216
`
`NOTES & TIPS
`
`High-Resolution Preparative-Scale Purification
`of RNA Using the Prep Cell‘
`
`Peristaltic
`
`Pump
`
`UV
`9 .
`monitor
`
`Chart Recorder
`
`Digitizer
`
`Tu H. Nguyen,2 Lynette A. Cunninghamf"
`Kendra M. Hammond, and Yi Lu3
`
`Department of Chemistry, University of Illinois.
`Urbana. Illinois 61801
`
`Received December 2, 1998
`
`Milligram-scale purification of RNA with high reso-
`lution is required for spectroscopic and X—ray crystal-
`lographic characterizations, as well as for clinical tri-
`als. The presence of many conformations of the same
`RNA sequence makes it particularly difficult to purify
`RNA using column chromatography. Therefore, dena-
`turing gel electrophoresis is commonly used for RNA
`purification (1, 2). However, most gel electrophoresis
`methods suffer low capacity, requiring 8—15 gels for a
`typical spectroscopic or X-ray structural experiment
`(3). We previously reported a method for automated
`large-scale purification of an RNA ribozyme from other
`transcription components using the Bio-Rad Prep Cell
`(4). Here we test a new model of the Prep Cell (Fig. 1)
`which is three times longer than the original Model 491
`apparatus and also investigate the sample recovery of
`the Model 491 apparatus. We found that the new ap-
`paratus offers significant improvement in resolution
`and loading capacity, and that the sample recovery of
`~90% for Model 491 apparatus is better than that
`typically obtained from either electroelution or crush-
`and-soak methods. This method can be easily adapted
`to large-scale purification of other nucleic acids.
`
`Materials and Methods
`
`A 100-ml-scale transcription was performed as pre-
`viously described to obtain milligram quantities of the
`34-mer 5'-GGCGACCGUGAUGAGGCCGAAAGGC-
`
`CGAAACAUU-3’ (4). The crude transcript was etha-
`nol-precipitated and reconstituted in 5 ml of 1.5x TBE“
`(135 mM Tris—borate, 3 mM EDTA). The sample was
`then concentrated using Centricon-lO units (Amicon,
`Beverly, MA) to a final volume of 1—2 ml. During this
`concentration procedure, the centricon was washed
`several times with 1.5x TBE to partially remove un-
`
`' This research was supported by the NIH FIRST Award
`(GM53706) and the Donors of the Petroleum Research Fund, admin-
`istered by the American Chemical Society.
`2 Indicates an equal contribution to this work.
`3 To whom all correspondence should be addressed at Department
`of Chemistry, University of Illinois, Box 8-6 Chemical and Life Sci-
`ences Laboratory, Urbana, IL 61801. E-mail: yi—lu@uiuc.edu.
`‘ Abbreviation used: TBE, Tris— borate—EDTA.
`
`Analytical Biochemistry 269. 216 ~218 (1999)
`Article ID abio.1999.4030
`0003-2697/99 $30.00
`Copyright © 1999 by Academic Press
`All rights of reproduction in any form reserved.
`
`(_ )
`
`Fraction
`\ Collector
`
`Gel mflce
`
`Migrating RNA
`
`Bela Prep Cd!
`
`FIG. 1. Schematic diagram of the modified Prep Cell. This model is
`identical to the Model 491 except the lower chamber has been elon-
`gated to accommodate a longer gel tube. The RNA is loaded onto the
`cylindrical PAGE gel where it migrates down the gel toward the
`positive electrode (see the arrows), just as in traditional electro«
`phoresis. The separated RNA species are then pulled through the
`capillary tubing located in the center of the cooling core by a peri-
`staltic pump into a UV monitor and finally into a fraction collector.
`A cellulose membrane with a molecular weight cutoff below that of
`the RNA is placed at the bottom of the Prep Cell gel to prevent the
`purified RNA from escaping the system while still allowing passage
`of conducting ions. A digitizer was also added to allow collection of a
`computerized chromatogram in addition to the hardcopy chromato-
`gram generated by the strip chart recorder.
`
`incorporated NTPs and thus increase the solubility of
`the transcript.
