ANALYTICAL ‘
`
`BIOCHEMISTRY
`
`1
`
`MTX1030
`
`

`

`ANALYTICAL BIOCHEMISTRY
`
`Volume 145, Number 1, February 15, 1985
`
`Copyright © 1985 by Academic Press, Inc.
`
`All Rights Reserved
`
`No part of this publication may be reproduced or transmitted in any form or by any means,
`electronic or mechanical, including photocopy, recording, or any information storage and retrieval
`system. without permission in writing from the copyright owner.
`The appearance ofthe code at the bottom ofthe first page ofan article in thisjournal indicates
`the copyright owner’s consent that copies of the article may be made for personal or internal
`use, or for the personal or internal use ofspecific clients. This consent is given on the condition,
`however, that the copier pay the stated per copy fee through the Copyright Clearance Center,
`Inc. (27 Congress Street, Salem, Massachusetts 01970), for copying beyond that permitted by
`Sections 107 or 108 of the U. S. Copyright Law. This consent does not extend to other kinds
`of copying, such as copying for general distribution, for advertising or promotional purposes,
`for creating new collective works, or for resale. Copy fees for pre-l985 articles are as shown on
`the article title pages; if no fee code appears on the title page, the copy fee is the same as for
`current articles.
`
`0003-2697/85 $3.00
`
`MADE IN THE UNITED STATES OF AMERICA
`\‘f
`
`
`ANALYTICAL BIOCHEMISTRY
`
`Published monthly (except semimonthly in February, May, August, and November) by Academic Press,
`Inc.. 111 Fifth Avenue, New York, N. Y. 10003
`
`1985: Volumes 144—151. Price: $740.00 U. S. A. and Canada; $857.00 all other countries
`All prices include postage and handling.
`
`Information concerning personal subscription rates may be obtained by writing to the Publisher. All
`correspondence and subscription orders should be addressed to the office of the Publishers at 111 Fifth
`Avenue, New York, N. Y. 10003. Send notices of change of address to the office of the Publishers at least
`6 to 8 weeks in advance. Please include both old and new addresses. POSTMASTER: Send changes of
`address to Analytical Biochemistry, 111 Fifth Avenue, New York, New York 10003.
`Second class postage paid at New York, N. Y. and at additional mailing offices.
`Copyright © 1985 by Academic Press, Inc.
`
`2
`
`

