ABSTRACT
The divergent synthesis of branched DNA (bDNA) comb structures is described. This new type of bDNA contains one unique oligonucleotide, the primary sequence, covalently attached through a comb-like branch network to many identical copies of a different oligonucleotide, the secondary sequence. The bDNA comb structures were assembled on a solid support and several synthesis parameters were investigated and optimized. The bDNA comb molecules were characterized by polyacrylamide gel electrophoretic methods and by controlled cleavage at periodate-cleavable moieties incorporated during synthesis. The developed chemistry allows synthesis of bDNA comb molecules containing multiple secondary sequences. In the accompanying article we describe the synthesis and characterization of large bDNA combs containing all four deoxynucleotides for use as signal amplifiers in nucleic acid quantification assays.
Nucleic acids are key molecular targets for human diagnostic tests and microorganisms can now be detected at low levels in clinical specimens (1 ). Target amplification methods, such as the polymerase chain reaction (PCR) (2 ) and ligase chain reaction (LCR) (3 ), have been introduced. These methods are employed to produce large quantities of the nucleic acid of interest, which can then be detected with conventional detection schemes. In contrast, we have utilized a direct analysis of nucleic acids which involves amplification of the signal, rather than replication of the target, to detect the nucleic acid of interest. The key molecule in the signal amplification method is a branched DNA (bDNA) molecule which allows specific incorporation of many labels (4 ). We have been interested in producing large bDNA molecules to act as amplification multimers.
Various approaches have been used to construct branched nucleic acid polymer structures. Hudson et al. used a convergent growth procedure to prepare a structure containing 87 nt with six adenosine branch point units and 12 terminal ends (5 ,6 ). Divergent solid phase phosphoramidite strategies have been used to assemble smaller natural lariat molecules in which three oligomers were connected through a natural ribonucleotide branch point (7 ,8 ). However, due to low coupling yields, these convergent and divergent synthesis approaches are not readily applicable to high yield synthesis of large branched nucleic acids. Other groups have employed solid phase synthesis using 3'-deoxy-[beta]-D-psicothymidine (9 ), 4'-C-(hydroxymethyl)thymidine (10 ) and 1-(2-methyl-[beta]-D-arabinofyranosyl)uracil (11 ) as branch point monomers to prepare small branched DNA and/or RNA oligonucleotides respectively. In earlier work our group has described the chemical synthesis of bDNA molecules of the `fork' and `comb' type containing the same primary and secondary sequences (a total of 107 nt) (12 ). The comb structure appeared to offer a more open structure permitting better control over the number of and spacing between secondary hybridization sites. Thus strategies to improve synthesis of the bDNA `comb' structures were pursued.
In this communication we detail the development of an optimized chemistry for the synthesis of bDNA comb structures by a solid phase method and describe several synthetic parameters that were investigated to achieve higher quality and yield of product. We also present a novel chemical degradation technique that utilized a periodate-cleavable moiety incorporated in its phosphoramidite form during bDNA oligomer synthesis. We found that specific cleavage of the product bDNA provided information regarding the quality of the secondary sequence and have used this approach to determine the optimal length of the secondary sequence. Here we focus on the synthesis of large bDNA comb molecules containing as many as 400 nt. In the accompanying paper (13 ) we describe the synthesis and characterization of large bDNA combs containing all four deoxynucleotides.
All chemicals and biochemicals were reagent grade or better from commercial sources and were used without further purification. Nucleoside and nucleotide standards were purchased from Sigma (St Louis, MO) or Pharmacia (Piscataway, NJ). The 3'-O-(2-cyanoethyl)-N,N-diisopropylphosphoramidites of dAN-Bz, dCN-Bz, dGN-ibu and T were purchased from Glen Research (Sterling, VA). Anhydrous acetonitrile (<30 p.p.m. water content) was from Baxter or Fisher. All other DNA synthesis ancillary reagents were purchased from Applied Biosystems (a Division of Perkin Elmer, Foster City, CA). The SDT reagent was prepared by dissolving 1 g solid sodium dithionite (technical grade, 80% purity) in 20 ml 1 M triethylammonium bicarbonate, pH 7.2, followed by addition of 20 ml dioxane (14 ). Underivatized controlled pore glass (CPG) with pore sizes of 500, 1000, 1400, 2000 and 3000 Å, all with particle size 120/200, were obtained from CPG Inc. (Piscataway, NJ). Ion spray mass spectrometry (ESI) measurements were run on a Perkin-Elmer PE SCIEX API III electrospray quadrupole instrument. NMR spectra were recorded on a Varian 300 MHz instrument. The 31P NMR spectra were run in CH3CN with d6-DMSO for a lock and aqueous 85% H3PO4 as an external reference. Elemental analyses were made by the Microanalytical Laboratory, University of California (Berkeley, CA).
