Members and sponsors are encouraged to share ideas, applications research and experiences, or technical problems by contributing to this column. In addition, short reviews covering aspects of biotechnology that are of interest to members are also encouraged. All contributions will be subject to peer review and the deadline for submitting articles for the March issue is February 1, 1993. Periodically, articles will also be solicited from non-members by the editors. Contributions may be mailed or faxed to the Editor, Ken Williams, W.M. Keck Foundation Biotechnology Resource Laboratory, P.O. Box 9812, 295 Congress Avenue, New Haven, CT 06536-0812, (203) 737-22()6, Fax (203) 737-2638. Contributions will not necessarily reflect views of the editorial board or the Association.
Interest in oligoribonucleotides has been revitalized recently. particularly in the area of antisense and "ribozyme" RNA. There are presently two approaches to the synthesis of RNA - enzymatic and chemical. Enzymatic methods. which take advantage of DNA template dependent synthesis of RNA by T7 or SP6 RNA polymerases, can produce substantial quantities of RNA that is up to several hundred nucleotides in length (1). However, short RNA fragments are produced with lower yield. In addition, RNA polymerase undergoes abortive reactions and may add undesired bases at the 3' end (1-2) Chemical synthesis is the method of choice to obtain RNA fragments in large scale that are suitable for structural analysis, therapeutic purposes, and/or for RNA containing modified nucleotides or unusual chemistry. Automated RNA synthesis, using a variety of chemical approaches, has made the solid phase .synthesis of RNA fragments possible on a large scale for oligoribonucleotides up to 20-3() nucleotides in length (3-5) and in smaller quantities for RNA of up to 77 nucleotides (6). This progress has underscored development of effective and large scale RNA purification techniques. In this paper I will briefly discuss strategies used for preparing highly purified samples of oligoribonucleotides.
Methods for the large scale purification of small cellular RNA, (tRNA and their fragments) were developed in the 197()s. However, oligoribonucleotides obtained by chemical synthesis represent a new challenge. Obviously, RNA samples of varying levels of purity are required for different experiments (protein:RNA binding experiments. fluorescence studies, NMR, crystallization. therapeutic applications, etc.). It is important to define the purity requirements prior to purification and to select techniques that will meet the chosen criteria. Purity of the RNA may be of particular significance when the RNA fragments are going to be administered as therapeutic agents. Chemically synthesized oligoribonucleotides are not homogeneous in size or quality. The level of purity depends on the chemistry used for synthesis, the length of the oligonucleotides, the scale of synthesis and the quality of reagents, and is progressively poorer with increasing size of the RNA fragment. Moreover, crude synthetic RNA is contaminated by shorter oligoribonucleotides more so than comparable DNA samples, not only due to abortive reactions but also due to significant internucleotide cleavage during synthesis and deblocking. In fact, limiting the exposure to ammonia during cleavage of RNA from the solid phase support and deblocking procedures is critical to minimize cleavage of the phosphodiester bond. Various deblocking procedures as well as last oligoribonucleotide deprotection chemistry have been developed to cope with this problem (3). Full length RNA is also not homogeneous because of RNA modification resulting from side reactions occurring during .synthesis (depurination, branching, 5'-2' linkages, dimerization, etc.) an(l incomplete deprotection procedures. Appropriate procedures for purification should be determined based on the desired purity, scale of synthesis, length, sequence, secondary structure of the oligoribonucleotide and available equipment.
Existing purification protocols take advantage of a wide range of techniques developed for purification of DNA fragments and for naturally occurring small RNA and their fragments ( I, 7 10). Problems encountered in handling oligoribonucleotides differ from those faced in DNA purification in a few important aspects: i) RNA can undergo rapid hydrolysis under a variety of conditions due to the presence of the 2-'OH group, which makes the phosphodiester bond of oligoribonucleotides significantly less stable than in DNA (3) and, ii) RNases, enzymes that hydrolyze RNA, are very common and are extremely stable. The accidental introduction of RNases via poor techniques or contaminated reagents represents a serious problem when working with RNA. Extreme caution is suggested when dealing with RNA samples. All glassware used for purification should be cleaned with chromic acid, rinsed exhaustively with glass-distilled water and baked for at least 4 hr at 300deg.C. Plastic containers, if needed, can be cleaned with 15% hydrogen peroxide in ().1 M NaOH. Buffers should be prepared with glass-distilled water and reagents of highest available purity should be autoclaved and stored at 2()deg.C, or filtered through nitrocellulose or nylon filters and used fresh. Disposable pipette tips and tubes should always be used. Disposable gloves are strongly recommended when contact with the RNA sample is likely. Special care must be taken to prevent contamination of the laboratory with RNases which can detect an otherwise successful purification. Chromatographic columns (HPLC, FPLC) should be used exclusively for RNA purification.