`

`High-Performance Anion-Exchange Chromatography of Oligonucleotides
`
`R. R. DRAGER AND F. E. REGNIER
`
`Department of Biochemistry, Purdue University, W. Lafayette, Indiana 47907
`
`Received August 3, 1984
`
`Several types of high—performance silica-based supports have been found to be effective in
`the separation of polynucleotides. The principal difference in these materials is the type of
`bonded phase and the method by which it is attached to the silica support. One approach is
`the coupling of stationary-phase groups to the surface through siloxane bonding. This technique
`is simple and produces a material of high capacity and resolution, but it sulfers from poor
`bonded-phase stability. An alternative approach is the adsorption of low-molecular-weight
`polyethylene imines (PEI) that are crosslinked into a surface film. The stationary phase is held
`in place by adsorption of the film at multiple sites. A previous report on this material showed
`the resolution of Oligonucleotides containing up to 30 bases. This paper reports further
`optimization of the PEI bonding chemistry in the preparation of HP—IEC columns for
`Oligonucleotides and tRNA species. Quaternization of the ion—exchange matrix was found to
`increase resolution of Oligonucleotides from 30 to 50 bases. The same support was found to be
`capable of resolving multiple tRNA species. Separations were achieved on small (0.42 X 5 cm)
`columns, using a 60- to 120-min ammonium sulfate gradient. The initial solvent was 15%
`acetonitrile in 0.05 M potassium phosphate (pH 5.9). The addition of 1 M ammonium sulfate
`© I985 Academic Press. Inc.
`to the initial solvent was used to prepare the final solvent.
`
`Fractionation of Oligonucleotides has been
`accomplished on a variety of chromato-
`graphic suppor’ts ranging from ion—exchange
`IEC)‘ through mixed mode to strict reversed-
`phase materials. Organic based anion-ex—
`change (AEC) packings such as Dowex-l (1)
`and diethylaminoethyl (DEAE) derivatives of
`Sephadex and cellulose (2—5) have been used
`for many years, but suffer from poor resolu-
`tion and recovery of Oligonucleotides in excess
`of 10 bases in length. Their low mechanical
`strength also requires that they be operated
`at
`low mobile phase velocities. With the
`development of rigid packing materials, high-
`performance liquid chromatography has be-
`come a major tool
`in oligonucleotide frac-
`tionation. Alkylamine derivitized (6,7) and
`DOIyethyleneimine coated [PEI (8,9)] silicas
`
`ion-exchange chromatog-
`'Abbreviations used: IEC,
`réphy; PEI, polyethyleneimine; BUDGE, 1,4-butane-
`d'OIdlglyCidyl ether; lPC, ion-pairing capacity.
`
`have been used successfully in anion-exchange
`separations of Oligonucleotides. Another an-
`ion-exchange material, RFC-5, has been pre-
`pared by adsorbing methyltrioctylamine onto
`particulate polychlorotrifluoroethylene beads
`(10—12). The hydrophobic character of both
`the support and the amine imparts charac-
`teristics that result
`in mixed-mode separa-
`tions. However, the coating of RFC-5 leaches
`from the support surface during operation.
`Reversed-phase separations of oligonucleo—
`tides have been achieved on both uncoated
`
`polychlorotrifluoroethylene beads (13) and
`alkyl silane-derivatized porous silica (14,15).
`The above three types of oligonucleotide
`separations were recently compared (16). For
`short Oligonucleotides (up to 10 bases) all
`three modes (anion—exchange, reversed-phase,
`and mixed—mode) were equally good. For
`larger Oligonucleotides anion exchange was
`the superior mode. However,
`the relative
`chemical instability of the aminoalkyl silane
`
`47
`
`0003-2697/85 $3.00
`Copyright © I985 by Academic Press. Inc.
`All rights of reproduction in any form reserved,
`
`
`
`3
`
`

`

`DRAGER AND REGNIER
`
`anion—exchangesupport materials used in the
`study was considered a significant disadvan—
`
`The pellicular PEI anion-exchange supports
`(8,9) have some advantages in that they are
`more stable than silane-derivatized materials,
`
`exhibit high loading capacity, and can resolve
`oligonucleotides differing by one base in
`length up to a total of 30 bases (8). A slightly
`different form of the PEI—bonded phase sup-
`
`port has also been successfully applied to
`protein fractionation (17). Recently,
`it was
`shown that quaternization of the PEI anion-
`exchange material, by methylation with
`iodide,
`improves protein resolu-
`
`This paper examines the effect of quater-
`nization on the chromatographic properties
`of PEI-bonded phases in the fractionation of
`oligonucleotides.
`
`MATERIALS
`
`Hypersil was purchased from Shandon
`Southern Products Limited (Sewickley, Pa.)
`LiChroma stainless-steel, precision—bore col-
`umns (0.41 X 5.0 cm) fitted with 0.5 pm frits
`were obtained from Anspec (Warrenville,
`111.). Polyethyleneimine (PEI-6) was purchased
`from Polysciences (Warrington, Pa.)
`1,4—
`Butanedioldiglycidyl ether
`(BUDGE) was
`purchased from Aldrich (Milwaukee, Wisc.).
`Methyl iodide was obtained from Columbia
`Organic Chemical Company, Inc. (Columbia,
`S. C.). 1,4-Dioxane was from Fisher (Spring—
`field, N. J- )- HPLC-grade dichloromethane,
`methanol, and acetonitrile were from J. T.
`Baker (Phillipsburg, N.
`J .). Triethylamine
`was obtained from Malinckrodt (Paris, Ky.).
`Picric acid was purchased from MCB (Nor-
`wood, Ohio). Oligodeoxyadenylates and oli-
`godeoxythymidylates were gifts from P. L.
`Biochemicals (Milwaukee, Wisc.). Heteroge-
`neous oligonucleotides were gifts from Ap-
`Plied Biosystems (Foster City, Calif). Transfer
`
`EXPERIMENTAL
`
`Support preparation. A modification of
`previous methods (8,17,19,20) was used to
`prepare silica-based, PEI-coated anion-ex-
`change supports. Hypersil (3 pm) silica was
`used exclusively. Support synthesis was ac-
`complished in three steps: (i) adsorption of
`PEI to silica, (ii) crosslinking the adsorbed
`PEI to form a stable polymeric coating, and
`(iii) methylation of the adsorbed, crosslinked
`polyamine.
`PEI was adsorbed from a 1% (w/v) solution
`in methanol. Suspension of silica in the PEI
`solution (10 ml PEI solution/g dry silica) was
`accomplished by brief sonication under vac-
`uum. The suspension was allowed to settle
`for 20 min at room temperature and then
`
`filtered to dryness on a sintered glass funnel.
`Thickness of the PEI coating could be reduced
`to varying extents by washing with appropri-
`ate volumes of methanol at this point (17).
`Adsorbed PEI was crosslinked on silica in
`
`a 10% (v/v) solution of BUDGE in 1,4-
`dioxane (10 ml BUDGE solution/g dry silica).
`Coated silica was suspended in the crosslink-
`ing solution by brief sonication, after which
`it was allowed to settle overnight (12—18 h)
`at room temperature to complete the cross-
`linking reaction. The support was washed
`with methanol
`(500 ml/g dry silica) and
`filtered to dryness on a sintered glass funnel.
`Methylation of the amine-coated silica was
`achieved with an excess of 20% (v/v) methyl
`iodide in methanol (10—20 ml/g silica). The
`degree of methylation was varied by control-
`ling reaction time. A 500 umol N/g PEI
`support was methylated to the extent that 40
`to 50% of the ion-pairable amines were made
`quaternary in a 5-min reflux. All methylation
`reactions were carried out at 49°C, under
`reflux. Following methylation, the quaternary
`amine—coated silica was washed extensively
`with methanol (300 ml/g dry silica), acetone
`(300 ml/g dry slica), deionized water (300
`ml/g dry silica), and, finally, methanol (100
`
`4
`
`