Polyacrylamide gel electrophoresis (PAGE) was carried out using 10 or 20% cross-linked slab gels (1 or 3 mm thick with 19:1 acrylamide:bis-acrylamide) with the following running buffer: 100 mM Tris-borate, 1 mM EDTA, 7 M urea, pH 8.3 (15 ) (diluted from a 10× stock). Bromophenol blue (BPB) was used as tracking dye. Gels were transferred to a thin layer chromatography plate containing a F254 indicator (20 × 20 cm; EM Sciences) and DNA bands were visualized under UV light.
Ninhydrin test. Ninhydrin (100 mg) was dissolved in 100 ml ethanol. In a test tube, 1 ml ninhydrin solution was added to a 100 mg sample of CPG and then heated to a quick boil. A positive sample turned dark blue under these conditions.
To a solution of N-4-(6-hydroxyhexyl)-5'-DMT-5-methyl- 2'-deoxycytidine in 200 CH2Cl2 ml (17 mmol) (12 ) was added pyridine (40 mmol). The mixture was cooled to 0°C. Then a solution of MAC-Cl (20 mmol) in 200 ml CH2Cl2 was added dropwise and left stirring for 10 min. Thin layer chromatography analysis showed that the starting material had been completely consumed. The reaction mixture was diluted with 400 ml ethyl acetate and the organic phase extracted twice with 300 ml 5% NaHCO3 followed by 80% saturated aqueous NaCl. After the organic phase was dried over Na2SO4 for 30 min and filtered the solvent was removed in vacuo. The product, DMT-BM(MAC), was purified by silica gel chromatography using a gradient of methanol (0-6%) in CH2Cl2 to give 13 g pure product (85% yield). 1H NMR (CDCl3): [delta] 1.4 (m, 4H), 2.4 (s, 3H), 1.6-1.7 (m, 4H), 3.4 (m, 2H), 3.5 (m, 4H), 3.8 (s, 6H), 4.2 (s, 2H), 4.6 (m, 1H), 4.8 (m, 1H), 6.4 (t, 1H), 6.8 (d, 4H), 7.2-7.4 (m, 9H), 7.6 (s, 1H), 7.8 (m, 3H), 8.3 (m, 4H) p.p.m. A sample was detritylated and purified by silica gel chromatography to yield BM(MAC) for analysis. ESI MS, mol. wt calculated for C32H35N3O9 605.6; found 606.2. 1H-NMR (CD3-OD): [delta] 1.4 (m, 4H: -CH2-CH2-), 1.6 (m, 2H: -CH2-C-N), 1.7 (m, 2H: -CH2-C-O), 1.9 (s, 3H: C-5-CH3), 3.4 (t, 2H: -CH2-NH-), 3.7-3.8 (2H, q: 2' H2), 3.9 (t, 1H: 4' H), 4.2 (d, 2H: CH2-OH), 4.4 (t, 1H: 3' H), 4.8 (s, 2H: Ph-CH2-O-CO-O), 6.2 (t, 1H: 1' H), 7.7 (s, 1H: C-6 H), 7.8 (m, 3H: aromatic H), 8.2 (d, 4H: aromatic H) p.p.m.
The nucleoside N-4-(O-2-anthraquinonemethoxycarbonyl-6-oxyhexyl)-5'-DMT-5-methyl-2'-deoxycytidine (14.4 mmol) was converted to the ME phosphoramidite using a published phosphitylation procedure (12 ) to give 12.4 g (80% yield) of a slightly yellow solid. 31P NMR: 146.5, 145.8 p.p.m. Other carbonate-protected derivatives of DMT-BM were prepared in a similar fashion.
