Q-DI 1993 Oxford University Press
`
`Nucleic Acids Research, 1993, Vol. 21, No. 5 1061-1066
`
`High-resolution liquid chromatography of DNA fragments
`on non-porous poly(styrene-divinylbenzene) particles
`
`Christian G.Huber, Peter J.Oefner1, Eugen Preuss2 and Gunther K.Bonn1
`Institute of Radiochemistry, Leopold-Franzens-University, Innrain 52a, A-6020 Innsbruck, 1Department
`of Analytical Chemistry, Johannes-Kepler-University, Altenbergerstrafle 69, A-4040 Linz and
`2Institute of Medicinal Chemistry and Biochemistry, Leopold-Franzens-University, Fritz-Pregl-StraBe 3,
`A-6020 Innsbruck, Austria
`
`Received January 11, 1993; Accepted February 2, 1993
`
`ABSTRACT
`DNA restriction fragments and PCR products were
`separated by means of ion-pair reversed-phase high-
`performance liquid chromatography on alkylated non-
`porous poly(styrene-divinylbenzene) particles with a
`mean diameter of 2.1 Am. Optimum resolution was
`obtained by using an acetonitrile gradient in 100 mM
`of triethylammonium acetate and a column temperature
`of 50 OC. This allowed the separation of DNA fragments
`differing in chain length by 1 - 5% up to a size of 500
`base pairs. PCR products could be analyzed directly
`in less than two minutes with a concentration
`sensitivity of at least 300 ng/ml. Compared with anion-
`exchange chromatography or gel electrophoresis no
`desaltation of the purified DNA molecules is required
`because the volatile buffer system can be readily
`evaporated. Subsequently, the method was used for
`the semiquantitative evaluation of the expression of
`multidrug resistance genes in mononuclear white blood
`cells.
`
`INTRODUCTION
`Restriction enzyme mapping and polymerase chain reaction have
`gained tremendous popularity in molecular biology and clinical
`diagnostics. The standard method for the separation of DNA
`fragments is electrophoresis in polyacrylamide and agarose gels
`Despite
`(1).
`impressive advances, however, modern
`electrophoresis is still, for the most part, a collection of manually
`intensive methodologies that cannot be run unattended and that
`are prone to irreproducibility and poor quantitative accuracy.
`The recent introduction of fused silica capillaries as the
`migration channel in electrophoresis has been a big step forward
`in putting electrophoretic separations on the same instrumental
`footing as high-performance liquid chromatography (2, 3). With
`regard to the separation of double-stranded DNA fragments, three
`different approaches have been reported. The first one is based
`on the use of low- or zero-crosslinked polyacrylamide-filled
`capillaries, which allowed the realization of up to five million
`theoretical plates per meter and, hence, the resolution of DNA
`
`fragments that differed by less than 10 base pairs, if their overall
`length was not larger than approximately 500 base pairs (4, 5).
`However, gel-filled capillaries are difficult to prepare and, due
`to the destruction of the gel matrix at high current densities,
`resolving power cannot be maintained for long periods of time.
`Moreover, it is not possible to deliver the same amount of each
`DNA species, because samples can only be injected
`electrokinetically. Therefore, several investigators have explored
`the feasibility of separating DNA fragments in free solution using
`micellar buffer systems containing
`cetyltrimethyl-
`either
`ammonium bromide (6) or sodium dodecyl sulfate (7). However,
`this approach does not yield the same degree of resolution as
`obtained by means of gel-filled capillaries. Finally, the combined
`use of hydroxyethylcellulose and ethidium bromide was found
`to offer both the same resolution as seen with gel-filled capillaries
`as well as the advantage of hydrostatic injection (8).
`However, all capillary electrophoretic approaches to the
`analysis of DNA restriction fragments and,
`especially,
`polymerase chain reaction (PCR) products require their
`desaltation because otherwise the salts will cause considerable
`band broadening due to the conductivity difference between the
`sample zone and the surrounding buffer. In this respect, liquid
`chromatographic techniques are far less affected. The method
`most commonly used for the separation of DNA molecules is
`anion-exchange chromatography. The earliest attempts were
`carried out on columns filled with polychlorotrifluoroethylene
`powder coated with a trialkylmethylammonium chloride (9-11).
