Jack Pollard


Several methods existfor the purification of oligonucleotides following chemical synthesis.The advantages of purification on denaturing polyacrylamide gels are speed,simplicity, and high resolution. Denaturing polyacrylamide gels can resolveoligonucleotides from 2 to 300 bases, depending on the percentage of polyacrylamideused (see Table 1).This method is thus useful not only for isolating chemically synthesizeddeoxyribonucleotides but also small RNAs or other single-stranded oligonucleotides.After gel setup, samples are loaded onto a urea-based denaturing gel, separatedby electrophoresis, and finally recovered from the crushed gel slice.



10x and 1x TBE buffer,pH 8
38% acrylamide/2%bisacrylamide
TEMED (N,N,N',N'-tetramethylethylenediamine)
10% ammonium persulfate(in water 2 1 month old, store at 4° C)
urea loading buffer
3 M sodium acetate,pH 5.2
TE buffer
Thin-layer chromatography(TLC) plate with fluorescent indicator (e.g., Silica Gel F-254 or IB-F)
Glass plates, spacers,and combs for pouring gels
Acrylamide gel electrophoresisapparatus
DC power supply
0.2 micron filter(Gelman Sciences)

Additional reagentsand equipment for phenol extraction and ethanol precipitation.


Prepare the sample

1. Follow the appropriatedeprotection protocol to prepare the sample for electrophoresis.

Be sure to lyophilizethe sample to dryness. The samples will generally appear as a an off-whitepowder following deprotection and lyophilization. If a yellowish liquidor crusty pellet remains, rather than an off-white powder, resuspend thepellet in 0.5 ml distilled water and add 1/10 volume 3.0 M sodium acetatepH 5.2. Add 3 volumes of absolute ethanol to precipitate by chilling to-80 °C for approximately 20 minutes. Centrifuge for 10 minutes at 16000xgand 4°C. Decant and save the supernatant. Wash the pellet in 70 percentethanol Lyophilize the pellet to dryness.

Prepare the gel

2. Assemble the gelcasting apparatus.

Gel spacer and castingsystems have been developed to avoid leakage. Those which avoid sealingthe gel with tape are best, and recently, gel casting boots that lack bottomspacers have become available (GibcoBRL). Greasing the side /bottom spacersor pouring an agarose plug for the gel is not necessary if some care istaken to ensure that the bottom of the plate assembly is completely sealed.Clean the gel plates thoroughly by washing them with warm soapy water followedby an ethanol:water rinse. However, if the plates are particularly dirtyor if the complete removal of any residual nucleic acids is required, theplates may be soaked in an 0.1 M NaOH for 30 minutes prior to washing.If the gel is particularly thin (<1 mm), silanizing one or both of theplates facilitates post-electrophoretic separation of the gel from theplate.

3. Prepare the gelsolution (see Table 1 for appropriate acrylamide concentrations for resolvingsingle stranded DNAs). For a denaturing acrylamide gel of 20 cm x 16 cmx 1.6 mm, 60 ml of gel solution is sufficient, and it can be made by mixingthe following:

25.2 g urea (finalconcentration of 7 M)

6 ml 10x TBE buffer

desired amount of 40% acrylamide/ 2 % bisacrylamide needed for resolution

water to a final totalvolume of 60 ml

Pick a concentrationof acrylamide that will allow the single stranded nucleic acid to migrateapproximately one-half to three-fourths the way through the gel when theloading dye has reached the bottom of the gel. This allows for good separationof non- and full-length products.

Use a flask thathas a wide mouth and a spout for pouring.

Caution: Alwayswear gloves, safety glasses, and a surgical mask when working with acrylamidepowder since it is a neurotoxin.

Commercially preparedpolyacrylamide solutions (National Diagnostics) are available and highlyrecommended since they have long shelf lives and do not involve massingthe neurotoxic acrylamide powder.
Table1 Concentrations of Acrylamide Giving Optimum Resolution of DNA FragmentsUsing Denaturing PAGEa

Acrylamide (%)


separated (bases) 

migrationof bromophenol blue (bases)migrationof xylene cyanol (bases)
302to 8620
208to 25828
1025to 3512 55
835to 451975
645to 7026105
570to 30035130
4100to 500~50~230
aData from Maniatis et al., 1975.

4. Heat the mixtureby immersing the flask in a 60 °C hot water bath or under running tapwater to speed the dissolution of the urea and acrylamide. Once most ofthe urea and acrylamide have gone into solution, vigorously agitate thesolution for approximately 20 minutes with magnetic stirring to ensurecomplete mixing.

