J. Biochem. Biophys. Methods 47 (2001) 5–19
`
`www.elsevier.com/locate/ jbbm
`
`Mutation detection by capillary denaturing
`high-performance liquid chromatography using
`monolithic columns
`
`*
`a
`b
`Christian G. Huber
`, Andreas Premstaller , Wen Xiao ,
`a
`a
`b
`¨
`Herbert Oberacher , Gunther K. Bonn , Peter J. Oefner
`
`a ,
`
`aInstitute of Analytical Chemistry and Radiochemistry, Leopold-Franzens University, A-6020 Innsbruck,
`Austria
`bGenome Technology Center, Stanford University, Palo Alto, CA 94304,USA
`
`Received 29 June 2000; accepted 10 July 2000
`
`Abstract
`
`The high resolving power of the chromatographic separation of single- and double-stranded
`nucleic acids in 200 mm i.d. monolithic poly(styrene–divinylbenzene) capillary columns was
`utilized for mutation screening in polymerase chain reaction amplified polymorphic loci.
`Recognition of mutations is based on the separation of homo- and heteroduplex species by ion-pair
`reversed-phase high-performance liquid chromatography (IP-RP-HPLC) under partially denaturing
`conditions, resulting in characteristic peak patterns both for homozygous and heterozygous
`samples. Six different single nucleotide substitutions and combinations thereof were confidently
`identified in 413 bp amplicons from six heterozygous individuals each of which yielded a different
`unique chromatographic profile. Alternatively, mutations were identified in short, 62 bp PCR
`products upon their complete on-line denaturation at 758C taking advantage of the ability of
`IP-RP-HPLC to resolve single-stranded nucleic acids of identical length that differ in a single
`nucleotide. Separations in monolithic capillary columns can be readily hyphenated to electrospray
`ionization mass spectrometry and promise increased sample throughput by operating in arrays
`similar to those already used in capillary electrophoresis.
`2001 Published by Elsevier Science
`B.V.
`
`Keywords: Denaturing high-performance liquid chromatography; Monoliths; Mutation detection
`
`¨
`¨
`*Corresponding author. Institut fur Analytische Chemie und Radiochemie, Leopold-Franzens-Universitat,
`Innrain 52a, 6020 Innsbruck, Austria. Tel.: 143-512-507-5176; fax: 143-512-507-2767.
`E-mail address: christian.huber@uibk.ac.at (C.G. Huber).
`
`0165-022X/ 01/ $ – see front matter
`PII: S0165-022X( 00 ) 00147-0
`
`2001 Published by Elsevier Science B.V.
`
`1
`
`MTX1046
`
`(cid:211)
`(cid:211)
`

`

`6
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`1. Introduction
`
`With the whole sequence of the human genome available in the year 2000, the
`detection of mutations and polymorphisms is becoming increasingly important in the
`fields of genetics, molecular diagnostics, and cancer research. Methods used for the
`detection of unknown mutations include single-strand conformation polymorphism
`analysis [1], denaturing gradient gel electrophoresis [2],
`temperature gradient gel
`electrophoresis [3], mismatch cleavage methods [4,5], direct sequencing [6], oligo-
`nucleotide arrays [7], as well as gel electrophoresis [8] and high-performance liquid
`chromatography-based [9] heteroduplex detection.
`Denaturing high-performance liquid chromatography (DHPLC) has shown great
`potential in detecting efficiently and sensitively single-base substitutions as well as small
`deletions and insertions in DNA fragments ranging from 100 to 1500 base pairs in size
`[10,11]. Partially denaturing HPLC is based on the separation of homo- from heterodup-
`lex species generated by polymerase chain reaction (PCR) amplification of polymorphic
`loci containing one or more mismatches by ion-pair reversed-phase HPLC (IP-RP-
`HPLC) at elevated column temperatures in the range 48–678C [12]. Recently, it has
`been demonstrated that the high resolving power of IP-RP-HPLC in combination with
`the sequence dependence of the retention of single-stranded nucleic acids enables the
`detection of mutations in short PCR products ( , 100 bp) under completely denaturing
`conditions [13].