Synthesis of N-4-(O-N,N-diisopropylamino-2-cyanoethoxyphosphinyl-6-oxyhexyl)-5'-DMT-2',3'-dibenzoylcytidine[DMT-CM1 (Bz2) BCE]. Uridine (24.5 g, 100 mmol) dissolved in pyridine was treated with DMT-Cl (50 mmol) for 3 h. The solvents were removed in vacuo and the residue in 500 ml ethyl acetate was washed with 2 × 500 ml 5% aqueous NaHCO3 and 500 ml 80% saturated aqueous NaCl solution. After drying over solid Na2SO4 and filtration the solvent was removed in vacuo. Crude 5'-DMT-uridine (28 g) was then treated directly with benzoyl chloride (300 mmol) in pyridine. After aqueous work-up as described above and silica gel chromatography [using a gradient of methanol (0-6%) in CH2Cl2] 5'-DMT-2',3'-dibenzoyluridine was isolated in 85% overall yield. This intermediate was converted to 4-(triazolo)-5'-DMT-2',3'-dibenzoyl-ribofuranosyl-2(1H)-pyrimidone, isolated as a white foam in quantitative yield, using the published procedure (12 ). To a stirred solution of this compound (85 mmol) in 250 ml acetonitrile was added 6-aminohexanol (11.9 g, 101.5 mmol) as a solid and stirring was continued for 18 h. The reaction mixture was diluted with 700 ml ethyl acetate, extracted and dried as described above. The product was purified by silica gel chromatography eluted with a 0-50% gradient of ethyl acetate in methylene chloride to give 35.4 g (41.5 mmol) N-4-(6-hydroxyhexyl)-5'-DMT-2',3'-dibenzoylcytidine (DMT-CM1(Bz2). ESI MS, mol. wt calculated for C50H51N3O10 854.0, found 853.4. 1H NMR (CDCl3): [delta] 1.35 (m, 4H), 1.5 (m, 4H), 3.4 (m, 2H), 3.6 (m, 4H), 3.8 (s, 6H), 4.4 (m, 1H), 5.8 (m, 1H), 5.9 (m, 1H), 6.7 (m, 1H), 6.8 (d, 4H), 7.2-7.6 (m, 15H), 7.7 (d, 1H), 7.9 (m, 4H) p.p.m. Analysis calculated for C50H51N3O10, C 70.32, H 6.02, N 4.92; found C 69.04, H 6.13, N 4.78. For further characterization a small sample was fully deprotected with trichloroacetic acid and ammonium hydroxide before itwas purified by silica gel chromatography to give the parent compound CM1 = N-4-(6-hydroxyhexyl)-cytidine. 1H-NMR (CD3OD): [delta] 1.4 (m, 4H: -CH2-CH2-), 1.6 (m, 4H: -CH2-C-N/-CH2-C-O), 1.9 (s, 3H: C-5-CH3), 3.4 (t, 2H: -CH2-NH-), 3.55 (t, 2H: C-CH2-OH), 3.7-3.8 (2H, q: 2' H2), 3.9 (t, 1H: 4' H), 4.15 (t, 2H: CH2-OH), 3.75 (t, 1H: 4' H), 3.85 (t, 1H: 3' H), 4.0 (t, 1H: 2' H), 5.8 (t, 1H: 1' H), 7.9 (s, 1H: C-6 H) p.p.m.
A portion of purified DMT-CM1(Bz2) (8.9 mmol) was converted to the BCE phosphoramidite using a published phosphitylating procedure (12 ) to give N-4-(O-N,N-diisopropylamino-2-cyanoethoxyphosphinyl-6-oxyhexyl)-5'-DMT-2',3'-dibenzoylcytidine (7.25 g, 6.9 mmol, 78%). 31P-NMR: 148.0 p.p.m.
Periodate cleavage of bDNA. Purified bDNA (1-3 A260 units) containing 1,2-diol cleavable monomers (CM1) were treated with freshly prepared aqueous sodium periodate solution at a concentration of 25 mM in a total volume of 12 µl for 60 min at room temperature. Excess periodate was reacted with 4 µl aqueous 1 M glycerol solution. After addition of n-propylamine (5 µl) and 0.1 M triethylammonium acetate (25 µl) the mixture was heated at 60°C for 90 min and evaporated to dryness. The residue was dissolved in 10 µl water and analyzed by PAGE.
