µM of each metal and 250 µM of EN. The reactions were evaluated after 1 and 24 h of incubation at 37 °C and assayed by PAGE. Several water-soluble transition metals were assayed for their metallosalen-DNA formation ability, and the metal ions that supported metallosalen-DNA assembly were Mn(III), Ni(II), VO(II), Zn(II), Cu(II), and Rh(III) (see Supporting Information). As demonstrated previously, Mn(OAc)2 and Ni(OAc)2 provided the highest yields of metallosalen-DNA conjugates. Iron (II) also displayed early promise, but DNA aggregation has been observed above concentrations of 40 µM (60), which limited further studies. Copper(II) produced metallosalen-DNA conjugate in 36% yield, and the yields of VO-4 and Zn-4 were 15 and 21%, respectively. Formation of Rh-4 proceeded poorly, in a maximal yield of 5%. Other studies have shown that metallosalen-DNA conjugate assembly reactions also accept Al(III) (61), opening the possibility for template directed synthesis of more diverse metallosalen-DNA conjugates. All of the work described to date utilized ethylenediamine (EN) (55, 56) as the cross-linker for the two precursor SAL-DNA strands. To explore the tolerance of the metallosalen-DNA assembly reaction to other diamines, two diamines were examined for metallosalenDNA conjugation reactions: racemic trans-1, 2-diaminocyclohexane (CA) and 1,2-phenylenediamine (PA). CA was chosen because its molecular structure is more constrained than EN, while PA was selected to investigate formation with a rigid, achiral diamine. Utilizing the standard assembly conditions for Mn-4 and Ni-4, the comparative ability of EN, CA, and PA to support metallosalen-DNA formation was assessed by gel electrophoretic studies. The yield of metallosalen-DNA, composed of different diamines, was dependent on the combination of metal and diamine used in the assembly reaction. For assembly with Mn(II), metallosalen-DNA yields for CA were similar to EN (54% and 55%, respectively), whereas the metallosalen derived from PA was formed in significantly lower yield (16%). The opposite trend occurred with nickel-metallosalen assembly: PA and EN yielded 44% and 40% of the metal-DNA respectively, while CA yielded 3% of the Ni metallosalen-DNA conjugate. These investigations suggested that use of diverse diamines should support the construction of a broad range of metallosalen-DNA backbones, including chiral diamine cross-linkers. Physical Properties of the Metallosalen-DNA Conjugates. A second motivation for selection of metallosalen-DNA as a target for DNA template-directed synthesis was our interest in using metallosalen-DNA for applications in targeted DNA cleavage (62), catalysis, and materials design. As a result, we became interested in the inherent chemical and thermal stability of metallosalen-DNA and the stability of DNA duplexes con-