`However, insufficient
`coating
`of the
`matrix with
`trialkylmethylammonium chloride resulted in considerable loss
`of DNA. The subsequently introduced porous anion-exchange
`supports Mono Q and Mono P allowed the separation of DNA
`restriction fragments within a wide range of size, but resolution
`of fragments larger than 200 base pairs generally declined rapidly
`and separations of DNA molecules up to a length of 500 base
`pairs took almost four hours (12). A considerable reduction in
`time of analysis to less than 15 minutes as well as a significant
`enhancement in resolution was achieved by the introduction of
`non-porous particles to which diethylaminoethyl groups had been
`bonded chemically (13, 14). However, retention behaviour was
`
`MTX1063
`ModernaTX, Inc. v. CureVac AG
`IPR2017-02194
`
`1
`
`

`

`1062 Nucleic Acids Research, 1993, Vol. 21, No. S
`
`not found to be a direct function of chain length and, secondly,
`the isolated DNA fragments were eluted in a salt gradient which
`had to be removed before the DNA molecules could be employed
`in subsequent experiments such as DNA sequencing.
`liquid
`high-performance
`reversed-phase
`Ion-pair
`chromatography (IP-RP-HPLC) is a well-established technique
`for separating mono- and oligonucleotides (15- 18). The principal
`feasibility of separating DNA restriction fragments by means of
`IP-RP-HPLC on uncoated polychlorotrifluoroethylene powder
`has been demonstrated as early as 1979 (19), but attempts to
`transfer the results to commercial packing materials were not
`successful till 1986, when Eriksson et al. (20) published the
`separation of DNA molecules on a PepRPC (C2/C18) column,
`the packing of which is based on porous silica. However,
`complete separation of restriction fragments up to a size of 600
`base pairs took more than 3 hours, which makes the method
`impractical for both analytical as well as preparative purposes.
`Nevertheless, both studies confirmed that separation of restriction
`fragments in IP-RP-HPLC depends mainly on the chain length.
`Based on recent advances in the synthesis of non-porous
`styrene-divinylbenzene copolymers and their application to the
`rapid analysis of proteins (21) and in combination with an ion-
`pairing reagent to the separation of oligonucleotides (17, 18),
`we now explore the applicability of non-porous alkylated
`poly(styrene-divinylbenzene) particles (PS-DVB) to the separation
`of DNA restriction fragments and PCR products.
`
`MATERIALS AND METHODS
`Chemicals
`Styrene, divinylbenzene (DVB), polyvinylalcohol (PVA), and
`1-chlorooctadecane were purchased from Riedel-de Haen (Seelze,
`FRG). HPLC gradient-grade acetonitrile and methanol as well
`as tetrahydrofurane and aluminium chloride were obtained from
`Merck (Darmstadt, FRG). Buffers were prepared using a stock
`solution of 2 M HPLC-grade triethylammoniumacetate (TEAA,
`Applied Biosystems, San Jose, CA, USA) and high-purity water
`(Epure, Barnstead Co., Newton, MA, USA). Acrylamide, Tris,
`acid, and EDTA were purchased from Bio-Rad
`boric
`Laboratories (Richmond, CA, USA).
`
`DNA size standards
`Size standards of double-stranded DNA were purchased from
`Boehringer Mannheim (pBR322 DNA-HaeIl digest, Mannheim,
`FRG) and USB (¢DX-174 DNA-Hincd
`digest, United States
`Biochemical, Cleveland, OH, USA).
`
`Isolation of blood cells
`Peripheral venous blood from controls and patients suffering from
`B-CLL was collected in 10-ml tubes containing heparin as
`anticoagulant. Mononuclear cells were isolated by density
`gradient centrifugation (Lymphoprep, Nycomed AS, Oslo,
`Norway). After removing these cells from the plasma-Ficoll
`interface, they were washed three times with MOPS (DMEM
`buffered with 25 mM 3-N-morpholinopropanesulfonic acid).
`
`RNA isolation and reverse transcription
`Following two washes with ice-cold phosphate buffered saline,
`total RNA was isolated from mononuclear cells by the guanidine
`thiocyanate-acid phenol extraction method (22). Subsequently,
`first-strand synthesis of cDNA was accomplished in a 0.6 mL
`Eppendorf tube by adding 500 ng of RNA to a total volume of
`
`50 ,uL of 50 mM Tris HCl, pH 8.3, 50 mM KCl, 8 mM
`MgCl2, 5 Ag BSA (Pharmacia, Uppsala, Sweden), 1 mM of
`each dNTP, 250 pmol random hexanucleotide primers, 20 mM
`g-mercaptoethanol, 1.25 U rRNasin®
`ribonuclease inhibitor
`(Cat. No. N251 1, Promega, Madison, WI, USA), and 120 U
`M-MLV reverse transcriptase (M5301, Promega). The reaction
`was carried out at 37 °C for 1 hour and then stopped by boiling
`the sample for 2 min and cooling it on ice afterward.