`

`psi with a pneumatic pump (Haskel, Burbank,
`Calif.) as described by Pearson and Regnier
`(8). Approximately 150 to 170 ml of solvent
`was pumped through columns by the packing
`pump.
`Chemical analysis of supports. Dry anion-
`exchange supports were assayed for ion-pair-
`ing capacity (IPC) by the picric acid method
`of Alpert and Regnier (17). Duplicate assays
`diflered by 2% or less.
`The extent of methylation, expressed as
`the percent of ion-pairable amine quaternized
`(% QUAT), was calculated as
`
`% QUAT
`
`__ (IPC)unmcthylatcd _ (IPC)mclhylated
`(IPC)unmethylated
`
`X 100.
`
`Under the conditions of this assay (in dichlo-
`romethane) picric acid does not ion pair with
`quaternary amines.
`While halide ions can interfere with the
`
`formation of the amine—picric acid ion pair
`in the IPC assay (21), washing of the amine—
`coated silica with 0.01 M NaOH eliminates
`
`the problem. Halide interference in the IPC
`assays of the materials used in this study was
`not evident.
`IPC assays of materials not
`washed with NaOH were in agreement with
`IPC assays of the same materials washed
`with NaOH.
`
`Carbon, nitrogen, and silicon analyses were
`performed by H. D. Lee, Chemistry Depart-
`ment, Purdue University. Accuracy was
`:0.3%.
`
`in
`Packing efiiciency. Packing efficiency,
`terms of number of theoretical plates (N),
`was calculated as
`
`N = 5.54(TR/ATR,,2)2,
`
`where TR is retention time of an unretained
`species and ATR”,
`is the half-height peak
`width for the unretained species (21). The
`chromophore used in these analyses was so-
`dium nitrite. Sodium nitrite was not retained
`
`on these columns when operated isocratically
`
`counts are the averages of five measurements.
`Chromatography. HPLC separations were
`performed on a Varian 5500 pumping system
`equipped with either an Altex Model 153
`fixed-wavelength (254 nm), uv detector (Altex
`Scientific Inc., Berkeley, Calif), or a Spectro-
`flow 773 variable-wavelength detector (Kratos
`Analytical Instruments, Ramsey, N. J .). Data
`were collected on a Linear Model 555 re-
`
`Instruments Corp. Reno,
`(Linear
`corder
`Nev.). The system was fitted with a Valco
`Model C6U injector with a lOO-pl injection
`loop (Valco Instruments C0,,
`Inc., Hous-
`ton, Tex.).
`Mobile phases. A binary gradient was em-
`ployed for all separations. Buffer A was 0.05
`M potassium phosphate, 15% (v/v) acetoni—
`trile, pH 5.9. Buffer B was 1.0 M ammonium
`sulfate, 0.05 M potassium phosphate, 15% (v/
`v) acetonitrile, pH 5.9, unless otherwise noted.
`The gradient rate was 0.25%/min at a flow
`rate of 0.5 ml/min, unless otherwise noted.
`Buffers for tRNA separations were 5% (v/v)
`acetonitrile; all other parameters were as
`described above. Gradients were as indicated
`
`in the figure legends.
`Sample preparation. Oligodeoxyadenylates
`and oligodeoxythymidylates were dissolved
`in buffer A (1.0 A254 U/100 pl) and used
`directly. Oligomers of DNA were identified
`by coelution with purified DNA Oligomers
`of specific length. Heterogeneous oIigonucle—
`otides were solubilized and used as described
`
`for Oligodeoxyadenylates, except the samples
`from Applied Biosystems, which were ana-
`lyzed as supplied (in solution). Transfer RNA
`was dissolved in buffer A [5% (v/v) acetoni-
`trile] to 1 mg/ml for use.
`Radioactive labeling and electrophoresis.
`The heptadecanucleotide (dTGAAGAATT-
`CGGCGTTT) was end labeled with 32F and
`analyzed by polyacrylamide gel electropho-
`resis, after chromatography, according to
`standard methods.
`
`Comparative oligonucleotide
`Columns were compared on the basis of
`
`
`
`5
`
`