`However, despite its high sensitivity and productivity, further improvements in
`DHPLC technology are necessary to achieve higher throughput, information content,
`cost-effectiveness, and reliability of the method required for routine mutation screening.
`One attractive means to enhance sample throughput
`is the miniaturization of the
`separation channel by using capillary columns of 50–320 mm inner diameter [14]. Such
`capillaries can then be bundled into arrays similar to those used in capillary electro-
`phoresis [15]. In addition, the low flow rates applied in capillary HPLC are highly suited
`for on-line hyphenation with mass spectrometry [16,17].
`Due to their excellent chemical and physical stability, micropellicular, alkylated
`poly(styrene–divinylbenzene) (PS/DVB-C18) beads of 2–3 mm diameter packed into
`columns of 50 3 4.6 mm i.d. have become the premier column packing material for
`separating single- and double-stranded nucleic acids with high resolution [18,19]. Novel
`frit designs have enabled the reduction of the inner column diameter to 200 mm and
`allowed the analysis of femtomole amounts of DNA by capillary IP-RP-HPLC [20].
`Subsequently, the in situ synthesis of a continuous, rod-shaped, porous copolymer of
`styrene and divinylbenzene inside a 200 mm i.d. fused-silica capillary eliminates the
`need for tedious frit preparation due to the covalent attachment of the polymer to the
`inner surface of the capillary [17,21]. In addition, such monolithic PS/DVB capillary
`columns have been found to provide up to 40% improved separation efficiency over
`packed column beds.
`In this communication, we report on the applicability of monolithic poly(styrene–
`divinylbenzene) (PS/DVB) capillary columns to the detection of single nucleotide
`polymorphisms (SNPs) in genomic DNA under both partial and complete thermal
`denaturation of PCR products. The chromatographic performance of the monolithic
`
`2
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`7
`
`columns is critically evaluated and compared with the performance of conventional
`state-of-the-art 4.6 mm i.d. columns packed with 2 mm micropellicular PS/ DVB-C18
`particles.
`
`2. Materials and methods
`
`2.1. Sequences and reagents
`
`The following three sequences, with priming regions typed in lower case and positions
`and chemical nature of polymorphic sites indicated in parentheses, were investigated:
`
`Sequence 1, 209 bp, aggcactggtcagaatgaagTGAATGGCACACAGGACAAGTCC
`AGA CCCA GGAAGGTCCAGT AAC ATGGGAG AA GAACGG AAGGAGTTCTAAA
`ATT CAGGGCT CCCTT GGGCTCCCCTGT TTAAAAATGTAGGTTTTATTATTATA
`TTT CATT GTT AACAAAAGTCC (A / G) TGAGA TCTG TGGAGGATAAAGggggagct
`gtattttccatt
`
`Sequence 2, 413 bp, gggggtataagtataaacaaaacTGACCCCATCGCTGCCCTCTTG
`GAGCTGAG AGT CTC ATAAA CAGCT TTAA GGTAATAAAATCATTTT(C / A)TGT
`GCCACAGGAT (G / A) TGAGTTGGTTT GATGACCCTAAAAACACCACTGGAGCA
`TTGACT ACCAGGCT CGCCAATGATGCTG CTCA AGTTAA AGGGGTACGTGCC
`TCCTTTCTACTGGT(G / A)TTTGTCTTAATTGGC(C / T)ATTTTGGACCCCAGCAT
`GAAAC TAATTTTCTC(C / A)TTA(C / T)GGGTGTT AGTTATCATCATTAAGAAAA
`TGTT GAATAAATATCT AACCTACGAATATA TCACATGCTTTTTGTAGCAACAT
`GTT AACTATTT AAACAT TATATACT GTAGAGCATATAGATA ACTTATAAAccat