Czlapinski and Sheppard

taining metallosalen-DNA. To evaluate the metallosalenDNA conjugates for thermal and chemical stability, radioactive samples of Mn-4 and Ni-4 were prepared using 0.4 nmol of 1, 2, template 3, and a trace amount of radiolabeled 1 with the appropriate concentrations of metal and EN. The assembly reactions were incubated at 37 °C for 1 h (Mn-4) or 24 h (Ni-4) and purified by PAGE. The purified metallosalens, Mn-4 or Ni-4, were subjected to various conditions, and the rates of decomposition were assayed by PAGE. The effect of temperature on metallosalen-DNA stability was explored. At room temperature, there was no detected decomposition of Mn-4 or Ni-4 over several days. Incubation at higher temperatures also demonstrated little effect on the conjugates: no Mn-4 or Ni-4 decomposition was seen over 2 days at 50 °C. The degradation of Mn-4 was slow (1-2% per h) at 80 °C and produced a product that comigrated with the SAL-DNA starting material. Thus, we hypothesized that the major decomposition mechanism was hydrolysis of the metallosalen. Conjugate Ni-4 displayed no degradation at 80 °C over 24 h. Repeated thaws from -20 °C of nonradiolabeled Mn-4 and Ni-4 samples over several months to a year proceeded with no detectable loss of the conjugates. Given the demonstrated thermal stability of conjugates Mn-4 and Ni-4, we examined the stability of the metallosalen groups to the metal chelator, EDTA. Early in our investigations of template-directed metallosalen-DNA syntheses, EDTA was removed from all buffers to prevent removal of the metal from the salen ligand. The extent of metallosalen-DNA EDTA decomposition was assayed by incubation of Mn-4 and Ni-4 radiolabeled samples with increasing concentrations of EDTA (0-40 mM) for 15 min. The only visible products were the starting strand 1 and the metallosalen-DNA conjugate, 4, suggesting that the major mechanism for decomposition was salen hydrolysis following metal removal. The Mn-4 conjugate was stable in buffers containing up to 5 mM EDTA, above which the fraction of Mn-4 remaining steadily decreased. At and above 40 mM EDTA, the amount of Mn-4 remained at 6% of the starting amount. In contrast to the instability of Mn-4, Ni-4 was stable to all EDTA concentrations examined, with no decomposition to component SAL-DNA strands being detected. Thus, the thermal and EDTA stability assessments of Mn-4 and Ni-4 indicated that nickel metallosalen-DNA conjugates were the more robust complexes, and the manganese metallosalen-DNA were more labile. With the stabilities of the metallosalen-DNA conjugates established, the helical properties of the metallosalen-DNA conjugates 4 and 8 within duplex DNA were examined by circular dichroism (CD) spectroscopy. Metallosalen-DNA strands were hybridized to their complementary templates (4•3 and 8•7, TT spacers) in 10 mM sodium phosphate, 150 mM NaCl, pH 7.0. For comparison, all-DNA duplexes of templates 3 and 7 with complementary oligonucleotides (23 and 24, respectively) also were evaluated. Representative CD spectra for Mn4•3 and Ni-4•3 and 3•23 are compared in Figure 3A. The spectrum of each duplex displays a positive band at 280 nm and a negative band at 250 nm, indicative of a B-form helix (63). All metallosalen-DNA conjugate duplexes provided similar CD spectra (see Supporting Information), indicating that the metallosalen does not disrupt B-form helicity. The thermal stabilities of the metallosalen-DNA/ template duplexes were evaluated by UV-thermal denaturation studies. Utilizing the same annealing conditions as the CD studies, the UV melting profiles were obtained

Template-Directed Metallosalen−DNA Assembly

Bioconjugate Chem., Vol. 16, No. 1, 2005 175

conjugates was shown to be specific and templatedirected, even in the presence of noncomplementary SAL-DNA strands. The optimal templates for efficient metallosalen-DNA conjugate synthesis contained dinucleotide spacers, usually TT, compared to flexible organic spacers or abasic site analogues. CD spectroscopy demonstrated that duplexes containing metallosalens adopted B-form helices, and UV-thermal denaturation studies indicated that the metallosalen destabilizes 32mer duplexes by 5-6 °C. These studies establish the key features necessary for the template-directed syntheses and applications of metallosalen-DNA. ACKNOWLEDGMENT

We acknowledge funding from the ACS Petroleum Research Fund (34740-G4). J.L.C. was supported by an Institutional NRSA Training Grant in Molecular Biophysics (GM08382). T.L.S. is a recipient of a Burroughs Wellcome Fund New Investigator Award in the Toxicological Sciences. The MALDI-TOF MS instrument was purchased with funds provided by NIH Scientific Instrumentation grant 1-S10-RR13810. We acknowledge the use of instruments in the Keck Biophysics Facility at Northwestern University.

Figure 3. Biophysical properties of metallosalen-DNA Mn4•3 (circles) and Ni-4•3 (triangles) duplexes compared to the control duplex of 3•23 (squares). (A) CD spectra of duplexes (1 µM each strand) in 10 mM sodium phosphate, 150 mM NaCl, pH 7.0. CD spectra were recorded at 25 °C. (B) UV thermal denaturation data were acquired under conditions identical to part A at 260 nm with a temperature ramp rate of 0.5 °C/min.