`

`DRAGER AND REGNIER
`
`relative resolution of a standard test mixture
`of deoxyadenylate oligomers of 12 to 18
`in length. This average resolution
`(RS(,2_13)) was calculated as described by
`Pearson and Regnier (8). Average resolution
`values for columns 1 and 2 were calculated
`as described above but
`for a mixture of
`
`deoxythymidylate oligomers of 19 to 24 bases
`in length. Error limits on average resolution
`values are 150%.
`
`RESULTS AND DISCUSSION
`
`Table 1 describes the properties of all the
`anion-exchange materials used in this study.
`The material used in column 2 was derived
`from the material used in column 1. The
`material used in column 4 was derived from
`the material used in column 3. The materials
`used in columns 6—9 were derived from the
`
`material used in column 5. In every case,
`except for column 6, the packing efficiency
`(N) for unmethylated materials (columns 1,
`3, 5) was greater than or equal to same for
`the respective methylated columns. Improve-
`ments in chromatographic performance for
`the methylated columns therefore reflect dif-
`ferences in silica coating and not packing
`efficiency. The elemental analyses for col-
`umns 3 and 4 show that, within experimental
`
`limits, methylation of the PEI coating in-
`creased the amount of carbon on the material
`
`without changing the amount of nitrogen
`present. This was as expected.
`Initially, IEC of oligonucleotides on PEI
`supports was accomplished with a mobile
`phase that was 30% (v/v) methanol (8).
`It
`has since been found that the use of 15% (v/
`v) acetonitrile in the mobile phase improves
`the performance of PEI supports in oligo-
`nucleotide separations, relative to methanolic
`mobile phases (22). The performance of these
`mobile phases is compared directly in terms
`of the fractionation of oligodeoxyadenylates
`0f 40 to 60 bases in length on column 2
`(Table 1, Fig. 1).
`We observed that quaternization of a PEI
`support produced a 15% to 20% increase in
`theaverage resolution (R3) of DNA homo-
`polymers of 10 to 20 bases in length as
`reflected by columns 5 and 6 (Table 1).
`Subsequently, we observed that quaterniza-
`tion of PEI supports extended the fraction-
`ation range in oligonucleotide separations
`from the 30-base (8) to the 50-base range
`(Fig. 2).
`Chromatographic performance of the qua-
`ternized and unquaternized supports was
`compared directly (Fig. 3). The comparison
`was in terms of selectivity (Trim — TR") and
`
`TABLE 1
`
`CHROMATOGRAPHIC AND CHEMICAL PROPERTIES OF ANION-EXCHANGE MATERIALS
`
`1pc
`
`
`(umolN/g support)
`% QUAT.
`% C
`% N
`% Si
`R5"
`N
`
`465
`0.98
`—
`—
`—
`0
`583
`450
`1.35
`—
`—
`—
`58.8
`240
`566
`1.98
`42.95
`1.03
`3.34
`0
`514
`566
`2.92
`40.68
`0.93
`3.79
`57.5
`2l8
`668
`1.42
`—
`—
`—
`0
`480
`945
`1.85
`—
`—
`—
`30
`337
`581
`2.07
`—
`—
`—
`44
`270
`569
`1.61
`—
`—
`—
`51
`237
`654
`1.42
`—
`—
`—
`74
`124
`
`
`
`
`
`
`57.5 3.79 0.93 40.68 —218 3431 ’7
`
`6
`
`

