(24k)Although cleaved, deprotected oligonucleotides from automated DNA synthesizers may often be used without purification, they are always contaminated with products of incomplete synthesis or deprotection, modified or improperly synthesized oligonucleotides, and protecting groups cleaved from the bases after synthesis. The decision as to whether or not purification is required depends on an evaluation of the probable degree of contamination of the desired full-length product (n-mer) with these undesired side products and the possible effects of these contaminants on the intended experiment. For example, a 10 % contamination of an oligonucleotide with a termination product one nucleotide short of full length (n-l mer) may have no appreciable impact on its use as an antisense inhibitor or as a primer for sequencing or polymerase chain reaction (PCR) amplifications. However, this level of contamination might well confound interpretation of primer extension reactions, prevent subcloning of an appropriate linker or adapter, or inhibit crystallization of a protein:nucleic acid complex. Thus, many investigators routinely purify oligonucleotides prior to use.
In general, purification strategies perform three significant functions: 1) separation of full-length n-mers from incomplete products (n-1 mer, n-2 mer, etc.); 2) removal of modified oligonucleotides resulting from incomplete deprotection, depurination, dimerization, branching, etc.; and 3) desalting of the oligonucleotide and removal of cleaved blocking groups. Common purification methods include ethanol (1) or butanol precipitation (2), thin layer chromatography (TLC) on silica gel plates (3), ion exchange (IE) or reverse phase (RP) HPLC (4,5), use of commercial hydrophobic purification cartridges (6) such as the oPC (Applied BioSystems, Inc.), or denaturing polyacrylamide-urea gel electrophoresis (PAGE,1). The purification method selected depends on the size of the oligonucleotide, the degree of purity required, the quantity to be purified, availability of instruments, the time available, the number of samples to be purified, and the cost of the method.
In DNA core labs where many oligonucleotides must be processed, single-step purification procedures provide adequate purity for most applications. Table 1 compares the efficiency and adaptability of several single-step protocols commonly used in core laboratories to purify moderate quantities (up to 1 umole) of synthetic oligonucleotides.
(24k)
Although none of these single-step procedures provides absolute purity, linkage of two or more of these strategies in series will provide extremely pure reagents. However, since we are primarily concerned with protocols that are readily adapted to routine use on multiple oligonucleotides in an active core laboratory, optimized multi-step strategies that yield very pure products required for some applications are only briefly discussed.
Probably the most straightforward, rapid and inexpensive single-step oligonucleotide purification strategies are direct precipitation with ethanol (1) or butanol (2). In these protocols the synthetic DNA is directly precipitated while salts and most of the cleaved hydrophobic blocking groups are eliminated. Thus, these approaches are frequently used to rapidly desalt and concentrate synthetic oligonucleotides, but are not recommended when contamination of the full-length product with incomplete or modified products would interfere with the planned use (e.g., cloning, crystallization, etc.).
Chromatography and gel electrophoresis strategies provide more highly purified oligonucleotide product than direct precipitation. Small ( ~ 25 residue) oligonucleotides are quite efficiently purified directly by TLC or HPLC. If HPLC instruments are not available, TLC is a low-cost and technically straightforward purification strategy for ~ 25 residue oligonucleotides, since full-length and n-1 mer resolution is very good for this size range using this technique. Moreover, multiple samples are easily processed in parallel on a single 20 X 20 cm TLC plate when only 50-100 ~g of product is required, and several plates can be developed in a single chamber. Thus, TLC is a method of choice to purify multiple small oligonucleotides if no HPLC is available.
HPLC is a good alternative purification strategy if an autosampler is available for automatic serial processing of multiple samples. In our experience IE-HPLC is considerably more efficient than RP-HPLC (or TLC) for resolution of full-length and n-1 mer oligonucleotides up to 25 residues in length. Moreover, both of these HPLC strategies efficiently remove both the cleaved hydrophobic protecting groups and modified or incompletely deprotected oligonucleotides from the final product. IE-HPLC efficiently resolves oligonucleotides up to ~25 residues in length, although established protocols require gradients of non-volatile salts (e.g., NaCl) which often must be removed from the sample in additional purification steps. However, our preliminary data suggests that gradients of volatile salts (ammonium- or triethylammonium-acetate) provide nearly as effective resolution as gradients of nonvolatile salts in IE-HPLC. Both RP- and IE-HPLC are relatively high capacity, low cost purification strategies for small oligonucleotides if an instrument with an autosampler is available. However, until IE-HPLC protocols using volatile salts are definitively established, RP-HPLC remains the HPLC method of choice for purification of S 25 residue oligonucleotides.