Supporting Information Available: Synthesis of oligonucleotides 23-26, characterization, CD, and UV thermal denaturation data for Mn-8 and Ni-8, assays of metal ions and templates 3, 7, and 14-17 for the assembly of metallosalen-DNA conjugate 4. This material is available free of charge via the Internet at http:// pubs.acs.org/bc. LITERATURE CITED

by heating the samples from 30 °C to 90 °C at 0.5 °C/ min. The melting profiles of Mn-4•3 and Ni-4•3 and 3•23 are shown in Figure 3B. Metallosalen DNA conjugate duplexes of Mn-4•3 and Ni-4•3 demonstrated sharp melting transitions at 70.6 ( 0.1 °C and 70.5 ( 0.1 °C, respectively. The control duplex comprised of 3•23 displayed a slightly higher melting temperature (Tm) of 75.6 ( 0.3 °C. Thus, the replacement of two base pair interactions with the metallosalen led to a thermal stability loss of 5 °C in these duplexes. A similar decrease was observed in the Ni-8 duplex Tm (∆Tm ) -6 °C) compared with the unmodified duplex of 7•24. Duplex Ni-8•7 displayed a melting transition of 72.1 ( 0.5 °C compared to the Tm of 78.0 ( 0.1 °C for the control duplex 7•24. The Mn-8•7 duplex melted with a Tm of 69.1 ( 0.1 °C, but with a broad curvature, indicating a less cooperative denaturation profile. The UV-thermal denaturation studies indicated that the metallosalen of a metallosalen-DNA conjugate creates a minor destabilization of duplex DNA (∆Tm ) -5 to -6 °C). The melting profiles of the conjugate duplexes were similar to unmodified duplexes 3•23 and 7•24 indicating that insertion of a metallosalen did not alter the mode of duplex denaturation. Although a duplex destabilization was expected with the replacement of two base pair interactions with a metallosalen, the disruption of duplex stability was relatively minor for these 32-nucleotide duplexes. In conclusion, we have demonstrated the generality of template-directed metallosalen-DNA synthesis through the preparation and characterization of a new metallosalen-DNA conjugate (8). A second-generation purification methodology was developed that involved the use of shorter DNA templates, which permitted the isolation and characterization of four metallosalen-DNA conjugates. Our assembly approach for metallosalen-DNA

(1) Saenger, W. (1984) Principles of Nucleic Acid Structure, Springer-Verlag, New York. (2) Zarrinkar, P. P., and Williamson, J. R. (1994) Kinetic Intermediates in RNA Folding. Science 265, 918-924. (3) Feig, A. L., and Uhlenbeck, O. C. (1999) in The RNA World (Gesteland, R. F., Cech, T. R., and Atkins, J. F., Eds.) pp 287319, CSHL Press, Cold Spring Harbor, NY. (4) Pyle, A. M. (1993) Ribozymes: A Distinct Class of Metalloenzymes. Science 261, 709-714. (5) Dreyer, G. B., and Dervan, P. B. (1985) Sequence-specific cleavage of single-stranded DNA: oligodeoxynucleotide-EDTAFe(II). Proc. Natl. Acad. Sci. U.S.A. 82, 968-972. (6) Bashkin, J. K., Frolova, E. I., and Sampath, U. (1994) Sequence-Specific Cleavage of HIV mRNA by a Ribozyme Mimic. J. Am. Chem. Soc. 116, 5981-5982. (7) Bergstrom, D. E., and Gerry, N. P. (1994) Precision SequenceSpecific Cleavage of a Nucleic Acid by a Minor-GrooveDirected Metal-Binding Ligand Linked through N-2 of Deoxyguanosine. J. Am. Chem. Soc. 116, 12067-12068. (8) Chen, C. B., and Sigman, D. S. (1988) Sequence-specific scission of RNA by 1,10-phenanthroline-copper linked to deoxyoligonucleotides. J. Am. Chem. Soc. 110, 6570-6572. (9) Sakamoto, S., Tamura, T., Furukawa, T., Komatsu, T., Ohtsuka, E., Kitamura, M., and Inoue, H. (2003) Highly efficient catalytic RNA cleavage by the cooperative action of two Cu(II) complexes embodied within an antisense oligonucleotide. Nucleic Acids Res. 31, 1416-1425. (10) Hall, J., Husken, D., and Haner, R. (1997) Sequencespecific cleavage of RNA using macrocyclic lanthanide complexes conjugated to oligonucleotides: a structure activity study. Nucleosides Nucleotides 16, 1357-1368. (11) Dubey, I., Pratviel, G., and Meunier, B. (2000) Synthesis and DNA cleavage of 2′-O-amino-linked metalloporphyrinoligonucleotide conjugates. J. Chem. Soc., Perkin Trans. 1 3088-3095. (12) Magda, D., Crofts, S., Lin, A., Miles, D., Wright, M., and Sessler, J. L. (1997) Synthesis and Kinetic Properties of Ribozyme Analogs Prepared Using Phosphoramidite De-