`Purification of the RNA transcripts was performed
`using the Model 491 Prep Cell and prototype Prep Cell
`along with equipment donated by Bio-Rad Laborato-
`ries, including the Model 1327 Econo-Recorder, EM-l
`Econo UV Monitor, Model 2128 Fraction Collector,
`EP-l Econo Pump, and the Powerpac 1000 power sup-
`ply. In addition to the chart recorder, the data were
`simultaneously digitized using a ComputerBoard
`DAS-08 digitizer. A 20% acrylamide/S M urea gel (120
`ml for a 13-cm gel, 260 ml for a 30-cm gel) was pre-
`pared in the 37-min i.d. large Prep Cell gel tube and
`allowed to polymerize for 3 h. During the polymeriza-
`tion process, the gel was cooled using the recirculation
`pump connected to the cooling core and a l-liter beaker
`of ice-water. To prevent crystallization of the urea dur-
`ing the cooling process, room-temperature water was
`passed through the pump until polymerization began
`and the gel started to get warm, at which point ice was
`added to the beaker. For each run, 130 pl of crude
`transcript was combined with an equal volume of for-
`mamide, heat denatured, and loaded onto the Prep Cell
`gel. We found that increasing the power to 15 W al-
`lowed for a shorter running time without loss in reso-
`lution. Therefore, the following separation conditions
`were used: 20% denaturing polyacrylamide, 15 W con-
`
`CU REVAC EX2020
`CUREVAC EX2020
`Page 1
`Page 1
`
`

`

`NOTES & TIPS
`
`217
`
`l l
`
`iI
`
`
`
`0.05 i"
`
`-
`
`-
`
`~
`
`—- ~
`
`0.04 4
`
`0.03 f
`
`0.02 7
`
`0.01 -
`
`Absotbanceat260nm
`
`30 cm
`
`0.00 .
`
`0
`
`100
`
`200
`
`300
`
`400
`
`Relative time (minutes)
`
`3":~.ge-aau=: :-
`".1“
`
`B
`
`(A) Overlaid chromatograms from purifications using 13-
`FIG. 2.
`and 30-cm Prep Cell gels. Because the elution time differs for the two
`gel lengths, the origin of the time axis shown corresponds to a point
`just before peak elution and does not correspond to the start of each
`run. The origin corresponds to 15.5 h (930 min) and 47.5 h (2850 min)
`for the 13- and 30-cm runs, respectively. (B) Analytical PAGE gel of
`Prep Cell peak fractions from a 30-cm Prep Cell run in which an RNA
`transcript containing 7 mg of 34- and 35-mer hammerhead ribozyme
`was loaded. Aliquots of every other fraction from the RNA peak were
`loaded onto a 19 cm X 29 cm X 0.7 mm gel and stained with ethidium
`bromide.
`
`is loaded. This method requires a significantly lower
`amount of acrylamide gel solution (~260 ml for the 30
`cm Prep Cell) than a typical preparative PAGE gel
`(~720 ml for a 40 X 60 X 0.3 cm gel), needs no attended
`operation once the sample is loaded, and requires min-
`imal postelectrophoresis manipulation. Unlike the
`crush-and—soak (6), electroelution (5), or ultracentrifu-
`gation methods, the gel is not destroyed and can be
`used up to three times without significant loss of res-
`olution. These features make the Prep Cell method a
`viable alternative to traditional purification methods,
`especially for laboratories that need large quantities of
`RNA on a routine basis. This method has also been
`
`adapted to high-resolution purification of milligram
`quantities of other nucleic acids, such as phosphoro-
`thioate DNA and RNA in our laboratory.
`
`Acknowledgments. We thank J. J. Dunn and A. H. Rosenberg for
`plasmid pAR1219 containing the T7 RNA polymerase gene, Pascale
`Legault and Arthur Pardi for the transcription protocol, Eric Val-
`lender for technical assistance, and Mary Ann Ireland, Lauri Heerdt,
`and Linda Castle at Bio-Rad Laboratories for generously providing
`materials, equipment, and technical support. Y.L.
`is a Sloan Re-
`search Fellow of the Alfred P. Sloan Foundation, a Beckman Young
`Investigator of the Arnold and Mabel Beckman Foundation, and a
`Cottrell Scholar of the Research Corporation.
`
`stant power, 1.5x TBE running buffer, 1 ml/min elu-
`tion rate, and 8 ml/fraction.