`

`.OI5
`.010
`.005
`
`A254
`
`TIME (min)
`
`FIG. 1. A comparison ofthe 15% acetonitrile (A) and
`30% methanol (B) mobile phases, in the fractionation of
`40- to 60-base oligodeoxyadenylates on column 2. The
`gradients were 0.5%/min, beginning at 0%. B. Flow rates
`were 0.5 ml/min. The injection volumes were 50 [11.
`
`average peak width [(ATR” + ATR,,H)2], where
`TR" is the retention time of an oligonucleotide
`of n bases and ATR,I is the peak width of the
`same oligonucleotide. TR.“ and ATR"+I are
`the corresponding retention time and peak
`width of an oligonucleotide one base longer.
`This comparison showed that quaternized
`PEI supports improve resolution and frac—
`tionation range in oligonucleotide separations
`
`maintaining these improvements over a large
`oligonucleotide size range.
`The data in Fig. 3 were taken from one
`separation on column 3 (Table l), and one
`separation on column 4 (Table l). The elution
`gradient in both cases was begun at 0% buffer
`B. This is not optimal for the fractionation
`of larger oligonucleotides on either column.
`It was observed that better resolution was
`
`obtained in the 40- to 50-base range when a
`smaller gradient window was used (Fig. 2).
`Suboptimal conditions, and therefore poorer
`resolution, were necessary in Fig. 3 to permit
`a comparison of supports under identical
`conditions. Peaks of 19 to 24 base oligonu-
`cleotides on column 2 (quaternized) were
`relatively small, making accurate peak width
`measurements difficult. This favors overesti-
`
`mation of peak width and underestimation
`of resolution.
`
`The effect of mobile phase pH on the
`resolution of oligodeoxyadenylates on an un-
`quaternized PEI support was studied to es-
`tablish whether the observed effects of qua-
`ternization could be explained simply in
`terms of increased charge density on the PEI
`surface (Table 2). We conclude from these
`data that the etTect of quaternization is not
`solely due to increased charge density. The
`
`
`
`
`
`O
`
`20
`
`4O
`
`60
`TIME (minutes)
`
`80
`
`'
`IOO
`
`06
`
`5l—l
`m
`05 8
`(\l
`A
`f04 Z
`[—1
`
`03
`
`02
`
`002
`
`ct
`'0
`N
`<
`
`001
`
`O
`
`FIG. 2. Fractionation of 40- to 60—base oligodeoxyadenylates on column 2 (Table l). The gradient was
`begun at 20% buffer B, other elution conditions were as described under Experimental. The injection
`volume was 75 #1.
`
`
`
`7
`
`