176 Bioconjugate Chem., Vol. 16, No. 1, 2005 rivatives of Dysprosium(III) Texaphyrin. J. Am. Chem. Soc. 119, 2293-2294. (13) Lewis, F. D., Helvoigt, S. A., and Letsinger, R. L. (1999) Synthesis and spectroscopy of Ru(II)-bridged DNA hairpins. Chem. Commun. 327-328. (14) Khan, S. I., Beilstein, A. E., Tierney, M. T., Sykora, M., and Grinstaff, M. W. (1999) Solid-Phase Synthesis and Photophysical Properties of DNA Labeled at the Nucleobase with Ru(bpy)2(4-m-4′-pa-bpy)2+. Inorg. Chem. 38, 5999-6002. (15) Bannwarth, W., Pfleiderer, W., and Mueller, F. (1991) Energy transfer within oligonucleotides from a lumazine (pteridine-2,4(1H,3H)-dione) chromophore to bathophenanthroline-ruthenium(II) complexes. Helv. Chim. Acta 74, 19911999. (16) Yoo, J., Delaney, S., Stemp, E. D. A., and Barton, J. K. (2003) Rapid radical formation by DNA charge transport through sequences lacking intervening guanines. J. Am. Chem. Soc. 125, 6640-6641. (17) Ortmans, I., Content, S., Boutonnet, N., Kirsch-De Mesmaeker, A., Bannwarth, W., Constant, J. F., Defrancq, E., and Lhomme, J. (1999) Ru-labeled oligonucleotides for photoinduced reactions on targeted DNA guanines. Chem. Eur. J. 5, 2712-2721. (18) Meade, T. J., and Kayyem, J. F. (1995) Electron transfer through DNA: site-specific modification of duplex DNA with ruthenium donors and acceptors. Angew. Chem., Int. Ed. 34, 352-354. (19) Meggers, E., Kusch, D., and Giese, B. (1997) An efficient synthesis of enantiomerically pure D- and L-ruthenium(II)labeled oligonucleotides. Helv. Chim. Acta 80, 640-652. (20) Telser, J., Cruickshank, K. A., Schanze, K. S., and Netzel, T. L. (1989) DNA oligomers and duplexes containing a covalently attached derivative of tris(2,2′-bipyridine)ruthenium(II): synthesis and characterization by thermodynamic and optical spectroscopic measurements. J. Am. Chem. Soc. 111, 7221-7226. (21) Schubert, F., Kluetsch, T., and Cech, D. (1997) Anionic manganese meso-tetraphenyl porphine/oligonucleotide conjugates - synthesis and utilization in chemiluminescent DNA-probe detection. Nucleosides Nucleotides 16, 277-289. (22) Bannwarth, W., Schmidt, D., Stallard, R. L., Hornung, C., Knorr, R., and Mueller, F. (1988) Bathophenanthrolineruthenium(II) complexes as non-radioactive labels for oligonucleotides which can be measured by time-resolved fluorescence techniques. Helv. Chim. Acta 71, 2085-2099. (23) Hurley, D. J., and Tor, Y. (1998) Metal-Containing Oligonucleotides: Solid-Phase Synthesis and Luminescence Properties. J. Am. Chem. Soc. 120, 2194-2195. (24) Manchanda, R., Dunham, S. U., and Lippard, S. (1996) Automated Solid-Phase Synthesis of Site-Specifically Platinated Oligodeoxyribonucleotides. J. Am. Chem. Soc. 118, 5144-5145. (25) Zimmermann, N., Meggers, E., and Schultz, P. G. (2002) A Novel Silver(I)-Mediated DNA Base Pair. J. Am. Chem. Soc. 124, 13684-13685. (26) Weizman, H., and Tor, Y. (2001) 2,2′-Bipyridine Ligandoside: A Novel Building Block for Modifying DNA with IntraDuplex Metal Complexes. J. Am. Chem. Soc. 123, 3375-3376. (27) Meggers, E., Holland, P. L., Tolman, W. B., Romesberg, F. E., and Schultz, P. G. (2000) A Novel Copper-Mediated DNA Base Pair. J. Am. Chem. Soc. 122, 10714-10715. (28) Tanaka, K., Yamada, Y., and Shionoya, M. (2002) Formation of Silver(I)-Mediated DNA Duplex and Triplex through an Alternative Base Pair of Pyridine Nucleobases. J. Am. Chem. Soc. 124, 8802-8803. (29) Tanaka, K., Tengeiji, A., Kato, T., Toyama, N., and Shionoya, M. (2003) A Discrete Self-Assembled Metal Array in Artificial DNA. Science 299, 1212-1213. (30) Weizman, H., and Tor, Y. (2001) Oligo-ligandosides: a DNA mimetic approach to helicate formation. Chem. Commun. 453-454. (31) Wagenknecht, H.-A. (2003) Metal-mediated DNA base pairing and metal arrays in artificial DNA: Towards new nanodevices. Angew. Chem., Int. Ed. 42, 3204-3206.

