into highly reactive activated compounds (Figure 1.9) is one of the most remarkable aspects of phosphoramidite synthesis chemistry.223,234 This activation is performed by mixing a solution of nucleoside phosphoramidite with a solution of a very mild acid (activator). Initially, the activators were tertiary amine hydrochloride salts such as N,N-dimethylaniline hydrochloride, but since these reagents were hygroscopic and difficult to work with 1H-tetrazole soon became the preferred activator.102 1H-tetrazole is a stable, nonhygroscopic solid that can be readily obtained in high purity by either crystalization or sublimation and 0.45 to 0.5 M solutions in acetonitrile are widely used. Although tetrazole and its substituted derivatives are potentially explosive (especially when attempting to sublime 1H-tetrazole), the room-temperature solutions required for DNA synthesis are not considered unduly hazardous.

The first step in the mechanism of phosphoramidite activation is protonation of the trivalent phosphorus, followed by slow displacement of the secondary amine (usually N,N-diisopropylamine) by tetrazolide. The tetrazolide then immediately reacts with the 5'-hydroxyl group of the oligonu-cleotide attached to the solid-phase support. The activation and coupling reaction is very fast, and for deoxyribonucleoside phosphoramidites coupling times of between 30 to 60 sec are generally used. Oligoribonucleotide synthesis requires longer coupling times (up to 600 sec) because the phosphoramidites are more hindered by the adjacent 2'-hydroxyl protecting group (especially with the more bulky 2'-0-i-butyldimethylsilyl protecting group). However, the 1H-tetrazole solutions are acidic enough to cause small amounts of detritylation during prolonged coupling reactions.235 This detritylation can result in the formation of unwanted oligonucleotide impurities that are longer than the expected product (i.e., N + 1, N + 2, etc. long-mers). Fortunately, however, normal small-scale synthesis conditions produce undetectable, or barely detectable, amounts of these impurities.

As with every other step in solid-phase oligonucleotide synthesis, the activation step has been extensively studied, and a variety of other activating reagents have been proposed. More acidic 5-(p-nitrophenyl)-1H-tetrazole205,216,218 and 5-ethylthio-1H-tetrazole210,211,221,236 have been used for RNA synthesis; benzimidazolium triflate,237 imidazolium triflate,170 and pyridine hydrochloride/imidazole169 have been used for O-selective phosphitylation; and more nucleophilic N-meth-ylanilinium trifluoroacetic,227 1H-tetrazole/DMAP,238 1H-tetrazole/N-methylimidazole239 solutions, as well as 4,5-dicyanoimidazole239 and other azole activators240 have been developed for difficult couplings. Other activators include pyridinium salts,241 pyridinium trifluoroacetate/N-methylimida-zole,242 trimethylchlorosilane,243 and 2,4-dinitrophenol.244 These activators are generally faster and more soluble than 1H-tetrazole but are also usually more expensive or more hygroscopic, and so none of them has replaced 1H-tetrazole for general oligodeoxyribonucleotide synthesis.

The high reactivity of the activated nucleoside phosphoramidites also leads to phosphitylation of guanine bases at the O-6 position.171 These guanine base modifications are unstable as long as the phosphorus atom remains in the trivalent oxidation state. Therefore, the newly formed internu-cleotide phosphite linkages cannot be oxidized into the desired pentavalent phosphate oxidation state until the guanine base modifications are removed. Fortunately, these modifications are simultaneously removed at the same time unreacted 5'-hydroxyl groups (typically <0.5 to 1% after each coupling reaction) are acetylated.