The simplest step in RNA purification is desalting of the oligoribonucleotide and it should precede any other purification .steps. Following .synthesis, the RNA can by purified from cleaved blocking groups and other contaminants by precipitation with ethanol or isopropanol in the presence of sodium or potassium acetate at pH 5 A number of RNA contaminants can also be extracted with n-butanol. It has been shown that n-butanol can effectively remove various impurities and simultaneously concentrate oligonucleotides. n-Butanol can remove organic reagents, as well as salts, SDS, urea and mononucleotides and, since it does not cause RNA precipitation but rather RNA dehydration, it can be used to purify even short oligoribonucleotides (11). Desalting of oligoribonucleotides using gel filtration on Sephadex G-10, 15, 25, or BioGel P-2 gels and volatile buffer is commonly used. Binding RNA to Sep-Pac C-18 cartridges and eluting with acetonitrile or, on a larger scale, binding to a DEAE-cellulose column and eluting with a volatile triethylammonium bicarbonate buffer at pH 7.5 (TEABC) followed by vacuum drying can also be used (5, X). Much better purification can be accomplished by polyacrylamide gel electrophoresis. Purification of small amounts ((). I - ().2 mg) of multiple samples of RNA up to 50-100 N in length can be done most conveniently in a single step using polyacrylamide gel electrophoresis under denaturing conditions (7-8 M urea at 60deg.C, 90-95% formamide, note that
formamide is toxic and both reagents should be deionized) (9). It has been observed that solubilization of RNA in formamide protects it from degradation during long term storage (12). RNA samples must be denatured before loading onto a gel by dissolving in 50% formamide or 8 M urea and heating to 75deg.C. However, RNA's with a high degree of secondary structure may need to be dissolved in 100% formamide. After electrophoresis, the RNA is localized by UV shadowing and is recovered by the "crush and soak" procedure (usually in the presence of SDS) or by electroelution. RNA can be recovered by ethanol precipitation or by binding to DEAE-cellulose and eluting with TEABC buffer. Complex mixtures of oligoribonucleotides can be separated on two-dimensional polyacrylamide gel electrophoresis (9). The first dimension is usually run at pH 3.3 in the presence of 7 M urea where RNA mobility depends more on base composition than length. The second dimension is run at pH 8.0 and at a higher gel concentration in the absence of urea. This method is extremely effective in separating very complex mixtures of small RNAs but it is limited to analytical rather than preparative scale. While polyacrylamide gel electrophoresis is the preferred procedure for small scale purification of oligoribonucleotides (up to 1-2 mg per gel) from N-l and shorter products, it is inefficient in removing partly deprotected fragments.
Large scale purification of RNA is best achieved with column chromatography. This method of purification can resolve milligram quantities of oligoribonucleotides and can be easily scaled up. A large variety of columns is available for both HPLC and FPLC techniques; this provides the opportunity to optimize resolution and speed. Column chromatography is also more effective in removing modified RNA since purification is based more on differences in primary sequence and/or secondary and tertiary structure than length (10). Reversed phase chromatography (RPC) is most commonly used to purify crude mixtures of RNA fragments. C-18 RPC has been used successfully for short oligoribonucleotides. For longer RNA (>14 N), wide pore C-4 and C-8 columns are better. The use of reversed phase chromatography is particularly effective when the dimethoxytrityl group (DMT) "On/Off" approach is used. In the first step, the S'-DMT protected oligoribonucleotide is purified on RPC HPLC (5-10 mg of crude RNA can be injected onto a I x 25 cm RPC column). Full length oligoribonucleotides carrying the DMT group can be easily separated from other unprotected products. After removal of the DMT group, the RNA is chromatographed (3-4 mg of RNA per injection) on the same column to remove other impurities. Usually, separation is done in the presence of 100 mM triethylammonium acetate (TEAA) buffer at pH 6.0-6.5 and RNAs are eluted with an acetonitrile gradient. RNA fragments with strong secondary structure can be separated at elevated temperatures of up to 60deg.C. Although both acetonitrile and TEAA are volatile, RNA should not be lyophilized from this buffer system since the solution will acidify during this procedure causing partial RNA depurination. It is better to transfer the oligonucleotide to TEABC buffer prior to lyophilization. Large RNA samples can also be purified over strong anion exchange columns with either an HPLC or an FPLC. This chromatographic step permits resolution of the N- I product from full length RNA of up to 30 N in 3-4 mg scale when appropriately chosen isocratic elution with salt (usually NaCl) is used (13). The advantage of this method is that RNAs can also be purified under denaturing conditions in the presence of 7 M urea or high pH if needed. Separation can be carried out for example on DEAE-Fractogel (BioRad) using HPLC ( I ) or on MonoQ (Pharmacia) using FPLC (14). The drawback of these methods is that they are usually unable to resolve modified oligoribonucleotides and the RNA has to be desalted in a separate step.