Czlapinski and Sheppard (32) Weizman, H., and Tor, Y. (2002) Redox-active metalcontaining nucleotides: Synthesis, tunability, and enzymatic incorporation into DNA. J. Am. Chem. Soc. 124, 1568-1569. (33) Goodwin, J. T., and Lynn, D. G. (1992) Template-directed synthesis: use of a reversible reaction. J. Am. Chem. Soc. 114, 9197-9198. (34) Gryaznov, S. M., and Letsinger, R. L. (1993) Chemical ligation of oligonucleotides in the presence and absence of a template. J. Am. Chem. Soc. 115, 3808-3809. (35) Fedorova, O. A., Gottikh, M. B., Oretskaya, T. S., and Shabarova, Z. A. (1996) Cyanogen bromide-induced chemical ligation: mechanism and optimization of the reaction conditions. Nucleosides Nucleotides 15, 1137-1147. (36) Xu, Y., Karalkar, N. B., and Kool, E. T. (2001) Nonenzymatic autoligation in direct three-color detection of RNA and DNA point mutations. Nat. Biotechnol. 19, 148-152. (37) Sando, S., Abe, H., and Kool, E. T. (2004) Quenched AutoLigatind DNAs: Multicolor Identification of Nucleic Acids at Single Nucleotide Resolution. J. Am. Chem. Soc. 126, 10811087. (38) Zhan, Z.-Y. J., and Lynn, D. G. (1997) Chemical Amplification through Template-Directed Synthesis. J. Am. Chem. Soc. 119, 12420-12421. (39) Li, X., Zhan, Z.-Y. J., Knipe, R., and Lynn, D. G. (2002) DNA-Catalyzed Polymerization. J. Am. Chem. Soc. 124, 746747. (40) Summerer, D., and Marx, A. (2002) Highlight: DNAtemplated synthesis: More versatile than expected. Angew. Chem., Int. Ed. 41, 89-90. (41) Wu, T., and Orgel, L. E. (1992) Nonenzymatic templatedirected synthesis on oligodeoxycytidylate sequences in hairpin oligonucleotides. J. Am. Chem. Soc. 114, 317-322. (42) Bruick, R. K., Dawson, P. E., Kent, S. B. H., Usman, N., and Joyce, G. F. (1996) Template-directed ligation of peptides to oligonucleotides. Chem. Biol. 3, 49-56. (43) Letsinger, R. L., Wu, T., and Elghanian, R. (1997) Chemical and photochemical ligation of oligonucleotide blocks. Nucleosides Nucleotides 16, 643-652. (44) Gryaznov, S. M., Schultz, R., Chaturvedi, S. K., and Letsinger, R. L. (1994) Enhancement of selectivity in recognition of nucleic acids via chemical autoligation. Nucleic Acids Res. 22, 2366-2369. (45) Liu, J., and Taylor, J. S. (1998) Template-directed photoligation of oligodeoxyribonucleotides via 4-thiothymidine. Nucleic Acids Res. 26, 3300-3304. (46) Luther, A., Brandsch, R., and von Kiedrowski, G. (1998) Surface-promoted replication and exponential amplification of DNA analogs. Nature 396, 245-248. (47) Kozlov, I. A., De Bouvere, B., Van Aerschot, A., Herdewijn, P., and Orgel, L. E. (1999) Efficient Transfer of Information from Hexitol Nucleic Acids to RNA during Nonenzymatic Oligomerization. J. Am. Chem. Soc. 121, 5856-5859. (48) Mattes, A., and Seitz, O. (2001) Mass-spectrometric monitoring of a PNA-based ligation reaction for the multiplex detection of DNA single-nucleotide polymorphisms. Angew. Chem., Int. Ed. 40, 3178-3181. (49) Mattes, A., and Seitz, O. (2001) Sequence fidelity of a template-directed PNA-ligation reaction. Chem. Commun. 2050-2051. (50) Gartner, Z. J., Grubina, R., Calderone, C. T., and Liu, D. R. (2003) Two enabling architectures for DNA-templated organic synthesis. Angew. Chem., Int. Ed. 42, 13701375. (51) Gartner, Z. J., and Liu, D. R. (2001) The Generality of DNA-Templated Synthesis as a Basis for Evolving NonNatural Small Molecules. J. Am. Chem. Soc. 123, 69616963. (52) Calderone, C. T., Puckett, J. W., Gartner, Z. J., and Liu, D. R. (2002) Directing otherwise incompatible reactions in a single solution by using DNA-templated organic synthesis. Angew. Chem., Int. Ed. 41, 4104-4108. (53) Li, X., and Liu, D. R. (2003) Stereoselectivity in DNATemplated Organic Synthesis and Its Origins. J. Am. Chem. Soc. 125, 10188-10189.