Preparation of CPG-propyl-NH-CO-CH2CH2-CO-NH-(CH2)6- NH2 (CPG-HDA). CPG-propyl-NH2 (44 g) in a round-bottom flask were suspended in 350 ml THF containing succinic anhydride (16.3 g) and N,N-dimethylaminopyridine (DMAP) (2.2 g) and shaken for 18 h. The CPG was collected, washed with 5 × 400 ml methanol and 5 × 250 ml ether and air dried. Negative ninhydrin and Fluram tests confirmed the absence of amino groups. The resulting CPG-propyl-NH-CO-CH2CH2-COOH was suspended in 265 ml N,N-dimethylformamide (DMF) containing 1,1-carbonyldiimidazole (17.6 g) and left in vacuo for 18 h. The DMF solution was decanted off and the CPG washed with 2 × 250 ml DMF which was decanted off. The activated CPG-(CH2)3-NH-CO-CH2CH2-CO-imidazolide was suspended in 265 ml DMF containing 1,6-hexanediamine (17.6 g) and left shaking for 18 h. The resulting CPG-propyl-NH-CO-CH2CH2-CO-NH-(CH2)6-NH2 was washed extensively in a fritted funnel with 5 × 400 ml methanol, 5 × 400 ml methylene chloride and 5 × 400 ml ether. The final CPG was thoroughly dried in an incubator at 60°C for 18 h. Positive ninhydrin and Fluram tests confirmed the presence of amino groups.
Capping of residual silanol sites. CPG-HDA (44 g) was washed with 2 × 200 ml pyridine and then suspended in a mixture of trimethylchlorosilane (55 ml) and hexamethyldisilazane (110 ml). The suspension was gently shaken for 18 h. The capped CPG-HDA was washed extensively in a fritted funnel with 3 × 250 ml pyridine, 3 × 250 ml methanol and 3 × 400 ml ether and air dried.
General procedure for preparation of CPG-HDA-cytidine supports. 5'-DMT-CN-Bz-3'-O-hemisuccinate (1.16 g, 1.6 mmol) (16 ) was activated with p-nitrophenol (0.44 g) and dicyclohexylcarbodiimide (3.3 g, 16 mmol) to give 5'-DMT-CN-Bz-3'-O-CO-CH2CH2-CO-(p-nitrophenyl). After filtration the activated nucleoside succinate was added to CPG2000-HDA (8 g) and the suspension shaken for 7 h. To cap remaining amino groups the functionalized CPG was treated with a solution of pyridine (115 ml) containing acetic anhydride (12.8 ml) and DMAP (0.05 g). Loading 12 µmol/g. A similar loading was obtained starting from CPG2000-propyl-NH2 (8 g).
We have previously reported the synthesis of `fork' and `comb' bDNA molecules (12 ,17 ). Model building suggested that `comb' bDNA molecules would provide better access to both primary and secondary sequences and result in more favorable hybridization properties. Larger bDNA comb oligomers with an increasing number of secondary sequences can be synthesized by adding more branch point monomers during linear synthesis. The chemical synthesis of comb bDNA is outlined in Scheme 1 A. Scheme 1 B shows a graphical notation of a bDNA comb oligomer with the primary (Nx), spacing (Nss) and secondary (Nz) sequences and spacing nucleotides (Ny = branch point monomer to branch point monomer spacing). In the following text the notation BM<T5> is used to indicate that the nucleotide sequence 5'-TTTTT-3' is attached to the sidechain hydroxyl group of a particular branch point monomer (BM).
To be fully compatible with standard DNA synthesis a branch point monomer could, in principle, be derived from any compound with at least three hydroxyl functions, for example inexpensive and commercially available triol compounds like 1,1,1-tris(hydroxymethyl)ethane (1) and 1,3,5-tris(2-hydroxyethyl)cyanuric acid (2). However, the conversion to fully protected phosphoramidite reagents of these compounds was difficult due to the lack of selectivity in the introduction of dimethoxytrityl (DMT) and levulinate (LEV) protecting groups and the syntheses were difficult to run on larger scales. Small amounts of pure DMT-1(LEV) BCE and DMT-2(LEV) BCE were produced for testing in bDNA synthesis. Branch point monomer 1 performed favorably and was used to synthesize functional bDNA comb molecules. Branch point monomer 2 was found to be unstable under standard deprotection conditions, precluding its general use (data not shown).