An inexpensive and simple alternative is thin layer chromatography (TLC). Oligoribonucleotides can be purified by thin layer chromatography like DNA fragments (7). Silica gels, PEI and RPC TLC plates can be used for RNA purification. Although it is a very good method for purification of short RNA (up to 30 N) (recovery of RNA is simple and multiple samples can be processed in parallel), it is limited to rather small scale purifications and often, identification of the desired product can be difficult because mobility depends on factors other than length.
When the highest purity RNA is required, as for NMR or crystallization experiments, the purification procedure should involve at least two steps in which separation is based on different properties of RNA. For example, the procedure could combine RPC-HPLC chromatography (with either DMT blocking group on or off) with polyacrylamide gel electrophoresis or anion exchange HPLC chromatography in the presence of 7-8 M urea. Several scenarios are possible and their effectiveness should be evaluated considering the length of oligonucleotide, scale and resources that are available.
Lastly, the RNA samples should be tested using analytical methods for possible contaminations, sequence errors and biological activity if applicable. Several high resolution procedures are recommended which can provide important information about purity and homogeneity: polyacrylamide gel electrophoresis (particularly two dimensional); capillary electrophoresis, which is both highly effective and fast (3); and/or analytical chromatography on high resolution RPC or anion exchange columns using HPLC or FPLC. Two dimensional TLC (8) is clearly the most convenient for base composition analysis. Subsequently, sequence analysis of RNA should also be performed.
References
1. Milligan, J.F. and Uhlenbeck, O.C., (1989) in Methods in Enzymology 180, 51-62.
2. Martin, C.T., Muller, D.K. and Coleman, J. E., (1988) Biochemistry 27, 3966-3974.
3. Vinayak, R., Anderson, P., McCollum, C. and Hampel, A., (1992) Nucl. Acids Res. 20, 1265-1269.
4. Scaringe, S.A., Francklyn, C. and Usman, N., (1990) Nucl. Acids Res.. 18, 5433-5441.
5. Odai, O., Kodama, H., Hiroaki, H., Sakata, T., Tanaka, T. and Uesungi, S., (1990) Nucl. Acids Res. 18, 5955-5960.
6. Ogilvie, K.K., Usman, N., Nicoghosian, K. and Cedergren, R.J., (1988) Proc. Natl. Acad. Sci. USA X5, 5764-5768.
7. Buck, G.A., Wenner, R.J. and Reynolds, T.R., (1992) ABRF News 3, 9-11.
8. Silberklang, M. Gillum, A.M., and RajBhandary, U.L., (1979) in Methods of Enzymology 59, 58-109.
9. De Wachter, R. and Fiers, W., (1982) in Gel Electrophoresis of Nucleic Acids (Rickwood, D. and Hames, B.D., eds) p77-116.
10. Tanner, N. K., (1989) in Methods in Enzymology 180, 25-50.
11. Cathala, G. and Brunel, C., (1990) Nucl. Acids Res. 18, 201.
12. Chomczynski, P., (1992) Nucl. Acids Res. 20, 3791-3792.
13. Joachimiak, A., unpublished.
14. Dreef-Tromp, C.M., van den Elst, H., van den Boogaart, van der Marel, G.A. and van Boom, J.H., (1992) Nucl. Acids Res. 20, 2435-2439.
Return to the The ABRF Home Page