`Polymerase chain reaction
`PCR was performed in a 0.6 mL Eppendorf tube by adding 2
`AL of cDNA or 20 ng of total RNA to a total volume of 50 yIL
`of 10 mM Tris.HCl, pH 9.0, 50 mM KCl, 1.5mM MgCl2, 0.2
`mM of each dNTP, 0.1 mg/ml gelatin, 0.1 % Triton X-100, 0.5
`,uM of each primer, and 3 U TaqDNA Polymerase (Cat. No.
`M1861, Promega). The sequences of the primers used for the
`amplification of 02-microglobulin (5 'I-ACCCCCACTG-
`AAAAAGATGA and 5'I-ATCTTCAAACCTCCATGATG) as
`1
`multidrug
`resistance (MDR) gene
`the
`well
`as
`(5'-CCCATCATTGCAATAGCAGG and 5'-GTTCAAACTT-
`CTGCTCCTGA) were exactly as described by Noonan et al.
`(23). Together with the primers used for the amplification of
`MDR-3, which were MDR-3/2061, 5'-TGTCAGAAGAGCC-
`TTGATGTGG, and MDR-3/2193, 5 '-TGGCAATGGCACATA-
`CTGTTCC, respectively, they were synthesized on an automated
`DNA Synthesizer (Model 381 A, Applied Biosystems, San Jose,
`CA, USA) using phosphoramidite chemistry and purified by
`means of oligonucleotide purification cartridges (Cat. No.
`400771, Applied Biosystems). Thirty cycles of amplification were
`carried out in a thermocycler (MJ Research Inc., MA, USA)
`with a 94°C denaturation step for 35 s, a 57°C annealing step
`for 30 s, and a 73°C extension step for 60 s. Beginning at cycle
`16 each following DNA synthesis step was elongated for 5 s.
`Finally, samples were cooled to room temperature.
`Purification of PCR products
`In order to obtain a standard for the construction of a calibration
`curve several PCR samples containing a 404-bp product were
`pooled an purified by means of size-exclusion chromatography,
`for which prepacked Bio-Spin 30 (Bio-Rad Laboratories,
`Hercules, CA, USA) columns were used. The columns were
`washed three times with TE buffer, before 60 AL of PCR sample
`were applied to the top of the gel bed. The column was
`centrifuged in a swinging bucket at 110xg for 4 min. This
`procedure was repeated two times with 80 AL of TE buffer. The
`eluted DNA was then precipitated with 0.1 volume of 3 M sodium
`acetate, pH 5.0, and 1 volume of isopropanol, stored at -20
`°C overnight, and centrifuged at 15000 x g for 30 min. The
`supernatant was discarded and the pellet was rinsed with 75%
`ethanol, dried under vacuum and usually resupended in 22 ,tL
`of TE buffer prior to analysis both by means of spectrophotometry
`Perkin Elmer,
`(Lambda 2 UV/VIS Spectrophotometer,
`Uberlingen, FRG), in order to determine the content of purified
`as by high-performance liquid
`well
`PCR product,
`as
`chromatography.
`High-performance liquid chromatography
`The HPLC system consisted of a high-precision pump (Model
`480 GT, Gynkotek, Germering, FRG), a degasser (Liliput,
`Gynkotek), a column oven (Model STH 585, Gynkotek), a
`variable wavelength UV-monitor (Model UVD-160, Gynkotek),
`a biocompatible sample injection valve (Model 9125, Rheodyne
`Inc., Cotati, CA, USA) with a 10-,tl sample loop, and a PC-
`based data system (GynkoSoft, Version 3.12, Gynkotek).
`
`2
`
`

`

`PS-DVB beads were prepared by a two-step microsuspension
`method (17, 24). Subsequently, part of the material was alkylated
`under Friedel -Crafts conditions (Austrian patent filed, No.