We have chosen a more practical approach to the synthesis of new branch point monomers from nucleosides. A nucleoside preserves the natural internucleotidic distance and it can be modified to introduce a third hydroxy function linked to the nucleotide through a suitable spacer. We have developed a branch point monomer from a modified nucleoside, N-4-(6-hydroxyhexyl)-2'-deoxycytidine (BM), which was prepared in high yield from thymidine (12 ,18 ) with enhanced stability towards transamination and deamination reactions (19 ) due to the 5-methyl group. We found that this general approach in which the starting diol, thymidine, is protected then transformed into a partially protected triol greatly facilitates introduction of the desired protecting groups on all three hydroxyl functions, primary, secondary and sidechain hydroxyl groups, in the BM molecule. Removal of 3'-O-t-butyldimethylsilyl (TBDMS) and phosphitylation yielded the phosphoramidite reagent.
The synthesis of bDNA oligomers, as outlined in Scheme 1 , required an orthogonal protecting group for the sidechain hydroxyl that would be stable to all the commonly used reagents for synthesis of DNA oligonucleotides, but which could be removed selectively without affecting any other protecting groups or functionality in the oligonucleotide, including linkage between the oligomer and the solid support.
Several hydroxy protecting groups were investigated. Branch point monomer DMT-BM(R3), in either ME or BCE phosphoramidite form, was prepared with the following sidechain hydroxy protecting groups: R3 = LEV (see below), MAC (see below), NPEOC (p-nitrophenylethoxycarbonyl; 20 ); PTEOC (phenylthioethoxycarbonyl; 21 ); BNPEOC [2,2-bis(p-nitrophenyl)ethoxycarbonyl; 22 ]; FMOC (fluorenyloxycarbonyl; 23 ); TBDMS (24 ); TMSEOC [2-(trimethylsilyl)ethoxycarbonyl; 25 ]. Only LEV and MAC were found to be practical for bDNA synthesis work.
The LEV group was easy to introduce into bis-protected BM using levulinic anhydride (12 ). It could be removed rapidly and specifically with HPAA (4:1) reagent [0.5 M hydrazine hydrate in pyridine/acetic acid (4:1 v/v)] (26 ,27 ) without cleavage of the nucleoside-3'-O-succinate linkage to the support (28 ,29 ). The nucleobases were not modified when HPAA (1:1) reagent was properly buffered (28 ); this is especially important for cytidine residues, which are known to be particularly sensitive to transamination reactions (19 ). Synthesis of bDNA comb molecules with a multitude of secondary sequences required selective and complete removal of all LEV groups. The reported deprotecting procedure (26 ) was modified to accelerate the process while minimizing debenzoylation and base modification by increasing the acidic acid content from 20 to 50% in the solvent mixture. The accelerated de-levulinylation of BM(LEV) is probably due to facile dehydration of the initial adduct between hydrazine and the keto moiety to form the hydrazone under more acidic conditions. Complete removal of the LEV protecting group from the BMs in the solid supported linear sequence was achieved with HPAA (1:1) reagent in 90 min at room temperature (see below and Fig. 1 ). Both ME and BCE phosphate protecting groups were stable to all reagents required for bDNA synthesis using LEV and HPAA deprotection.