`ttgctattgctgttatt
`
`Sequence 3, 62 bp, cccaaacccattttgatgctT(G/T)ACTTAAaaggtcttcaattattattttcttaaat
`
`attttg
`
`All oligonucleotide primers were obtained from Life Technologies (Rockville, MD,
`USA). The oligonucleotide sizing marker containing 8–32-mers of
`the sequence
`d(GACT)
`and d(GACT) GT was obtained from Amersham Pharmacia Biotech
`n
`n
`(Uppsala, Sweden). MspI and HaeIII digests of pBR322 and pUC18, respectively, were
`obtained from New England Biolabs (Beverly, MA, USA) and Sigma (St. Louis, MO,
`USA). A stock solution of triethylammonium acetate (TEAA), pH 7.0, was obtained
`from PE Biosystems (Foster City, CA, USA) or prepared by dissolving equimolar
`amounts of triethylamine (Fluka, Buchs, Switzerland) and glacial acetic aced (Fluka) in
`water. HPLC-grade acetonitrile was purchased from J.T. Baker (Phillipsburg, NJ, USA)
`
`3
`
`

`

`8
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`or Merck (Darmstadt, Germany). High-purity water (E-pure, Barnstead Co., Newton,
`MA, USA) was used for preparing the mobile phase.
`
`2.2. Polymerase chain reaction
`
`Polymerase chain reactions were performed in a 50 ml volume containing 10 mM
`Tris–HCl, pH 8.3, 50 mM KCl, 2.5 mM MgCl , 0.1 mM each of the four dNTPs, 0.2
`2
`mM of each primer, 50 ng of genomic DNA, and 1 unit of AmpliTaq Gold (PE
`Biosystems). The PCR cycling regime carried out in a Perkin-Elmer 9600 thermal cycler
`comprised an initial denaturation step at 958C for 10 min to activate AmpliTaq Gold, 14
`cycles of denaturation at 948C for 20 s, primer annealing for 1 min at 63–568C with
`0.58C decrements, and extension at 728C for 1 min, followed by 20 cycles at 948C for 20
`s, 568C for 1 min, and 728C for 1 min. Following a final extension step at 728C for 5
`min, samples were chilled to 68C. Prior to DHPLC under partially denaturing conditions,
`all PCR products were denatured once more at 948C for 3 min and allowed to renature
`over 30 min by decreasing the temperature from 95 to 658C. This ensures the formation
`of equimolar ratios of homo- and heteroduplex species.
`
`2.3. High-performance liquid chromatography
`
`The instumentation for conventional HPLC consisted of an on-line degasser (DG1210,
`Uniflows Co., Tokyo, Japan), two SD-200 high-pressure pumps, an electronic pressure
`module, a 600 ml dynamic mixer, a six-port injection valve mounted into a MISTRAL
`column oven, an automated sample injector (AI-1A), a Dynamax UV-absorbance
`detector set at 254 nm, and a PC-based system controller and data analysis package
`(Varian, Walnut Creek, CA, USA). Preheating of the mobile phase was accomplished in
`an 80 cm 254 mm i.d. PEEK tubing that had been encased in a tin-alloy block
`(Timberline, Inc., Boulder, CO, USA). The stationary phase consisted of 2 mm
`micropellicular, alkylated PS/DVB particles [18] packed into 5034.6 mm i.d. columns,
`which are commercially available (DNASepE, Transgenomic, San Jose, CA, USA). The
`mobile phase was 0.1 M triethylammonium acetate at pH 7.0, containing in addition 0.1
`mM Na EDTA (Sigma). DNA restriction fragments and crude PCR products were
`4
`eluted with linear acetonitrile gradients at a flow rate of 0.9 ml/min.