`
`Results and Discussion
`
`Percentage sample recovery. To determine how
`much RNA can be recovered from the Prep Cell gel,
`duplicate runs were performed using the Model 491
`Prep Cell. For each run, 2—3 mg of purified 34-mer
`hammerhead ribozyme (quantified by UV absorption
`using 6250“,“ = 285,483 M‘1 cm“) was loaded onto a
`freshly prepared l3-cm, 20% PAGE Prep Cell gel. After
`completion of the run, the fractions containing RNA
`were combined, concentrated, desalted, and quantified
`by UV absorption. Of the 3.1 mg loaded in the first run
`and the 2 mg loaded in the second, 2.8 and 1.8 mg were
`recovered, respectively. The corresponding average re-
`covery is 90%. This recovery is better than that typi-
`cally obtained from both conventional electroelution
`(70—80%) (5) and crush~and-soak methods (60—80%)
`
`(6). The improved recovery is attributed to less post-
`electrophoresis manipulation.
`
`Effect ofgel height. The maximal gel height for the
`commercially available Model 491 Prep Cell is 13 cm,
`compared to 40 cm for most preparative PAGE gels. To
`investigate whether a longer gel would offer better
`resolution, we used a beta Prep Cell model with a
`maximal height of 40 cm, supplied by Bio-Rad Labora-
`tories (Hercules, CA). Figure 2A shows a comparison of
`chromatograms from purification runs using 13- and
`30-cm Prep Cell gels, each loaded with an RNA tran-
`script containing 0.8 mg of 34~ and 35-mer hammer-
`head ribozyme (as determined by quantifying these
`bands on an analytical PAGE gel prior to Prep-Cell
`purification). The operating conditions were constant
`for each run. As expected, better resolution is achieved
`with the 30-cm gel. Individual peaks corresponding to
`n — 1, n, n + 1, and n + 2 transcripts are visible. To
`obtain comparable resolution on the 13-cm gel, less
`than half as much crude transcript can be loaded (4).
`The Prep Cell can be used to purify even larger
`quantities of RNA. An RNA transcript containing ~7
`mg of 34- and 35-mer hammerhead ribozymes was
`loaded onto a 30-cm Prep Cell gel. Although resolved
`peaks corresponding to the 34- and 35-mer are not
`visible in the chromatogram, individual 8-ml fractions
`contain pure 34-mer RNA (Fig. 2B). Depending on the
`purity requirement of the particular application, those
`fractions containing less pure RNA can be pooled and
`repurified using the same Prep Cell gel to recover RNA
`of higher purity.
`In summary, the Prep Cell method provides high-
`resolution preparative purification of RNA with high
`percentage sample recovery. In addition, the longer
`apparatus has the advantage of allowing resolution of
`n from n + 1 transcripts even when several milligrams
`
`CU REVAC EX2020
`CUREVAC EX2020
`Page 2
`Page 2
`
`

`

`218
`
`REFERENCES
`
`NOTES & TIPS
`
`1. Milligan, J. F., Groebe, D. R., Witherell, G. W., and Uhlenbeck,
`0. C. (1987) Nucleic Acids Res. 15, 8783-8798.
`2. Heus, H. A., Uhlenbeck, O. C., and Pardi, A. (1990) Nucleic Acids
`Res. 18. 1103-1108.
`3. Heus, H. A., and Pardi, A. (1991) J. Mol. Biol. 217. 113-124.
`4. Cunningham. L., Kittikamron, K., and Lu, Y. (1996) Nucleic
`Acids Res. 24. 3647—3648.
`5. Zassenhaus, H. P., Butow, R. A., and Hannon, Y. P. (1982) Anal.
`Biochem. 125, 125—130.
`6. Grierson, D. (1982) in Gel Electrophoresis of Nucleic Acids: A
`Practical Approach (Rickwood, D.. and Hames, B. D., Eds), p. 11.
`IRL Press, Oxford.
`
`An Improved Method for the Purification
`
`of Large DNA Fragments from Agarose
`Gels Using Wizard Plus SV Columns
`
`Daniel Tillett and Brett A. Neilan1
`
`School of Microbiology and Immunology. University
`of New South Wales, Sydney, 2052, Australia
`
`Received December 8, 1998
`
`The isolation of DNA fragments from agarose gels is
`an integral step of many molecular biological protocols.
`Of the numerous techniques developed to recover DNA
`fragments from agarose gels (1), the direct elution of
`DNA from the agarose matrix by centrifugation
`through a filter is the simplest. Different filters have
`been used including cotton-filled pipet tips (2), glass
`wool (3), blotting paper (4), paper slurry (5), commer-
`cial barrier pipet tips (6), and Wizard minicolumns (7).