We also developed a new hydroxyl protecting group, MAC, the oxycarbonyl derivative of the 2-oxymethylene-9,10-anthraquinone (MAQ) moiety, as a complement to LEV(30 ). The 2-oxymethylene-9,10-anthraquinone moiety has been utilized for protection of carboxylic acids as the MAQ ester (14 ), amines as the MAQ urethane (31 ) and phosphates as the MAQ phosphotriester (32 ). However, use of the MAC group for hydroxyl protection has not previously been reported. The MAC protecting group was introduced specifically on to the sidechain primary hydroxy function of 5'-DMT-BM(OH)-OH without 3'-O protection using the corresponding chloroformate derivatives (33 ). Specific deprotection of MAC via elimination of 2-methylanthraquinone (14 ) was effected with an aqueous solution of sodium dithionite (SDT reagent) under neutral and mild reductive elimination conditions. The MAC group was found to be orthogonal to the protecting groups used in chemical synthesis of bDNA and complementary to LEV. Thus MAC was found to be stable to 3% trichloroacetic acid in CH2Cl2, HPAA (1:1), tetrazole, iodine/water oxidation and acetic anhydride capping conditions. The MAC was found to be less stable under basic conditions. The SDT reagent did not adversely affect succinate, N-acyl, N-(dibutylformamidine), LEV, TBDMS, DMT and methyl and 2-cyanoethyl phosphotriesters and modification of the DNA bases was not observed. In bDNA synthesis complete removal of the MAC protecting group from the BMs in the solid supported primary sequence was achieved with the SDT reagent in 30 min at room temperature. The mutual stability of BM(MAC) to HPAA and of BM(LEV) to SDT allowed us to deprotect one of the groups without affecting the other in the same DNA molecule. By incorporating the BM(LEV) and the BM(MAC) branch point monomers during primary synthesis, followed by differential deprotection and chain elongation from each type of branch point monomer, it was possible to synthesize bDNA with two different secondary sequences (data not shown).
We optimized the bDNA synthesis using the chemistry developed for BM protected with LEV. Several parameters were investigated to determine their effect on the quality of bDNA product, including: (i) LEV deprotection conditions; (ii) pore size of solid supports; (iii) length of spacing sequence.
The deprotection time for the LEV protecting group was optimized, since removal of multiple copies of identical protecting groups in a single DNA molecule could not be accomplished in the same time frame as for removal of a single copy of the same protecting group in a DNA molecule. As shown in Figure 1 , extended HPAA (1:1) reagent treatment appeared to eliminate most of the LEV groups when test molecule 5'-BM8<T5>8-T20-3' was exposed to the HPAA (1:1) reagent for 5, 15 and 70 min respectively on the solid support prior to secondary synthesis. Near quantitative LEV deprotection was achieved with a 70 min treatment, as indicated by a single product band with the lowest mobility (Fig. 1 , lane 4). For synthesis of bDNA with more BMs we routinely used a 90 min HPAA (1:1) deprotection step to insure complete removal of all LEV groups.
Steric hindrance inside the pores of the polymeric support has been shown to impose some limitations on the synthesis of DNA molecules (34 ). We found that the pore size of the CPG solid support had a profound effect on synthesis of bDNA comb oligomers. As a comparison we synthesized the bDNA comb oligomer 5'-T5-BM8<T5>8-T20-3' on 1000, 2000 and 3000 Å CPG supports using a 15 min HPAA (1:1) reagent deprotection step to remove the LEV groups. After standard ammonium hydroxide deprotection the crude mixtures were analyzed by PAGE (Fig. 2 ). Lanes 1 and 4 show the material from the 1000 Å support, whereas lane 2 shows the materials from the 3000 Å support and lane 3 shows the material that was synthesized on the 2000 Å support. The smearing pattern observed with the 1000 Å support resulted from insufficient elongation of secondary sequences. The discrete bands observed in syntheses using both 2000 and 3000 Å supports were due to incomplete LEV deprotection (see above). Increasing the pore size of the CPG solid support greatly improved the quality of the synthesis and both 3000 and 2000 Å CPG (lanes 2 and 3 respectively) were superior to the standard 1000 Å CPG (lanes 1 and 4). Since CPG with 3000 Å pores was too fragile for practical use, we have chosen 2000 Å CPG for routine bDNA synthesis. Further, it was found that we could successfully synthesize 5'-T5-BM8<T10>8-T20-3', which incorporated T10 secondary sequences using a 2000 Å CPG support (lane 5).