`2285/92). Depending on the dimensions of the stainless steel
`column used for packing, either 0.6 g (30 x4.6 mm I.D.) or 1.1
`g (5Ox4.6 mm I.D.) of the modified beads were suspended in
`tetrahydrofurane. Then the slurry was sonicated and packed into
`the column at a pressure of 70 MPa with the use of an air-driven
`pump (Model Maximator MSF 111, Ammann Technik, Kolliken,
`CH) and methanol as the driving solvent. Thereafter, methanol
`was replaced with water at the same inlet pressure.
`Gradient profiles are given on each chromatogram. The
`aqueous buffer was 0.1 M TEAA, pH 7.0, throughout the study.
`In order to keep the concentration of TEAA constant and not
`to be affected by volume contraction during mixing of organic
`solvents with water, the mobile phase was prepared as follows:
`for a 25 % solution of acetonitrile in 0.1 M TEAA, 50 ml of the
`2 M TEAA stock solution were added to 250 ml of acetonitrile
`in a 1000 ml volumetric flask and the final volume was made
`up to 1000 ml by the addition of water.
`Polyacrylamide gel electrophoresis
`Ten /tL of each PCR sample or chromatographically isolated
`DNA restriction fragment were separated in 10% polyacrylamide
`gels in TBE buffer (90 mM Tris-borate, 1 mM EDTA, pH 8.3).
`Separations were carried out in a small vertical Mini-Protean II
`electrophoresis cell (Bio-Rad Laboratories) at a constant voltage
`of 80 V for 90 min. Subsequently, the gels were stained in a
`solution of 0.5 ,ug/mL ethidium bromide in TBE buffer for 25
`min and photographed.
`
`RESULTS AND DISCUSSION
`Characterization of the stationary phase
`Because of their uniform size (2.1 0.12 itm, N =257), no
`sieving of the unmodified non-porous PS-DVB particles was
`required following their synthesis. Both the narrow size
`distribution as well as the non-porosity of the particles allow a
`minimization of the length of diffusion paths. This yields both
`an optimum of resolution as well as a significant reduction in
`time of analysis compared to a porous stationary phase (20).
`Another major advantage of the non-porous PS-DVB adsorbent
`is that the totally organic polymer is operable over a wide pH-
`range, typically 1-13, without any damage occurring to the
`packing. This results in high column life-time and allows the easy
`regeneration of a deteriorated column with aqueous sodium
`hydroxide or organic solvents. As seen in Figure la, dsDNA
`restriction fragments larger than approximately 100 base pairs
`in length could not be resolved sufficiently on unmodified PS-
`DVB particles. As in the case of oligonucleotides (18), an
`improvement in resolving power was obtained upon incorporation
`of polyvinyl alcohol during polymerisation (Fig. lb). An even
`higher degree of resolution, which even allowed the efficient
`separation of dsDNA fragments larger than 200 base pairs in size,
`was achieved by alkylating the PS-DVB particles with octadecyl
`groups (Fig. Ic). This is a consequence of the effective shielding
`of the aromatic rings of the PS-DVB matrix which otherwise
`unwantedly interact with the nucleobases that represent the
`hydrophobic moieties of the amphiphilic polynucleotides.
`Optinization of the chromatographic conditions
`Figure 2 shows, that the resolution of DNA restriction fragments
`increases with increasing concentration of TEAA in the mobile
`
`Nucleic Acids Research, 1993, Vol. 21, No. 5 1063
`
`A
`
`0.002 AU
`I
`
`I11
`
`^ AA
`
`Figure 1. Comparison of uncoated (a) and chemically modified (b,c) PS-DVB
`particles in the analysis of DNA restriction fragments. Columns: (a) PS-DVB,
`(b) PS-DVB-PVA and (c) PS-DVB-C18, 2.1 /Am, 30x4.6 mm I.D.; mobile
`phase: 0.1 M TEAA, pH 7.0; linear gradient from 7.5-13.75% acetonitrile in
`4 min, followed by 13.75-16.25% acetonitrile in 6 min; flow-rate: 1 ml/min;
`temperatre: 50 °C; detection: UV, 254 nm; sample: pBR322 DNA-HaeIII digest,
`0.5 Ag.
`
`6I % acetonitrile
`14A
`14.'
`3.
`
`11.!