`While the Wizard minicolumn is a convenient and re-
`
`producible means to directly elute DNA from agarose
`gels, it suffers from two limitations as a filter. First, the
`standard Wizard columns are of low capacity and its
`small opening can make the insertion of the agarose gel
`slice awkward. Second, the DNA yield is often poor,
`particularly with large DNA fragments. This problem
`is, however, not confined to the use of Wizard colums
`because low yields of large DNA fragments have been
`observed with other filter systems (2—6).
`We describe the use of Wizard Plus SV miniprep
`DNA purification columns (Promega, Madison, WI) for
`the rapid isolation of DNA fragments from agarose
`gels. The yield of large DNA fragments is improved by
`preequilibrating the gel slices in a neutral salt buffer
`and freezing before centrifugation (8).
`Two 30-ng samples of A—HindIII-digested DNA
`marker were electrophoresed in parallel on a 0.7%
`
`‘To whom correspondence and reprint requests should be ad-
`dressed. Fax: 61 2 9385 1591. E-mail: b.neilan@unsw.edu.au.
`
`Analytical Biochemistry 269, 218—219 (1999)
`Article ID abio.1999.4006
`0003-2697/99 $30.00
`Copyright © 1999 by Academic Press
`All rights of reproduction in any form reserved.
`
`TABLE 1
`
`Efficiency of DNA Recovery from Agarose Using either the
`Original Wizard Direct Elution Method (7) or the Presented
`Improved Wizard Plus SV Salt/Freeze Method
`
`Fragment size (kb) Original method (%)
`
`Improved method (%)
`
`23
`9
`6.5
`
`33
`45
`40
`
`75
`85
`80
`
`agarose gel in 1X Tris—acetatehEDTA buffer (1). The
`gel was stained with ethidium bromide and the 23-, 9-,
`and 6.5—kb bands were excised from both lanes under
`
`UV transillumination. Individual DNA fragments were
`eluted using either of the following two protocols:
`
`l. The original Wizard DNA gel elution protocol of
`Wolf and Hull {7). Briefly, gel slices were placed
`in individual standard Wizard columns held in 1.5-ml
`
`Eppendorf tubes and the DNA was eluted by centrifuga-
`tion at 14,000g for 12 min. DNA was precipitated by the
`addition of 0.1 vol of 3 M sodium acetate and 1 vol of iso-
`
`propanol, followed by centrifiigation at 14,000g for 10 min.
`The Wizard Plus S V column with salt equilibration
`and freezing protocol. Gel slices were placed within
`individual 2-m1 Eppendorf tubes containing 1 ml of salt
`buffer (300 mM sodium acetate, 50 mM Tris—HCl, 1 mM
`EDTA, pH 8.3). The gel slices were allowed to equilibrate
`for 30 min at room temperature before the gel slice was
`transferred, with minimal buffer, to individual Wizard
`Plus SV miniprep DNA purification columns held in
`1.5-ml Eppendorf tubes. After freezing the gel slices for
`10 min at —70°C, the excised DNA fragment was eluted
`by centrifugation at 14,000g for 12 min. DNA was pre-
`cipitated by the addition of 1 vol of isopropanol followed
`by centrifugation at 14,000g for 10 min.
`DNA samples were resuspended in 10 p.l of TE (10
`mM Tris—HCI, pH 7.4; 1 mM EDTA, pH 8.0) and the
`DNA recovery quantified (Table 1) using the Fluro-S
`Multilmager (Bio-Rad, Hercules, CA) after electro-
`phoresis on a 0.7% agarose gel with 30 ng of A—HindIII
`DNA marker (Fig. 1).
`The Wizard Plus SV column, in combination with a
`salt preequilibration and freeze step, provides a reli-
`able method for the direct elution of large DNA frag-
`ments fi'om agarose gels. Their large capacity and
`opening make them particularly convenient when ex-
`cising large gel volumes. The addition of a salt buffer
`preequilibration and freezing step provides for a sig—
`nificant increased recovery of large DNA fragments
`(Table 1). We have isolated DNA fragments using this
`technique from 0.4 to 3% agarose. Finally, DNA puri-
`fied using this method has proven suitable for a range
`of molecular biological procedures, including plasmid
`preparation, DNA ligations, DNA sequencing, PCR
`amplification, and restriction enzyme digestions (9).
`
`CU REVAC EX2020
`CUREVAC EX2020
`Page 3
`Page 3
`
`