`

`DRAGER AND REGNIER
`
`I
`fi— 1— 1
`fi—
`
`I4
`
`.2
`
`A
`o_
`V)
`3 IO
`1’
`
`E.
`
`:
`
`8
`
`’i‘
`CE
`3 6
`r;
`
`4
`
`2
`
`0
`
`1
`
`DIl.
`
`.’
`,’
`,
`I
`
`
`A —— — UNQUAT.
`QUAT
`
`
`
`
`I4
`
`.2
`
`A
`I
`d
`U
`'0 Z
`s
`
`s A.
`T
`
`gr“
`:
`6 :1 N
`+9
`S
`
`4
`
`2
`
`o
`45
`
`
`—l
`40
`
`10
`
`IS
`
`20
`
`25
`BASES,
`
`39‘
`(n)
`
`35
`
`1 and 2 respectively (Table
`FIG. 3. A graphic comparison of weak and strong anion-exchange columns.
`1), in terms of selectivity (TmH — TR"), and peak width [(ATR,,+1 + TRM)/2]. Where the solid lines cross,
`unit resolution on column 2 (quaternized) is shown. Where the broken lines cross, unit resolution on
`column 1 (unquaternized) is shown.
`
`data also reaffirm that the mobile phase pH
`previously established by Pearson and Regnier
`(8) and Lawson et a].
`(9)
`is optimal
`for
`oligonucleotide fractions on unquaternized
`PEI supports. This study was done with
`column 5 (Table 1). The mobile phase con-
`ditions were as in Fig. 4, except that the pH
`was varied as indicated. Resolution values
`
`TABLE 2
`
`EFFECT OF MOBILE PHASE pH ON AVERAGE
`RESOLUTION (RS) OF OLIGODEOXYADENYLATES
`
`(12—18 BASES)
`
`
`DH
`RS(l2—18)
`
`4.0
`4.5
`5.0
`5.5
`6.0
`6.5
`
`1.45
`1.51
`1.52
`1.62
`1.78
`1.63
`
`are averages as described under Experimental.
`The resolution values in Table 2 are high
`relative to those in Table 1 because the
`
`oligodeoxyadenylates used for Table 2 had
`the terminal phosphate groups removed. Re-
`moval of terminal phosphate groups increases
`the relative difference in anionic character
`
`between one oligonucleotide and the next
`larger oligonucleotide in a homologous series.
`The extent of quaternization profoundly
`influenced oligonucleotide resolution. The
`simple crosslinked polyamine support used
`in column 3 (Table l) was methylated to the
`extent that 57.5% of the ion-pairable amine
`was quaternized, and the quaternized material
`was used in column 4 (Table
`1). The
`R‘S(lz,|g) value obtained with column 4 was
`47% greater than that of column 3 (Table 1,
`Fig. 4). A more extensive study of the influ-
`ence of quaternization was carried out on
`the support materials used in columns 5—9
`
`8
`
`

`

`
`
`
`
`A254
`
`0005
`
`A254
`
`
`
`
`
`
`
`
`
`TIME (minutes)
`
`FIG. 4. Fractionation of 12- to 18-base oligodeoxyadenylates on (A) column 3 and (B) column 4 (Table
`l). Elution conditions were as described under Experimental. The injection volumes were 12 ul (A) and
`m»
`25 )1] (B).
`
`Rsuz—m values for these columns (Table 1)
`indicate that a moderate extent of quaterni-
`
`zation (44%, column 7) produced the best
`oligonucleotide resolution.
`The separation of oligodeoxythymidylates
`was also performed as an additional basis for
`comparison of quaternized and unquater—
`nized columns (Fig. 5). The quaternized col—
`umn 2 improved the average resolution for
`this mixture by 38% over the unquaternized
`column 1
`(Table 1). The support used in
`column 2 was obtained by methylation of
`the support used in column 1.
`A mixture of heterogeneous oligonucleo-
`tides from the synthesis of the octadecamer
`d(TCACAGTCTGGTCTCACT) was
`frac-
`tionated on column 4 (Fig. 6). The 5’-hy-
`droxyl of the octadecamer was blocked with
`
`a trityl group. A decrease of 6 min in the
`retention time of the detritylated octadecamer
`indicated that, even with a mobile phase
`containing 15% (v/v) acetonitrile, some sig-
`nificant hydrophobic interaction occurs on
`the quaternized PEI surface. The heteroge-
`neous octadecamer was retained some 30
`
`min longer than the homogeneous octadecyl
`adenylate on column 4 under the same con—
`ditions (Figs. 4 and 6). This suggests that
`sequence heterogeneity strongly influences
`anion—exchange chromatography of oligonu-
`cleotides.
`
`A mixture of heterogeneous oligonucleo—
`tides from the synthesis of the heptadecanu-
`cleotide (dTGAAGAATTCGGCGTTT) was
`fractionated on column 10 (Fig. 7A). The
`purity of the product peak was determined
`
`9
`
`