`
`81
`
`~
`
`70- ~
`
`~
`
`0
`
`0m
`
`0 0m
`
`T---am *
`I
`
`0
`
`100
`
`200
`
`m --nn
`
`500 600
`
`300
`400
`base pairs
`
`FIgure 2. Effect of the concentration of TEAA in the eluant on the chromatographic
`separation of DNA restriction fragments. Column: PS-DVB-C18, 2.1 Am,
`30x4.6 mm I.D.; mobile phase: (a) 0.025 M, (b) 0.05 M, (c) 0.075 M, (d)
`0.1 M, and (e) 0.125 M TEAA, respectively, pH 7.0; linear gradient from
`7.5-13.75% acetonitrile in 4 min, followed by 13.75-16.25% acetonitrile in
`6 min; flow-rate: 1 ml/min; temperature: 50 °C; detection: UV, 254 nm; sample:
`pBR322 DNA-HaeHI digest, 0.5 jig.
`
`phase. This is especially true for fragments larger than 300 base
`pairs in length. In this connexion, it is important to understand
`the principal mechanisms involved in IP-RP-HPLC (15). Firstly,
`the retention sequence is controlled by the charge of the
`polynucleotides, thus determining the number of ion pairs formed.
`This enables IP-RP-HPLC to separate polydeoxynucleotides
`reliably according to their chain length, largely irrespective of
`their base composition (19, 25). Secondly, the degree of retention
`is directly proportional to the chain length of the n-alkyl group
`of the ion pairing reagent. However, triethylammonium acetate
`
`3
`
`

`

`7S r__--.
`I~/I~~
`,
`
`392/495 --770_1057
`I--- 162M20
`-.-
`1
`.2
`.4
`.6
`.8
`1.2
`flow rate [ml/min]
`
`Rs
`
`6 5
`
`-
`
`4-
`
`3 2 1
`
`. 0
`
`Figure 5. Effect of eluant flow-rate on the resolution of selected pairs of DNA
`restriction fragments from the pBR322 DNA-HaeIl digest. Column: PS-DVB-
`C18, 2.1 Am, 30x4.6 mm I.D.; mobile phase: 0.1 M TEAA, pH 7.0; linear
`gradient from 7.5-13.75% acetonitrile in 4 min, followed by 13.75-18.75%
`acetonitrile in 12 min; flow-rate: 0.2-1 mI/min; temperature: 50 °C; detection:
`UV, 254 nm; sample: IX-174 DNA-HincIl digest, 0.65 pg.
`
`1064 Nucleic Acids Research, 1993, Vol. 21, No. 5
`
`0
`
`100
`
`200
`300
`400
`base pairs
`
`500
`
`600
`
`Figure 3. Percentage of acetonitrile needed for elution of restriction fragments
`from pBR322 DNA cleaved with HaelI, depending on the TEAA concentration.
`Chromatographic conditions as in Fig. 2.
`
`Figure 6. High-resolution liquid chromatographic separation of 31 DNA restriction
`fragments. Column: PS-DVB-C18, 2.1 pm, 50x4.6 mm I.D.; mobile phase: 0.1
`M TEAA, pH 7.0; linear gradient from 8.75- 11.25% acetonitrile in 2 min,
`11.25-14.5% acetonitrile in 10 min, 14.5-15.25% acetonitrile in 4 min, and
`15.25-16.25% acetonitrile in 4 min; flow-rate: 1 ml/min; temperature: 50 °C;
`detection: UV, 254 nm; sample: pBR322 DNA-HaeHI digest, 0.75 gg, and
`4X-174 DNA-HincI digest, 0.65 p4g, respectively.
`
`by evaporation (20). Thirdly, polynucleotides are retained the
`less the higher the solvent strength of the mobile phase.
`Figure 3 shows the effect of TEAA in detail. Above a
`concentration of 75 mM of TEAA, ion-pair formation has reached
`nearly its maximum with regard to DNA restriction fragments
`shorter than 200 base pairs, because their retention behaviour
`remains almost the same. This is reflected in the percentage of
`acetonitrile needed for their elution. Only for longer fragments,
`a further increase in retention is noted and therefore, more
`acetonitrile is required to elute them. As far as resolution is
`concerned, the increase in the concentration of TEAA from 25
`
`Figure 4. Effect of column temperature on the chromatographic separation of
`DNA restriction fragments. Column: PS-DVB-C18, 2.1 pm, 30x4.6 mm I.D.;
`mobile phase: 0.1 M TEAA, pH 7.0; linear gradient from 7.5 - 13.75 % acetonitrile
`in4min, followed by 13.75-16.25% acetonitrilein6min; flow-rate: 1 ml/min;
`temperature: 20-70 °C; detection: UV, 254 nm; sample: pBR322 DNA-HaeElI
`digest, 0.5 p4g.