Therefore, the next step in the synthesis cycle is treatment with an acetic anhydride reagent to cap (block) off residual 5'-hydroxyl sites and reverse the guanine base modifications.172 This capping step is very important because it prevents unreacted molecules from participating in further chain extension reactions. If this step were omitted, then a complex mixture of oligonucleotide impurities containing random deletions would result and isolation of the desired full-length product would be more difficult. Therefore, a fast and efficient capping step is required. Acetic anhydride is an inexpensive and effective acylating reagent as long as it is used in the presence of a nucleophilic catalyst such as 4-dimethylaminopyridine (DMAP) or N-methylimidazole (NMI). DMAP catalysis was used originally,245 but subsequent analysis found that it produced trace amounts of mutagenic 2,6-diaminopurine base modifications.246 Since catalysis with NMI does not produce these modifications, DMAP is no longer used. The acetylation reagent used for the capping step is packaged into two solutions: Cap A, acetic anhydride/2,6-lutidine/THF; and Cap B, NMI/THF; these are individually installed on the DNA synthesizer. During the capping step, the synthesizer simultaneously delivers both reagents to the synthesis column, where they mix and rapidly (10 to 60 sec) acetylate unreacted sites.

A variety of other capping reagents have also been proposed, but the only alternate reagent of significance is the t-butylphenoxyacetic anhydride reagent required for use with more labile t-butylphenoxyacetyl or phenoxyacetyl base protecting groups.247 In this case, small amounts of the base labile protecting groups may be lost during synthesis, and if they were replaced by the much more stable acetyl groups, then incomplete base deprotection would result.

The last step in the coupling cycle is oxidation of the trivalent phosphite linkage into a more stable pentavalent phosphate linkage. The phosphite compounds are easily oxidized and in Letsinger's original phosphite-triester coupling reactions95 the oxidation step was performed using a solution of iodine and water in THF. This oxidation is very fast (<10 sec), quantitative, and convenient. The progress of this step can be observed by the almost instantaneous disappearance of the dark iodine color on contact with the phosphite compounds. In early solid-phase oligonucleotide synthesizers, an iodine/water oxidation reagent containing 2,6-lutidine as an acid scavenger was employed.36 Unfortunately, these solutions had a limited shelf life because a black iodine containing precipitate eventually formed. However, iodine/water solutions containing pyridine as the acid scavenger are stable indefinitely248 and all automated oligonucleotide synthesizers now use these solutions for the oxidation step. This is somewhat surprising considering the high sensitivity of coupling reactions to moisture. However, in practice any residual moisture from the oxidation reagent is easily washed away from synthesis column using anhydrous acetonitrile. Therefore, nonaqueous oxidation reagents based on organic peroxides, such as m-chloroperbenzoic acid249 or t-butylhydroperoxide,250 have never been widely used.

The need for a separate oxidation step, which was originally considered an inconvenience, has turned out to be extremely beneficial because the oxidation step allows the specific introduction of a variety of internucleotide backbone modifications to be easily accomplished.40,41 In the simplest case, isotopically labeled phosphate linkages can be obtained by using water labeled with an oxygen isotope. Iodine and a substituted amine can be used instead of iodine/water to produce oligonucle-otides with modified phosphoramidate linkages useful for the synthesis of oligonucleotide conjugates. Elemental sulfur or selenium can also be used to produce phosphorothioate or phosphorose-lenoate linkages, which have one of the nonbridging oxygen atoms replaced with either a sulfur or selenium atom.

The phosphorothioate modification (Figure 1.14) is especially significant,251 since this is a rather conservative change. The native charge on the phosphate group is retained and the size of a sulfur atom is only slightly larger than an oxygen atom. Even though the modified phosphorus center is now chiral and each phosphorothioate linkage produces a set of two diastereoisomers, these compounds are excellent analogues for studying both the stereochemical course of reactions occurring at phosphorus and antisense inhibition of gene expression. The latter application is possible because phosphorothioate linkages are considerably more stable toward nucleases than are phos-phodiester linkages.

Elemental sulfur was first used in 1978 by Burgers and Eckstein to synthesize dinucleoside monophosphorothioates from phosphite triesters.252 Longer oligonucleotide phosphorothioates were later prepared by solid-phase techniques using similar sulfurization conditions.253 However, the sulfurization step with elemental sulfur was relatively slow (7.5 min) and required the use of unpleasant smelling carbon disulfide as the solvent. Since that time a great deal of effort has gone into the development of improved sulfurization reagents to increase the speed and efficiency of this

0 0

Post a comment