5. Add 40 ml of TEMEDand swirl the flask to insure thorough mixing. Immediately add 300 ml of10 % APS and mix thoroughly. Polymerization has begun so all SUCCEEDINGsteps must be PERFORMED promptly. Pour the acrylamide between the gel platesand insert the comb. Clamp the comb in place at the top of the gel to avoidseparation of the gel from the plates as the acrylamide polymerizes. Allowthe gel to polymerize for approximately 30 minutes.

For thick gels pourthe acrylamide directly from the mixing flask, but for thinner ones, asyringe fitted with the needle is useful. By pouring the gel slowly witha tilt 45û relative to the bench top and starting from one corner,bubbles may be largely avoided. Also, polymerize the gel while it is lyingflat to avoid undesirable hydrostatic pressure on the gel bottom.

TEMED may be storedindefinitely at 4°C, but the ability of APS to efficiently initiatethe free radical induced acrylamide polymerization diminishes greatly overtime. Make a new stock every month and store at 4°C.

Caution: Be sureto wear safety glasses while pouring the gel since splashing of the neurotoxic,unpolymerized acrylamide is common..

Run the gel

6. After polymerizationis complete, remove the comb and any bottom spacers from the gel. Washthe gel plates free of spilled acrylamide and be sure that the spacersare properly seated and clean.

7. Fill the lower reservoirof the electrophoresis tank with 1X TBE. Initially, place the gel intothe lower tank at an angle to avoid air bubbles forming between the platesand the gel bottom. Clamp the gel plates to the top of the electrophoresistank and fill the upper reservoir with 1X TBE so that the wells are covered.

A syringe with abent needle may be used to remove air bubbles trapped under the gel thatwill disrupt the current flow.

8. Use a DC power supplyto prerun and warm the gel for a least 30 minutes at 20-40 V/cm (constantvoltage).

9. Resuspend the oligonucleotidepellet obtained from step 1 in 1X urea loading buffer by heating it at90°C for 5 minutes.

The amount of samplethat can be loaded depends on the efficiency of the synthesis reaction.At least 10 mg of material in a single band 2 cm wide is required tocast a clear UV shadow. The longer the oligonucleotide, the less full lengthproduct obtained.

Use an amount of loadingbuffer that is consistent with loading approximately 25 % of a 0.2 m-molsynthesis of a 20-mer oligonucleotide per 2 cm X 2 cm X 1.6 mm well. Thiswill give sharp bands with good resolution. Up to 4 fold more may be added,but the resolution will suffer.

10. Rinse the wellsthoroughly with 1XTBE solution immediately prior to gel loading.

The 7 M urea dissolvedin the gel will start to diffuse from the wells thereby creating a denselayer at the bottom of the wells that prevents sample loading and decreasesresolution. Rinsing eliminates this problem.

11. Load the samples.

Tracking dyes suchas bromophenol blue and xylene cyanol (see table 2.12.1 for migration data)may be added to the samples or in empty lanes to monitor migration.

12. Electrophoresethe gel at 20-40 V/cm (constant voltage) until the tracking dyes indicatethat the oligonucleotide has migrated one-half to three-fourths the waythrough the gel.

The speed of electrophoresisis directly proportional to the voltage gradient across the gel. The currentin the circuit and the heat generated for higher percentage gels (>15 %acrylamide) are corresponding smaller since the increased acrylamide concentrationleads to greater resistance. While some heating of the gel during electrophoresisis desirable since it helps to denature the sample, temperatures in excessof 65° C should be avoided. All gels should be monitored to make surethat they do not generate so much heat that the plates crack. For example,while a 20 % gel can be electrophoresed at 800 V with few problems, an8 % gel under the same conditions would likely generate too much heat forthe apparatus to dissipate.

13. When the oligonucleotideis sufficiently resolved, turn off the power supply and detach the platesfrom the electrophoresis tank. Pry off the top plate. Cover the gel withplastic wrap (taking care to avoid bubbles and folds) and invert the plateonto a TLC plate with a fluorescent indicator. Using a spatula, peel acorner of the gel away from the plate and onto the plastic wrap. Pry offthe remaining plate and place another sheet of plastic wrap on top of thegel.

Recover the oligonucleotide

14. Visualize the bandson the gel by briefly exposing them to short-wave ( 254 nm) radiation froma handheld lamp. The bands will appear as black shadows on a green background.Outline the bands using a marking pen.

The desired bandis generally the darkest one on the gel (excluding material that runs atthe dye front); it should also be the slowest migrating band unless deprotectionwas incomplete. Lighter bands containing partially protected oligonucleotidesif they are present will migrate considerably above the major fully deprotectedband. If the stepwise efficiency of the synthesis is low, a smear may beseen instead of a clear band. Cut out the top of the smear.