`
`was not only chosen because it allows the faster elution of
`polynucleotides compared to
`the more commonly used
`tetrabutylammonium salts but also because it can be removed
`
`4
`
`

`

`-.-587
`-434
`.-..-267
`- 1 3 2
`--132
`---80
`bi- -5 1
`
`* N
`
`
`
`N co>
`
`'
`
`Nucleic Acids Research, 1993, Vol. 21, No. 5 1065
`
`peak half-width (min)
`
`.24
`.22
`.20
`.18
`.16
`.14
`.12
`.10
`.08
`
`.01
`
`.1
`Ig9 Injeted
`
`I
`
`* ^
`
`N.t le q
`
`cotco
`
`-un
`
`CL
`
`Figure 9. Dependence of the peak half-width on sample loading. Column: PS-
`DVB-C18, 2.1 itm, 50x4.6 mm I.D.; mobile phase: 0.1 M TEAA, pH 7.0;
`linear gradient from 12.5-17.5 % aceto-nitrile in 5 min; flow-rate: 0.75 ml/min;
`tenperature: 50 °C; detection: UV, 254 nm; sample: 404-bp PCR product, purified
`by size exclusion chromatography.
`
`T
`
`-
`
`6
`2 46
`
`-
`
`-
`
`T
`
`-
`
`8
`
`min
`
`mAU
`
`24
`
`dNTPs&
`primers
`
`ln
`
`18-
`
`12-
`
`6-
`
`a
`
`Figure 7. High-resolution liquid chromatographic separation of three PCR products
`spiked with restriction fragments from a pBR322 DNA-HaeJl digest. Column:
`PS-DVB-CI8, 2.1 Am, 50x4.6 mm I.D.; mobile phase: 0.1 M TEAA, pH 7.0;
`linear gradient from 7.5 - 13.75 % acetonitrile
`in 4 min, followed by
`13.75- 16.25% acetonitrile in 6 min; flow-rate: 1 ml/min; temperature: 50 °C;
`detection: UV, 254 am; sample: 120, 132 and 167-bp PCR products, respectively,
`and 0.5 jig of a pBR322 DNA-HaeIl digest; U=unspecific PCR product. The
`isolated restriction fragments of the pBR322 DNA-HaeIl digest were subsequendy
`separated on a 10% polyacrylamide gel in TBE buffer for 90 min at 80 V. Gels
`were then stained in TBE buffer containing 0.5 ,ug/mL ethidium bromide for 25
`min and photographed.
`
`10 pek wra (mAU-mln)
`
`T
`
`0.002 AU
`I
`
`2
`
`I
`
`4
`
`.01
`
`.1
`
`I pig
`
`Figure 8. Calibration curve for a 404-bp PCR product.
`
`to 125 mM in the eluant resulted in a linear enhancement in
`separation efficiency.
`Figure 4 illustrates the impact of column temperature on the
`resolution of DNA restriction fragments. It is evident that
`retention times increase with increasing column temperature.
`Concomitantly, resolution also improves and reaches its optimum
`at a temperature of 50 'C. The increase in retention is most
`probably the consequence of a defolding and a derotating of the
`double helix, which exposes more phosphate groups for ion-pair
`formation and, consequently, enables stronger interaction of the
`DNA molecules with the stationary phase. At even higher
`temperatures, however, a rapid decrease in separation efficiency
`is noted. This phenomenon is not related to the matrix, because
`similar results have been obtained with pure PS-DVB. The
`
`0
`
`2
`
`min
`
`Figure 10. High-speed chromatographic analyses of PCR-products. Column: PS-
`DVB-C18, 2.1 zm, 30x4.6 mm I.D.; mobile phase: 0.1 M TEAA, pH 7.0;
`linear gradient from 10-15% acetonitrile in 2 min; flow-rate: 1.5 ml/min;
`temperature: 50 °C; detection: UV, 254 am; sample: (a) normal mononuclear
`white blood cells, (b) chronic lymphatic leukemia; 1 = unspecific PCR product,
`2 = (32-microglobulin (120 bp), 3 = multidrug resistance gene II (132 bp), 4
`= multidrug resistance gene I (167 bp).