`

`DRAGER AND REGNIER
`
`A254
`
`TIME (min)
`
`to 24—base oligodeoxy-
`FIG. 5. Fractionation of 19-
`thymidylates on (A) column 2 and (B) column 1 (Table
`l). Elution conditions were as described under Experi-
`mental. The injection volumes were 50 pl.
`
`by polyacrylamide gel electrophoresis (Fig.
`7B). In a single chromatographic pass,
`the
`product was obtained at greater than 90%
`purity. Sequence determination verified the
`
`identity of the product (data not shown). The
`purified heptadecanucleotide was used suc-
`cessfully in a “site-specific” mutation exper-
`iment.
`
`Having tested the quaternized PEI mate-
`rials in the fractionation of 10— to 50— base
`
`oligonucleotides, we decided to attempt, with
`the same materials,
`the fractionation of
`tRNA, the next larger class of nucleic acid.
`The tRNA fractionation achieved on column
`
`2 (Table 1, Fig. 8) is comparable to those by
`reversed—phase
`(23) and by mixed-mode
`HPLC (24). When this separation was at-
`tempted with the mobile phase containing
`15% (v/v) acetonitrile, the peaks evident in
`Fig. 7 were poorly resolved (data not shown).
`Peak width and retention time increased as
`the concentration of acetonitrile in the mobile
`
`phase was decreased (data not shown). Res-
`oliition was best
`(as in Fig. 8) when the
`mobile phase contained 5% (v/v) acetonitrile.
`Because of the length (75—90 bases) and
`number (30) of tRNA species obtainable
`from Escherichia coli (25), we did not expect
`the separation to work as well as it did. The
`
` '—l_
`
`in1i!Wm"M‘00,3
`
`8[magmaM
`
`
`I
`g
`I _—I__——_—.L___l_‘— O
`20
`4O
`60
`BO
`IOO
`IZO
`TIME (minutes)
`
`OO
`
`FIG. 6. Fractionation of a mixture of oligonucleotides from the synthesis of the octadecanucleotide
`
`10
`
`

`

`
`
`.OZI
`
`5
`&V.OI4
`
`.007
`
`.3
`
`.1 g
`
`.23:O?U)
`”“
`IV
`
`—BPB
`
`0
`
`4O
`
`TIME (min)
`
`80
`
`IZO |38
`
`FIG. 7. Fractionation of a mixture of heterogeneous oligonucleotides from the synthesis of the
`heptadecanucleotide (dTGAAGAATTCGGCGTTT) on column 10 (Table l) (A), and polyacrylamide gel
`electrophoresis of the product peak (B). Elution conditions were as described under Experimental. The
`injection volume was 90 pl. The sample contained >50 A260 U/ml. (Tracking dyes: XYL, xylene cyanol;
`BPB, bromphenol blue).
`
`005
`
`0O4
`
`003
`
`A254 002
`
`OOI
`
`004
`
`003
`
`A254
`
`002
`
`
`
`
`
`TIME (minutes)
`
`FIG. 8. Fractionation of tRNA from Escherichia coli on column 2 (Table l). The gradient was begun
`at 50% Buffer B; other elution conditions were as described under Experimental. The injection volume
`Was 25 [.11.
`
`11
`
`11
`
`