`The capillary HPLC system consisted of a low-pressure gradient micro pump (LC
`Packings, Amsterdam, Netherlands) controlled by a personal computer, a vacuum
`degasser (Uniflows Co.), a MISTRAL column oven, a six-port valve injector with a 1 ml
`sample loop (Valco Instruments, Houston, TX, USA), a variable wavelength detector
`(UltiMate UV detector, LC Packings) with a Z-shaped capillary detector cell (UZ-ULT-
`NIO, 3-nl cell, LC Packings), and a PC-based data system (UltiChrom, LC Packings).
`Preheating of the mobile phase and consequent partial or complete denaturation of DNA
`fragments was obtained with a 20 cm piece of 25 mm i.d. fused-silica capillary
`positioned in the oven between the injector (mounted outside the oven) and the column.
`Monolithic capillary columns were prepared according to the previously published
`protocol [17].
`
`4
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`9
`
`3. Results and discussion
`
`3.1. Separation of single-stranded and double-stranded nucleic acids
`
`For the successful downscaling of a chromatographic separation from a conventional
`column of inner diameter 4.6 mm to a miniaturized inner diameter of 0.2 mm, a number
`of instrumental parameters have to be modified and optimized. An HPLC instrument
`capable of performing separations utilizing capillary columns of 50–320 mm i.d. must be
`configured with a low dispersion valve for the injection of typically 20 nl to 5 ml of
`sample. It must be able to reproducibly deliver gradients at low flow rates (100 nl/ min
`to 5 ml /min), and requires a sensitive detector with a low-volume detection cell (1–60
`nl) [22]. Reproducible gradients are conveniently obtained by splitting a relatively high
`primary flow of mobile phase delivered by a conventional gradient pumping system by
`means of a T-piece [23]. Thus, only a small portion of the mobile phase is passed via the
`injector onto the separation column, whereas the main flow of the mobile phase goes to
`waste. Since the primary flow is usually in the range of 100–500 ml/min, a reduction by
`a factor of 2–10 in solvent consumption is usually feasible compared to conventional 4.6
`mm i.d. columns operated at a flow of 1 ml/min. Another important consideration with
`microscale gradient HPLC separations is the so-called gradient delay time, the time
`elapsing between formation of the gradient in the mixing system and entering of the
`gradient into the separation column. The lower the flow rate, the more time is needed to
`flush a given volume between the mixing device and the separation column. Therefore,
`careful attention has to be paid to minimize the volumes of all connecting tubing in the
`HPLC system, especially after the splitting device.
`Fig. 1 illustrates the separation of a ladder of mixed-sequence single-stranded
`oligonucleotides ranging in size from eight to 32 nucleotide units and differing in length
`by two nucleotide units. It can be seen that all oligonucleotides are well separated,
`leaving enough peak capacity for obtaining baseline separation of oligonucleotides with
`single nucleotide resolution over the entire investigated size range. The significant
`influence of base sequence on the retention of single-stranded oligonucleotides has been
`reported previously [13,18,24] and can also be deduced from close examination of the
`retention differences of neighboring pairs of oligonucleotides in Fig. 1. The difference in
`retention times between the 8-mer and the 10-mer is bigger than that observed between
`the 10-mer and the 12-mer. The same alternating smaller and larger retention differences
`were observed with all subsequent peaks, suggesting that the addition of G and T
`contributes more to retention than addition of A and C.
`The slight peak asymmetry which is greater for early eluting peaks indicates an excess
`of extra-column volume. A reduction of the extra-column volume to less than 10% of
`the peak retention volume will require modifications of the existing instrumentation,
`including the positioning of the injection valve and sample loop in the column oven.