`
`decrease in resolution and the final failure to resolve the restriction
`fragments at all is rather due to the denaturation of the DNA
`molecules, which occurs in non-aqueous solutions earlier than
`in aqueous solutions (26).
`The flow-rate of the eluant also exerts a significant impact on
`resolution (Fig. 5). In the case of relatively small double-stranded
`polynucleotides up to a chain length of 200 base pairs an increase
`in the flow-rate of the eluant allows a concomitant enhancement
`in resolution due to the minimization of dispersion. But the longer
`the DNA molecules grow, the more prominent becomes their
`mass transfer resistance to move between the stationary and the
`mobile phase. This causes an increase in peak width with
`increasing flow-rate. Therefore, a flow-rate as low as 0.4 ml/min
`
`5
`
`

`

`1066 Nucleic Acids Research, 1993, Vol. 21, No. 5
`
`is preferable for the separation of double-stranded polynucleotides
`larger than approximately 600 base pairs.
`Figure 6 shows the separation of a mixture of 31 DNA
`restriction fragments obtained by the cleavage of pBR322 DNA
`with HaeIf and 4'X-174 DNA with HincH, respectively. It can
`be clearly seen that under optimized chromatographic conditions
`one base-pair resolution is obtained for fragments shorter than
`approximately 150 base pairs. Thereafter, separation efficiency
`decreases gradually.
`Quantitation of PCR products
`Figure 7 shows the separation of three PCR products spiked with
`a pBR322 DNA-HaeIH digest. Subsequently, the individual
`restriction fragments were isolated and subjected
`to
`polyacrylamide gel electrophoresis in order to confirm their
`elution order. It is evident that IP-RP-HPLC allows the separation
`of DNA molecules according to their size. This is in contrast
`to anion-exchange chromatography (12, 13) which failed to
`separate the same restriction fragments according to their length,
`because AT-pairs have been shown to interact more strongly with
`the stationary phase than GC-pairs (12).
`The chromatogram also demonstrates the high sensitivity of
`on-column UV detection, which permitted the determination of
`a PCR amplified 404-bp gene fragment with a lower mass
`detection limit of 0.006 pig. This is equal to a minimum detectable
`concentration of 300 ng/ml using a 20-/il sample loop. This value
`compares favourably to the concentration sensitivity reported for
`capillary zone
`on-column UV absorbance detection
`in
`electrophoresis (8). Moreover, the calibration curve (Fig. 8),
`which had been obtained for the 404-bp PCR product, showed
`excellent linearity (r=0.957) over a range of more than 2
`magnitudes (0.006-1.8 Atg), as assessed by the evaluation of a
`log (signal) versus log (concentration) plot (27). The relative
`standard deviation in peak area measurements was less than 5 %
`as determined with ten repeated injections of the 404-bp PCR
`product.
`The loading capacity was evaluated by injecting increasing
`amounts of a 404-bp PCR product purified chromatographically
`by means of a polyacrylamide size exclusion gel. Subsequently,
`the peak half-width was plotted against the sample load (Fig. 9).
`It remained constant at sample loads up to approximately 0.5 itg,
`and then increased with further increase in the sample load.
`However, for semipreparative purposes at least 5 jig of DNA
`can be applied with little decrease in resolution.
`The recovery of the 404-bp PCR product from the PS-DVB-
`C18 column was determined by measuring its peak area both
`with the column in place as well as after its replacement with
`an internal union. The recovery rate was 97.5%. Furthermore,
`no cross-contamination was detected. Therefore, IP-RP-HPLC
`on non-porous PS-DVB-C18 beads can be considered an
`excellent tool for the rapid analysis and micropurification of
`double-stranded polynucleotides without any prior sample
`pretreatment, as confirmed by the evaluation of the expression
`of 3-microglobulin and the multidrug resistance genes I and III
`in mononuclear white blood cells obtained from a control
`(Fig. lOa) as well as from a patient suffering from chronic
`lymphatic leukemia (Fig. lOb). Furthermore, analyses can be
`carried out with a high degree of reproducibility, the mean relative
`standard deviation of the retention times has been 3.4% for 42
`measurements.
`
`CONCLUSIONS
`It is concluded that IP-RP-HPLC on alkylated non-porous styrene-
`divinylbenzene copolymers represents the first real alternative
`to gel electrophoresis. Allowing an equal sample throughput,
`no time-consuming and laborious pouring of gels as well as no
`staining with the mutagenic ethidium bromide are required.