Oligonucleotides of ~25 residues are most frequently purified by PAGE or by RP-HPLC prior to removal of the hydrophobic S'-dimethy1trityl (DMT) blocking group (tritylon). Most commercial DNA synthesizers provide an option to synthesize oligonucleotides without removal of the S'-DMT group. Thus, full-length oligonucleotides with a highly hydrophobic S'-DMT terminus can be readily resolved from prematurely terminated products lacking the blocking group by RP-HPLC. These protocols extend the applicability of HPLC technology to purification of oligonucleotides 5 75 residues in length and provide a desalted product relatively free of cleaved protecting groups and prematurely terminated products. However, the RP-HPLC purified trityl-on oligonucleotide must be detritylated and repurified in subsequent processing steps for most applications. Hydrophobic purification cartridges provide a rapid alternative to RP-HPLC for purification of trityl-on oligonucleotides. Detritylation is performed directly on these cartridges and highly purified detritylated oligonucleotides are quickly prepared. However, these cartridges are quite costly, not reusable, and generally of relatively low capacity, thus obviating their use in many core laboratories. Denaturing PAGE still provides the best resolution of full-length and n-1 mer products for very large oligonucleotides (5150 residues), and probably best controls for the problems encountered with secondary structure in oligonucleotides of all sizes. However, incompletely deprotected products may co-migrate and therefore co-purify with the desired product on denaturing PAGE, and this method is laborious and time consuming in that multiple subsequent purification steps are required to separate the desired oligonucleotide from gel contaminants. Thus, PAGE is only a method of choice for purification of very large oligonucleotides or where secondary structure of the product is a serious potential problem.
Finally, some applications (e.g., crystallization) require that the oligonucleotidesbe more highly purified. Although the intent of this discussion is to compare convenient single-step protocols for purification of multiple oligonucleotides as in a core facility, coupling of these purification steps can provide very highly purified products. For example, IE-HPLC followed by RP-HPLC will efficiently separate fairly large quantities of full-length 525 residue oligonucleotides from prematurely terminated products, modified or incompletely deprotected oligonucleotides and cleaved protecting groups. For oligonucleotides >25 residues, RP-HPLC of trityl-on product followed by denaturing PAGE should provide very efficient purification although the capacity of PAGE is limited. Other combinations may be equally effective and must be evaluated in terms of the size of the oligonucleotide and the degree of purity required for the intended application.
Overall purity of an oligonucleotide is commonly determined, either before or after purification, by analytical IE- or RP-HPLC, denaturing PAGE or electrophoresis in gel-filled capillaries (CE). Although straightforward in application, both HPLC strategies have the same drawbacks for quality control as they have for purification; i.e., they must be run in series, size of the oligonucleotide i8 a limiting factor, and secondary structure can confuse interpretation. In practice, we find denaturing PAGE of a sample of the product after a rapid deprotection procedure (8) to be convenient for assessing purity because multiple samples ( > 20) can be directly visualized on a single gel using UV-shadowing. This method is rapid, inexpensive and generally not sensitive to secondary structure interactions, but it is unable to detect modified contaminants that co-migrate with the desired product. Recently, CE has proven to be an excellent alternative for analysis of purity of an oligonucleotide (7). This technique detects most common contaminants in oligonucleotide preparations and is not as sensitive to oligonucleotide secondary structure as other techniques. However, the equipment required for CE is relatively expensive and not widely available, and samples must be processed serially.
In summary, the best strategy for purification of a newly synthesized oligonucleotide depends on a number of factors, and each strategy has advantages and disadvantages. While denaturing PAGE clearly provides the most efficient resolution of full-length and prematurely terminated products of all sizes and readily fractionates oligonucleotides with significant secondary structure, incompletely deprotected or otherwise modified products may co-purify. Moreover, denaturing PAGE and the subsequent purification steps required render the technique quite cumbersome for routine use. In practice, we find that despite some limitations in resolution of full-length product from prematurely terminated products, RPHPLC with an autosampler is the most generally applicable single-step purification protocol for serial processing of multiple 5 25 residues oligonucleotides, since it removes many of the incompletely deprotected or otherwise modified contaminants as well as most of the prematurely terminated products. TLC is approximately equivalent to RP-HPLC in the purity of the product provided and is a good alternative for purification of small oligonucleotides if an HPLC is not available. Oligonucleotides > 25 residues are usually synthesized trityl-on and purified by RP-HPLC, although a subsequent detritylation and repurification step may be necessary. Purification cartridges are a good alternative although generally of low capacity and relatively high cost. Very large oligonucleotides or those with significant secondary structure are still best purified by PAGE. Combinations of these strategies implemented in series provide extremely high quality products. Quality of the final purified oligonucleotide can be assessed by HPLC or PAGE, or more sensitively by CE, although the latter approach requires a relatively sophisticated and expensive instrument.
References
1. ABI, Model 380,381 User Bulletin 13-Revised. April, 1987.
2. Sawadogo, M., and Van Dyke, M.W. (1991) Nucl. Acids Res. 19: 674.
3. Alvarado-Urbina, G., Sathe, G.M., Liu, W.C., Gillen, M.F., Duck, P.D., Bender, R., and Ogilvie, K.K. (1981) Science 214: 270-274.
4. Zieske, L.R. (1988) BioChromatog. 3: 112-117.
5. Pulaski, S.P., and Hatzenbuhler, N.T. (1989) BioChromatog. 4: 41-45.
6. ABI, Model 380A, 380B, 381A, 391, 392, 394. March, 1991.
7. Demorest, D., and Dubrow, R. (1991) J. Chromatog. 559: 43-56.
8. Reynolds, T.R., and Buck, G.A. (1992)BioTechniques 12: 518-521.
Return to the The ABRF Home Page