`Reduction of connecting volumes in the instrument will also help to reduce the relatively
`large gradient delay time of 6 min in the chromatogram of Fig. 1. Such a modification
`will ensure more rapid and improved resolution of single-stranded oligo- and polynu-
`cleotides with increasing column temperature as reported recently [14]. With the current
`instrumentation the resolution was highest at 608C, where the resolution values ranged
`
`5
`
`

`

`10
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`Fig. 1. Chromatogram of mixed sequence oligonucleotide sizing markers eight to 32 bases in length separated
`in a monolithic capillary column. Column, monolithic PS/ DVB, 5030.2 mm i.d.; mobile phase, (A) 100 mM
`TEAA, 0.1 mM Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, 25% acetonitrile; linear
`4
`4
`gradient, 0–5% B in 1.0 min, 5–15% B in 2.0 min, 15–25% B in 8.0 min, 25–30% B in 4 min; flow rate, 3.0
`ml/ min; temperature, 608C; detection, UV, 254 nm; injection volume, 150 nl; sample, 8–32-mer oligonucleo-
`tides, 0.5–1.7 ng/ ml each.
`
`from 7.09 between the 8-mer and the 10-mer to 2.69 between the 30-mer and the
`32-mer.
`The most significant advantage of capillary HPLC is the better signal height-to-sample
`mass ratio, as the peak concentration is proportional to the inverse square of the column
`diameter [25]. This means that the same amount of analyte will give, in theory, a signal
`that is approximately 500 times higher with a 0.2 mm than with a 4.6 mm column. In
`reality, this translates into significantly smaller injection volumes of the order of a few
`hundred nanoliters with the detection limit being as low as 3 fmol for an 18-mer
`oligonucleotide using UV absorbance at 254 nm. This is almost as sensitive as
`fluorescence detection of fluorescent dye-labeled nucleic acids using a 4.6 mm column,
`with lower limits of detection ranging from 0.5 to 3 fmol depending on the attached
`fluorophore [26]. Hence, while the savings in HPLC solvent may only be on the order of
`a factor of 2–10, far more significant savings can be accomplished in PCR reagent
`consumption as amplifications can be carried out in volumes as small as 0.5–1 ml.
`Fig. 2 shows the separation of 30 double-stranded DNA restriction fragments using
`both a 5030.2 mm i.d. monolithic capillary column and a conventional, microparticulate
`5034.6 mm i.d. packed column. As discussed above, the longer time required for
`
`6
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`11
`
`Fig. 2. Comparison of high-resolution IP-RP-HPLC of double-stranded DNA fragments obtained from HaeIII
`and MspI digests of pUC18 and pBR322, respectively, using either a conventional microparticulate 4.6 mm i.d.
`packed column or a 0.2 mm i.d. monolithic capillary column. Conditions in (a): column, 2 mm PS/ DVB-C18
`particles (DNASepE, Transgenomic), 5034.6 mm i.d.; mobile phase, (A) 100 mM TEAA, 0.1 mM
`Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, 25% acetonitrile; linear gradient,
`4
`4
`38–56% B in 4.0 min, 56–67% B in 6.0 min; flow rate, 0.90 ml/min; temperature, 508C; injection volume, 8
`ml. Conditions in (b): column, monolithic PS /DVB, 5030.2 mm; mobile phase, (A) 100 mM TEAA, 0.1 mM
`Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, 25% acetonitrile; linear gradient,
`4
`4
`30–50% B in 4.0 min, 50–65% B in 13.0 min; flow rate, 3.0 ml /min; temperature, 49.78C; detection, UV, 254
`nm; injection volume, 1 ml.
`
`7
`
`

`

`12
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`separation with the monolithic column is a consequence of the greater gradient delay
`time with the capillary HPLC instrument. Nevertheless, the separation efficiency of both
`columns is equivalent, allowing the separation of DNA fragments that differ by less than
`5% in size. The resolution obtained with monolithic capillaries may be further improved
`by reducing the dead volumes in the chromatographic system. A close inspection of the
`chromatograms reveals partial resolution of the two 160 bp fragments, whereas the 174
`and 180 bp fragments are hardly resolved. As demonstrated previously [27], such
`anomalies are a consequence of differences in base composition with very AT-rich
`fragments undergoing partial denaturation already at 508C, which results in their reduced
`retention. Such reductions in retention are also observed in the presence of mismatches
`that disrupt the DNA helix. This constitutes the basis for the successful detection of
`single-base substitutions and small insertions and deletions by denaturing high-per-
`formance liquid chromatography [9].