Avoid unnecessarilylong UV exposure which will damage the nucleic acids.

Unpolymerized acrylamideabsorbs strongly at 211 nm and may also cause shadowing that is confinedto the edges and wells of the gel.

15. Cut out the bandsdirectly with a clean scalpel or razor blade.

16. Chop the gel slabsinto fine particles by forcing the gel through a small bore syringe toaid the diffusion of the oligonucleotide from the matrix. Place the crushedgel slab in a 15 ml spin tube capable of withstanding high temperatures.

17. Add 3 ml of TEfor every 0.5 ml of gel slab. Freeze the sample for 30 minutes at -80°Cor until frozen solid. Quickly thaw it in a hot water bath and let soakfor 5 minutes at 90°C. Elute on a rotary shaker overnight at room temperature.

This freeze-rapidthaw approach (Chen Z., et al., 1996) greatly decreases elutiontime and increases yield by allowing ice crystals to break apart the acrylamidematrix. A 20-mer oligonucleotide is typically recovered in a 80 % yieldafter 3 hours of rotary shaking thereby making this technique comparableto electroelution.

Since elution isa diffusion-controlled process, more buffer will aid in elution efficiency.Also, note that longer oligonucleotides will take longer to diffuse fromthe gel. If speed is essential and high yields are dispensable, enoughsample can be obtained for most experiments in only a few hours of extraction.Increasing the temperature to 37°C will also speed the process. Yieldmay be increased upon repeated elutions.

18. Spin the tube topellet the gel fragments and use a syringe to remove the supernatant. Filteroff any remaining acrylamide fragments by passing the suspension throughan 0.2 micron filter and into a fresh 15 ml spin tube.

19. Concentrate thesample by extracting against an equal volume of n-butanol. Remove the upperbutanol layer and repeat until the lower aqueous volume is convenient forprecipitation.

About 1/5 volumeof the aqueous layer is extracted into the organic butanol layer for everyvolume of butanol used. If too much butanol is added and the water is completelyextracted in the butanol, simply add more water and concentrated again.

20. Add 3.0 Msodium acetate pH 5.2 to a final concentration of 0.3 M and use2 volumes of absolute ethanol to precipitate DNA and 3 volumes for RNA.Chill for 20 minutes at -20°C. Pellet the oligonucleotide by centrifugingat 12000xg for 10 minutes.

Do not attempt toprecipitate small oligonucleotides (220 bases) in the presence of ammoniumions. If the samples prove refractory to precipitation, use a 1:1 mix ofethanol:acetone or 6 volumes of acetone for precipitation. A rinse with95 % ethanol will remove undesired salts.

21. Redissolve theoligonucleotide in TE buffer if appropriate.



Urea loadingbuffer

8 M urea

20 mM EDTA

5 mM TRIS pH7.5

0.5 % dye by mass eitherxylene cyanol, bromophenol blue, or both.

add one volume of loadingbuffer to sample if a solution or enough to dissolve a powder.



Background Information

The traditional alternativeto gel purification of oligonucleotides has been high-performance liquidchromatography (HPLC). Although alkali perchlorate salts HPLC systems canachieve very high resolution of small and medium sized oligonucleotides(<60 bases), electrophoresis provides superior capacity and resolutionover a greater range of sizes and is simpler to set up and operate. Separationtimes using HPLC may be faster (<30 min) than for gels, but initializationof the system and product workup tend to negate this advantage.

Purification of oligonucleotideson low-pressure reverse-phase cartridges is technically simpler than gelelectrophoresis and faster (<2 hr). However, these cartridges offerno separation of desired product from failed sequences, and if not usedproperly, allow contamination of the final product with low-molecular-weightcompounds that often inhibit subsequent enzymatic manipulation of oligonucleotides.For short oligonucleotides synthesized with high yield, very simple purificationmethods (e.g., gel filtration or ethanol precipitation) are adequate forsome applications such as sequencing or PCR primers which do not requireabsolutely homogenous material.

The high resolutionand high capacity of polyacrylamide gels makes them the method of choicefor the purification of oligonucleotides. Urea disrupts hydrogen bondingbetween bases and thus allows oligonucleotides to be resolved almost exclusivelyon the basis of molecular weight as opposed to secondary structure. However,it should be noted that oligonucleotides of equivalent length but differentsequence will still migrate slightly differently. Thus, mixed sequenceswill appear as broader bands than homogeneous sequences (AppliedBiosystems, 1984). Also, RNA eletrophoreses through the gelmore slowly than does DNA of comparable size. Finally, the compatibilityof the chemistries of modified nucleotides incorporated into the nucleicacids and acrylamide matrix should be checked before PAGE purification(oligonucleotides bearing thio groups seem to undergo Micheal Additionto the acrylamide thereby rendering them irreversibly capped).