`

`(1983) J.
`
`DRAGER AND REGNlER
`
`fact that the separation was significantly im-
`proved by reducing the mobile phase aceto-
`nitrile concentration to 5% (v/v) suggests that
`the three-dimensional shape of the tRNA
`(reflecting secondary and tertiary structure)
`may be important in anion—exchange chro- .
`matography of the species. This would be
`consistent with the importance of three-di-
`mensional structure observed in surface me-
`
`diated separations of proteins (17,26).
`
`CONCLUSIONS
`
`anion-exchange
`Quaternization of PEI
`supports improves the resolution and frac-
`tionation range for oligonucleotide separa-
`tions in the 10- to 60-base size range. The
`extent of quaternization significantly afiected
`resolution. Optimal performance from the
`PEI supports examined here was obtained
`when 40 to 60% of the ion-pairable amines
`on the surface were quaternized. The conse-
`quence of this modification,
`in chromato-
`graphic terms,
`is that both peak width and
`selectivity are enhanced in oligonucleotide
`fractionation. The quaternized PEI supports
`are capable of fractionating a broad range of
`nucleic acids,
`including homo- and hetero—
`geneous nucleic acid oligomers and trans-
`
`ACKNOWLEDGMENTS
`
`We are deeply grateful to H. Zalkin and P. Mantsala
`for the electrophoretic analysis of the heptadecanucleotide
`(Figure 7). We thank H. L. Weith and T. G. Lawson for
`their many helpful discussions. We thank Ms. Camille
`Alexander for her clerical assistance. This work was
`supported by NIH Grant GM25431. This is Journal
`Paper NO- 9969 from the Purdue University Agricultural
`Experiment Station.
`
`REFERENCES
`
`1. Asteriadis, G. T., Armbruster, M. A., and Gilham,
`G. T. (1976) Anal. Bloc/rem. 70, 64,
`
`2. Tener, G. M. (1967) in Methods in Enzymology
`(Grossman, L., and Moldave, K., eds.), Vol. 12,
`pp. 398, Academic Press, New York.
`3. Staehlein, M. (1963) Progr. Nucl. Acid Res. 2, 169.
`:5
`Schott, H. (1982) J. C/iromaiogr. 237, 429.
`5. Satava, J., Mikes, G., and Strop, P. (1979) J. Chro-
`. malagr. 180, 31.
`6. Jost, W., Unger, K. K., Lipecky, R., Gassen, H. G.
`(1979) J. C/Iromalogr. 185, 403.
`7. Gait, M. J., and Sheppard, R. C. (1977) Nucl. Acids
`Res. 4, 1135.
`8. Pearson, J. D., and Regnier, F. E. (1983) J. Chro—
`matogr. 255, 137.
`9. Lawson, T. G., Regnier, F. E., and Weith, H. L.
`(1983) Anal. Biochem. 133, 85,
`10. BischolT, R., and McLaughlin, L. W.
`Chromatogr. 270,
`l 17.
`11. Wells, R. D., Haldies, S. C., Horn, G. T., Klien, B.,
`Larson, J. E., Nevendorf, S. K., Panayotatos, N.,
`Patient, P. K. (1980) in Methods in Enzymology
`(Grossman, L., and Moldave, K., eds), Vol. 65,
`Part I, pp. 2367, Academic Press, New York.
`12. Product Profile (1982) Cat No. 1501, Bethesda Re-
`search Laboratories, Inc., Rockville, Md.
`13. Usher, D. A. (1979) Nucl. Acids Res. 6, 2289.
`14. McFarland, G. D., and Borer, P. N. (1979) Nucl.
`Acids Res. 7, 1067.
`15. Edge, M. D., Greene, A. R.,Heathc1ifl”, R., Meacock,
`P. A., Schuch, W., Scanlon, D. B., Atkinson,
`T. C., Newton, C. R., and Markham, A. F. (1981)
`Nature (London) 292, 756.
`16. Haupt. W., and Pingoud, A. (1983) J. Chromatogr.
`260, 419.
`17. Alpert, A. J., and Regnicr, F. E. (1979) J. Clzronmiogr.
`185, 375.
`18. Kopacieweicz, W,, and Regnier, F. E. (1983) Anal.
`Bloc/rem. 133, 251.
`19. Vanecek, G., and Regnier, F. E.
`Biochem. 109, 345.
`20. Vanecek, G., and Regnier, F. E.
`Biochem. 121, 156.
`21. Pfannkoch, E., and Regnier, F. E., unpublished
`results.
`22. Lawson, G. T., PhD. Thesis, Purdue University
`(1984).
`23. Pearson, J. D., Mitchell, M., and Regnier, F. E.
`(1983) J. Liq. Chromalogr. 6, 1441.
`24. Pearson, R. L., Weiss, J. F., and Kelmers, A. D.
`(1971) Biochem. Biop/Iys, Acta 228, 770.
`25. Saenger, W. (1984) Principles ofNucleic Acid Struc—
`ture, Springer-Verlag, New York/Berlin.
`26. Brautigan, D. L., Ferguson-Miller, S., and Margoliash,
`E. (1978) J. Biol. Chem. 253, 130.
`
`(1980) Anal.
`
`(1982) Anal.
`
`12
`
`