`A comparison of the gradient profiles applied in Fig. 2 reveals that a lower
`concentration of acetonitrile is required to elute the DNA restriction fragments from the
`monolithic column. Taking into consideration a gradient delay time of 0.7 min with the
`conventional HPLC instrument, the DNA fragments elute at acetonitrile concentrations
`between 11.5 and 16.3% (Fig. 2a). On the other hand, with a gradient delay time of 6
`min in the capillary HPLC instrument, the fragments elute between 10.0 and 14.7%
`acetonitrile (Fig. 2b). This is due to the lower degree of non-polarity of the monolithic
`PS/DVB stationary phase, which is not alkylated, in contrast to the PS/DVB particles,
`which are derivatized with highly non-polar octadecyl groups [18].
`
`3.2. Partially denaturing HPLC for mutation detection
`
`Partially denaturing HPLC exploits the reduced retention of heteroduplex molecules
`containing one or more mismatches at elevated column temperatures compared to a
`corresponding perfectly matched homoduplex. The reliability of DHPLC in detecting
`single nucleotide substitutions as well as short deletions and insertions is due to the fact
`that most mismatches can be detected over a temperature range of several degrees [10].
`This is exemplified in Fig. 3, which shows the successful detection of an A to G
`transition in a 209 bp amplicon between 55.8 and 59.88C. The presence of a mismatch is
`generally recognized by the appearance of one or more additional peaks. The number of
`peaks detected is a function of the nature and the number of mismatches as well as the
`column temperature.
`As shown in Fig. 3, at a temperature of 57.88C all four species of hetero- and
`homoduplices can be separated. The order of elution of the four species is dictated by
`the degree of destabilization of the DNA helix, which is influenced by both neighboring
`stacking interactions [28] and the ability of certain bases to form atypical Watson–Crick
`base pairs [29]. The thermal stabilities of typical Watson–Crick and mismatched base
`pairs as a function of the flanking base pairs in otherwise homologous 373 bp DNA
`fragments have been determined previously by temperature-gradient gel electrophoresis
`[28]. According to the ranking of stabilities provided in that study, one would expect the
`homoduplex with the AT-base pair to be retained less than that with the GC-base pair.
`And indeed, separate injection of the two homoduplices as well as a mixture of the two
`
`8
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`13
`
`Fig. 3. Resolution of 209 bp homo- and heteroduplex DNA fragments containing a single Afi G transition
`under partially denaturing conditions. Column, PS /DVB monolith, 5030.20 mm i.d.; mobile phase, (A) 100
`mM TEAA, 0.1 mM Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, 25% acetonitrile;
`4
`4
`linear gradient, 30–50% B in 3.0 min, 50–70% B in 7.0 min; flow rate, 3.0 ml /min; temperature, 54.7–59.88C;
`detection, UV, 254 nm; injection volume, 500 nl; sample, sequence 1.
`
`that had not been denatured and reannealed confirmed the expected elution order. With
`regard to the two heteroduplices, the GT heteroduplex should be retained longer because
`the flanking base pairs have been determined to stabilize this mismatch more than the
`AC mismatch. In addition, two hydrogen bonds can form between G and T, and this
`atypical Watson–Crick base pair stacks within a DNA B-conformation duplex with
`relatively little distortion, as indicated by NMR and X-ray crystallographic studies [1].