Critical Parameters

For most applications,the separation of oligonucleotides from mononucleotides and protectinggroups provides adequate purification. In those cases where separationof oligonucleotides from nearby failure sequences is essential, however,the most critical parameters to be considered are the percentage of acrylamideand the amount of sample loaded. If maximum resolution is desired, thenonly 50 to 100 mg of material should be loaded per 2 cm x 2 cm x 1.6 mmwell. The percentage of acrylamide that will give optimal resolution isgiven in Table 2.12.1 and can be determinedempirically by running a small portion of the starting material on trialgels and staining with ethidium bromide. By running long (20- to 30-cm)gels, oligonucleotides of up to 100 bases can be cleanly separated fromn-1 and n+1 products. If an oligonucleotide contains extensive self-complementarysequences or polyguanosine tracts, it may not be completely denatured in7 M urea, and thus, cannot be cleanly separated from failed synthesis products.To overcome this difficulty, samples can be electrophoresed on gels containing20 M formamide instead of urea (Frank etal., 1981).


All of the problemsthat apply to nondenaturing PAGE are relevant here (see Commentary,UNIT 2.7). However, most failuresin purification will occur because the initial synthesis reaction has beeninefficient. In almost all cases, it is better to resynthesize a poor yieldingoligonucleotide than to attempt to isolate a small amount of full-lengthproduct from a starting material seriously contaminated with failure sequences.If the oligonucleotide cannot be resynthesized, relatively small amountsof product can be visualized by autoradiography providing the startingmaterial is end-labeled with polynucleotide kinase and [g-32P]ATP (thestarting material should not contain residual ammonium, which inhibitsthe enzyme) (reference kinasing unit). It is convenient to end label asmall amount of starting material with radioactive ATP; the remainder shouldbe phosphorylated using cold ATP so that it will not migrate differentlyfrom the labeled tracer. Smaller amounts of starting material should beloaded on thinner (approximately 0.75- to 1.0-mm) gels in narrower lanes(approximately1.0 cm).

Anticipated Results

In general, the yieldof purified oligonucleotides from denaturing PAGE decreases as the percentageof acrylamide increases. With crushed gel slices, an average yield of 50percent may be expected.

Greater recoveriescan be obtained by increasing the volume of elution solution added to thegel slice or by doing serial elutions from the same gel slice. Methodsemploying more active transfers (e.g., electroelution:Smith, 1980; Vorndam and Kerschner,1986) may give more efficient recoveries. Also, samples (especiallylarge synthetic RNAs) which prove particularly refractory to elution withaqueous buffers may be eluted easily with 6 volumes of formamide (>5 hat room temperature) followed by a brief elution with an aqueous buffer(approximately 1 h). Isoamyl-alcohol may be used to concentrate the formamide-aqueousbuffer extracts to a convenient precipitation volume (Urbach, J., personalcommunication).


Time Considerations

It is usually mostconvenient to set up and run the gel on one day, elute the oligonucleotideovernight, then phenol extract and ethanol precipitate the sample the followingday. However, a deprotected oligonucleotide can be ready for molecularbiology applications in as little as 6 hr: set up and polymerization ofgel, 1 hr; running gel, 2 hr; fragment elution, 2 hr; product recovery,1 hr.


Literature Cited

Applied Biosystems,1984. Evaluation and Purification of Synthetic Oligonucleotides. UserBulletin, Issue #13.

Frank, R., Muller,D., and Wolff, C. 1981. Identification and suppression of secondary structuresformed from deoxyoligonucleotides during electrophoresis in denaturingpolyacrylamide gels. Nucl. Acids Res. 9:4967-4979.

Maniatis, T., Jeffrey,A., and deSande, H.V. 1975. Chain length determination of small double-andsingle-stranded DNA molecules by polyacrylamide gel electrophoresis. Biochemistry14:3787-3794.

Smith, H.O. 1980. Recoveryof DNA from gels. Meth. Enzymol. 65:371-379.

Vorndam, A.V. and Kerschner,J. 1986. Purification of small oligonucleotides by polyacrylamide gel electrophoresisand transfer to diethylaminoethyl paper. Anal. Biochem. 152:221-225.

Chen Z. and RuffnerD.E. 1996. Modified crush-and-soak method for recovering oligodeoxynucleotidesfrom polyacrylamide gel. Biotechniques. 21:820-822.