O

In practice, the length of overnight deprotection is not a serious restriction. A more important consideration is the amount time that synthesis instrumentation is tied up performing the automated cleavage of the oligonucleotide from the support. In addition, certain oligonucleotide modifications cannot withstand the strongly basic conditions required for cleavage and deprotection. In these cases, replacement of the traditional succinic acid linker arm, i.e., 37 with the more labile hydro-quinone-O,O'-diacetic acid (Q-Linker) linker arm, i.e., 78 can be helpful.268 This allows automated cleavage to be significantly shortened (2 min vs. 60 min cleavage time) and, in conjunction with labile base protecting groups, allows base labile oligonucleotide modifications.

Finally, in the case of oligonucleotides purified by reverse-phase HPLC or solid-phase extraction cartridges, the 5'-terminal dimethoxytrityl group must be removed by manually performing an acidic detritylation step to complete the deprotection process. However, because the base protecting groups have been removed by this stage, the oligonucleotide is much less susceptible to depurina-tion, and this step can be safely performed with either 80% acetic acid or trifluoroacetic acid.

1.5 SYNTHESIS ON SOLID-PHASE SUPPORTS 1.5.1 The Solid-Phase Approach

Although solution-phase synthetic methods for coupling small units together were developed many years ago, the large number of couplings needed to assemble useful sequences presented a very daunting task. This was because each step required some type of workup, extraction, or purification, and the labor and cumulative loss of material from these manipulations rapidly became significant problems. The problems involved in performing so many repetitive steps were addressed by R. B. Merrifield with the introduction of solid-phase synthesis.76269270

In this strategy (see Figure 1.6), a large insoluble support is covalently linked to the end of the sequence being assembled. The product on the surface of the support is available to react with reagents in the surrounding solution phase. The extended products remain covalently linked to the insoluble support, while unreacted reagents remain free in solution. Therefore, at the completion of each step the products can be rapidly and conveniently isolated by simply washing the unbound reagents away from the support. This can be performed as easily as filtering off the support and washing it with solvent. The support with its attached product is then ready for immediate use in the next step, as long as unwanted moisture contamination has not been introduced (in which case the support must be dried before use). In practice, it is convenient to handle the supports inside sealed reactors or columns so exposure to the atmosphere is minimized. This is also ideal for automation and the necessary reagent additions and solvent washes are readily mechanized. The process of adding each unit is repeated over and over until the desired sequence has been assembled on the surface of the support. The product can then be released from the support by cleavage of the covalent attachment linking the product to the surface. Removal of any remaining protecting groups completes the synthesis.

The big advantage of solid-phase synthesis is the ease with which immobilized products can be separated from other reactants and by-products. Solid-phase supports also permit relatively small quantities of material to be synthesized. This is because the additional physical bulk of the support, which is «10 to 100 times the mass of the attached nucleoside, can be handled more easily than the nucleoside alone. Also, confinement of the support inside a synthesis column eliminates handling losses. A small synthesis scale is important because of the high cost of reagents. Very little material is required for many biochemical applications and, as instrumentation has improved, the synthesis scale has decreased. The simplicity and similarity of the steps required for each chain extension reaction also greatly facilitate the synthesis of modified oligonucleotides, and a large number of modified substituents are available as phosphoramidite derivatives. Chimeric oligonucleotides containing peptide or PNA (peptide nucleic acids) sequences can also be prepared.271,272 Finally, multiple bases ("mixed bases") can be incorporated at defined positions by using a mixture of different monomers, instead of a single monomer, in the chain-extension reaction. This is also the procedure used in "base doping," when only one base, at random, within a particular section needs to be mutated.273

Although a powerful technique, solid-phase synthesis has its drawbacks. The main limitation is the need for very high coupling yields in every chain-extension step. This is because the overall yield of product decreases rapidly as the number of consecutive chain extension steps increases. For example, if each base addition step had a yield of 90%, then the amount of dinucleotide produced (one base addition) is 90%, the yield of trinucleotide (two base additions) is 0.90 x 0.90 x 100% = 81%, the yield of tetranucleotide (three base additions) is 0.90 x 0.90 x 0.90 x 100% = 73%, and so on. Note that the first nucleoside is attached to the insoluble support prior to the start of oligonucleotide synthesis and the efficiency of that step is not included in the calculation. The mathematical relationship between the overall yield (OY) and the average coupling efficiency (AY) is either where n = the number of coupling steps and N is the length of the oligonucleotide. The second equation assumes that the synthesis was performed by extending the product by one base at a time, as is usual with phosphoramidite reagents. Therefore, very high yields are required in every step and coupling yields that would be acceptable for most solution-phase reactions (such as the 90% yield assumed in the above example) are not adequate. Phosphoramidite reagents, which can produce average coupling efficiencies of 98 to 99%, are the most preferred, and oligonucleotides of up to 150 to 200 bases in length have been prepared.

Another consequence of having less than 100% coupling efficiencies is the accumulation of failure sequences containing deletions. The number of these failure products can be greatly reduced by the addition of a capping step after each chain-extension reaction. This step, which typically uses acetic anhydride to acetylate nonextended molecules, prevents these failure sequences from participating in any further reactions. However, a series of failure sequences, each one base shorter than the desired full-length product, will be present at the end of the synthesis.

Separating the full-length product (of length N) from the shorter failure sequences, and especially the N - 1 failure sequence, is another significant problem. This purification problem becomes more difficult as oligonucleotide length increases, and for oligonucleotides greater than ~30 bases long, only PAGE (polyacrylamide gel electrophoresis) has sufficient resolving power to separate the full-length product from the N - 1 component. Fortunately, however, many biochemical applications do not have stringent purity requirements and, if the coupling efficiency is high enough, the mixture of products produced can often be used with either minimal (desalting) or no purifi-cation.274 However, the use of di- or trinucleotide blocks, which was a common approach in the older phosphotriester coupling method, has recently been proposed for phosphoramidite synthesis as well. In this strategy, the closest impurity to the full-length product (N-mer) is only N - b (b = block size) bases long instead of N - 1, and this makes large-scale purification much easier.188,189

Analysis of the synthetic products, while still attached to the surface of the insoluble support, also presents a major difficulty for those researchers developing new techniques or new solid-phase supports. This is an especially significant problem for applications using immobilized arrays because removal of the products for characterization is often difficult, if not impossible. Only a few studies have been performed, using techniques such as magic angle spinning NMR, ellipsometry, interfer-ometry, or optical wave guides on immobilized oligonucleotides,275-277 but these techniques have not been able to provide specific information about the quality of the oligonucleotide syntheses.

0 0

Post a comment