We have investigated the chemical synthesis of bDNA comb molecules using a flexible divergent approach. We have developed N-4-(6-hydroxyhexyl)-2'-deoxycytidine (the BM nucleoside) as a branch point monomer and found LEV to be the preferred sidechain hydroxyl protecting group. Improved synthesis protocols were developed to achieve efficient solid phase synthesis of large bDNA comb molecules with >30 branches. The protocols have been utilized in synthesis of bDNA comb molecules containing all four deoxynucleotides. This work is described in detail in the next paper of this issue of Nucleic Acids Research (13 ).
We wish to thank Yougen Gee for expert technical support in the preparation of bDNA synthesis reagents, Drs Celine Hu and Say-Jong Law for helpful suggestions during manuscript preparation, Dr F.Masiarz for ion spray MS measurements and Dr Linda Wuestehube for editorial assistance.
Fluram test. Fluram (fluorescamine; 15 mg) was dissolved in 100 ml acetone. In a screw cap vial 1 ml Fluram solution was added to a 100 mg sample of CPG to be tested. After exposure for 5 min the CPG was washed three times with acetone and dried. The vial was placed on top of a thin layer chromatography plate which contained F254 indicator and viewed under long wavelength UV light (350 nm). A positive test was indicated by strong fluorescence.
Preparation of 5'-DMT-N-4-(O-2-anthraquinonemethoxycarbonyl-oxyhexyl)-5-methyl-2'-deoxycytidine-3'-O-N,N-diisopropylmethylphosphoramidite [5'-DMT-BM(MAC) BCE]. A 0.1 molar solution of 2-hydroxymethyl-9,10-anthraquinone was prepared by dissolving 25 mmol (5.95 g) in 250 ml dioxane. The yellow solution was filtered and the solvent removed in vacuo. The residue was redissolved in 200 ml dioxane, before pyridine (2 ml, 25 mmol) was added. The solution was added dropwise to a stirred solution of triphosgen (2.5 g, 25 mequiv) in 50 ml CH2Cl2. The mixture was stirred at 20°C for 18 h, then diluted with 800 ml ethyl acetate. The organic phase was washed with 3 × 600 ml 80% saturated aqueous NaCl solution. After drying the organic phase over Na2SO4 the solvent was removed in vacuo to give a yellow solid of 2-anthraquinonemethoxycarbonyl chloride (MAC-Cl). The solid was dissolved in CH2Cl2 (250 ml, 0.1 M solution) and the solution used without further purification. ESI MS, mol. wt calculated for C16H9ClO4 300.7; found 302.1. Analysis calculated for C16H9ClO4, C 63.91, H 3.02; found, C 63.92, H 3.61.
Removal of LEV protecting group of BM during bDNA synthesis. The CPG column from the primary synthesis was attached to a 10 ml plastic syringe and rinsed with 10 ml acetic acid/pyridine (1:1 v/v). Approximately 10 ml HPAA (1:1) reagent [0.5 M hydrazine hydrate in acetic acid/pyridine (1:1, v/v)] was periodically pushed through over a period of 90 min at room temperature. The solid support was then rinsed with 10 ml acetic acid/pyridine (1:1 v/v) and detached from the syringe, followed by extensive rinsing with acetonitrile before brief drying under vacuum.
Preparation of CPG-propyl-NH2. CPG (2000 Å, 44 g) was suspended in 440 ml 95% ethanol in a round-bottom flask. The content was degassed under house vacuum for 5-10 min and a solution of 3-aminopropyltriethoxysilane (132 ml) was added and the flask shaken on a mechanical shaker for 18 h. The CPG was transferred to a 600 ml funnel with a medium fritted filter and washed with 5 × 400 ml ethanol and air dried. The CPG was thoroughly dried in an incubator at 60°C for 18 h. Positive ninhydrin and Fluram tests confirmed the presence of amino groups.
Scheme 1. (A) Synthesis of bDNA comb oligomer using BM(R3). (B) Schematic representation of bDNA comb oligomer, where Nx = primary sequence; Ny = spacing nucleotides; Nss = spacing sequence; Nz = secondary sequence.
REFERENCES
CiteULike 
Connotea 
Del.icio.us What's this?
This Article 
Abstract
Print PDF (131K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (9)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Horn, T.
Articles by Urdea, M. S.
Search for Related Content
PubMed
PubMed Citation
Articles by Horn, T.
Articles by Urdea, M. S.
Social Bookmarking
What's this?