`Moreover, quantitation can be achieved directly by means of UV
`absorbance at 254 nm without the need for an expensive two-
`dimensional gel scanner. And in comparison with capillary zone
`electrophoresis, not only faster separations are achieved, but PCR
`products may be also analyzed without prior purification.
`
`REFERENCES
`1. Stellwagen, N.C. (1987) Adv. Electrophoresis 1, 177-228.
`2. Gordon, M.J., Huang, X., Pentoney, S.L. and Zare R.N. (1988) Science
`242, 224-228.
`3. Kuhr, W.G. (1990) Anal. Chem. 62, 403R-414R.
`4. Guttman, A., Cohen, A.S., Heiger, D.N. and Karger, B.L. (1990) Anal.
`Chem. 62, 137-141.
`5. Heiger, D.N., Cohen, A.S. and Karger, B.L. (1990) J. Chromatogr. 516,
`33-48.
`6. Kasper, T.J., Melera, M., Gozel, P. and Brownlee, R.G. (1988) J.
`Chromatogr. 458, 303-312.
`7. Cohen, A.S., Najarian, D., Smith, J.A. and Karger, B.L. (1988) J.
`Chromatogr. 458, 323-333.
`8. Nathakamkitkool, S., Oefner, P.J., Bartsch, G., Chin, M.A. and Bonn, G.K.
`(1992) Electrophoresis 13, 18-31.
`9. Hardies, S.C. and Wells, R.D. (1976) Proc. Natl. Acad. Sci. U.S.A. 73,
`3117-3121.
`10. Landy, A., Foeller, C., Reszelbach, R. and Dudock, B. (1976) Nucl. Acids
`Res. 3, 2575-2592.
`11. Eshaghpour, H. and Crothers, D.M. (1978) Nucl. Acids Res. 5, 13-21.
`12. Westman, E., Eriksson, S., Laas, T., Pernemalm, P.-A. and Skold S.-E.
`(1987) Anal. Biochem. 166, 158-171.
`13. Kato, Y., Yamasaki, Y., Onaka, A., Kitamura, T., Hashimoto, T., Murotsu,
`T., Fukushige, S. and Matsubara, K. (1989) J. Chromatogr. 478, 264-268.
`14. Katz, E.D., Haff, L.A. and Eksteen, R. (1990) J. Chromatogr. 512,
`433-444.
`15. Jost, W., Unger, K. and Schill, G. (1982) Anal. Biochem. 119, 214-223.
`16. Kwiatkowski, M., Sandstrom, A., Balgobin, N. and Chattopadhyaya, J.
`(1984) Acta Chem. Scand. B38, 721 -733.
`17. Huber, C.G., Oefner, P.J. and Bonn, G.K. (1992) J. Chromatogr. 599,
`113- 118.
`18. Huber, C.G., Oefner, P.J. and Bonn, G.K. (1992) Anal. Biochem., in press.
`19. Usher, D.A. (1979) Nucleic Acids Res. 6, 2289-2306.
`20. Eriksson, S., Glad, G., Pernemalm, P.-A. and Westman, E. (1986) J.
`Chromatogr. 359, 265-274.
`21. Maa, Y. and Horvdth, C. (1988) J. Chromatogr. 445, 71-86.
`22. Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979)
`Biochemistry 18: 5294-5299.
`23. Noonan, K.E., Beck, C., Holzmayer, T.A., Chin, J.E., Wunder, J.S.,
`Andrulis, I.L., Gazdar, A.F., Willman, C.L., Griffith, B., Von Hoff, D.D.
`and Roninson, I.B. (1990) Proc. Natl. Acad. Sci. USA 87: 7160-7164.
`24. Wongyai, S., Varga, J.M. and Bonn, G.K. (1991) J. Chromatogr. 536,
`155-164.
`25. Haupt, W. and Pingoud, A. (1983) J. Chromatogr. 260, 419-427.
`26. Michelson, A.M. (1963) The Chemistry of Nucleosides and Nucleotides,
`Academic Press, London, pp. 509-512.
`27. Scott, R.P.W. (1986) Liquid Chromatography Detectors, Journal of
`Chromatography Library, Vol. 33, Elsevier, Amsterdam, pp. 12-14.
`
`6
`
`