Bioconjugate Chem., Vol. 16, No. 1, 2005 177

Template-Directed Metallosalen−DNA Assembly (54) Gartner, Z. J., Kanan, M. W., and Liu, D. R. (2002) Multistep small-molecule synthesis programmed by DNA templates. J. Am. Chem. Soc. 124, 10304-10306. (55) Czlapinski, J. L., and Sheppard, T. L. (2001) Nucleic Acid Template-Directed Assembly of Metallosalen-DNA Conjugates. J. Am. Chem. Soc. 123, 8618-8619. (56) Czlapinski, J. L., and Sheppard, T. L. (2004) TemplateDirected Assembly of Metallosalen-DNA Hairpin Conjugates. ChemBioChem 5, 127-129. (57) Pourbaix, M. (1966) Atlas of Electrochemical Equilibria in Aqueous Solutions, Pergamon Press, Oxford. (58) Marky, L. A., Kallenbach, N. R., McDonough, K. A., Seeman, N. C., and Breslauer, K. J. (1987) The melting behavior of a DNA junction structure: a colorimetric and spectroscopic study. Biopolymers 26, 1621-1634. (59) Katsuki, T. (2003) Some recent advances in metallosalen chemistry Synlett 281-297.

(60) Mandelkern, M., Dattagupta, N., and Crothers, D. M. (1981) Conversion of B DNA between solution and fiber conformations. Proc. Natl. Acad. Sci. U.S.A. 78, 42944298. (61) Brown, R. S., Nielsen, M., and Gothelf, K. V. (2004) Selfassembly of aluminium-salen coupled nanostructures from encoded modules with cleavable disulfide DNA-linkers. Chem. Commun. 1464-1465. (62) Czlapinski, J. L., and Sheppard, T. L. (2004) Site-Specific Oxidative Cleavage of DNA by Metallosalen-DNA Conjugates. Chem. Commun. 2468-2469. (63) Gray, D. M., Ratliff, R. L., and Vaughan, M. R. (1992) Circular dichroism spectroscopy of DNA. Methods Enzymol. 211, 389-406.

BC049753R