`As shown in Fig. 4, different mismatches or combinations of mismatches (Table 1)
`tend to generate different chromatographic profiles. Consequently, sequence analysis,
`which is still required to determine the exact location and nature of mismatches, can be
`limited to a few representative profiles. However,
`identical profiles for different
`mutations located within the same melting domain but not at the same nucleotide
`position have been reported [30]. The chromatographic profiles obtained with a
`monolithic 0.2 mm i.d. capillary column (Fig. 4) are very similar to those observed with
`a 4.6 mm i.d. microparticulate column (Fig. 5), although a 28C higher column
`temperature was required. Since the residence times in the preheating loops between
`injector and column with both the conventional and capillary HPLC instrument are
`
`9
`
`

`

`14
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`Fig. 4. Influence of the nucleotide position, the number of mismatches, and column temperature on the
`chromatographic separation profiles obtained for one homozygous control and six heterozygotes under partially
`denaturing conditions using a monolithic 200 mm i.d. capillary column. Column, monolithic PS /DVB, 6030.2
`mm i.d.; mobile phase, (A) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM
`4
`Na EDTA, pH 7.0, 25% acetonitrile; linear gradient, 43–50% B in 0.5 min, 50–54% B in 3.5 min; flow rate, 3
`4
`ml/ min; temperature, 61 and 578C; injection volume, 500 nl; sample, sequence 2. See Table 1 for the nature
`and location of mismatches underlying the chromatographic profiles.
`
`practically identical (2.7 versus 2.5 s), the difference in temperature is due to the more
`polar character of the monolith surface. Thus, lower concentrations of acetonitrile are
`required for the elution of the amplicons, which in turn results in higher stability of the
`double helical structure at a given temperature. All heterozygote profiles except one are
`clearly recognized as such. The only exception is Aus23, which displays only a trailing
`shoulder indicating that
`the melting domain in which the mutation is located has
`undergone almost complete denaturation both in the heteroduplex as well as in the
`homoduplex species. This was confirmed by re-running the sample at a column
`temperature 48C lower, where a clean heteroduplex profile is observed.
`
`3.3. Completely denaturing HPLC for mutation detection
`
`It has been noticed that DNA fragments shorter than approximately 150 bp are too
`unstable to allow the detection of mutations by partially denaturing HPLC. This
`limitation can be overcome by the complete on-line denaturation of short amplicons into
`
`10
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`15
`
`Table 1
`Position and nature of mismatches contained in 413 bp amplicons of six heterozygous individuals and a
`homozygous control. The sequence included exon 20 and flanking non-coding regions of the human
`P-glycoprotein (MDR1, Genbank Acc. No. M14758)
`
`Individual
`
`a
`
`Positions of mutations from 59-end of forward primer
`(positions within the MDR1 sequence)
`
`b
`
`91
`(2398211)
`
`105
`(2401)
`
`209
`(2481124)
`
`225
`(2481140)
`
`258
`(2481173)
`
`262
`(2481177)
`
`P218
`P100G
`SD18
`GM2064A
`P111G
`OM135
`Aus23
`
`C
`C
`C/A
`C
`C
`C
`C
`
`G
`G
`G
`G
`G
`G/A
`G
`
`G
`G/A
`G
`G/A
`G
`G
`G
`
`C
`C
`C
`C/T
`C/T
`C
`C
`
`C
`C
`C
`C/A
`C/A
`C
`C
`
`C
`C
`C
`C
`C
`C
`C/T
`
`a Individuals are from the Stanford human diversity panel.
`b Sequence 2, see Materials and methods.
`
`their single-stranded components which are then resolved on the basis of their sequence
`composition rather than their size [13]. Since differences in base composition as small as
`a single base out of 100 bases suffice to separate two single-stranded nucleic acids of
`identical size, the alleles of a given polymorphic locus can be resolved without the
`addition of a reference chromosome. The only exceptions to this rule have been C to G
`transversions [13]. Fig. 6 shows an example of allelic discrimination based on the
`separation of the single-stranded components of a 62 bp PCR product containing a single
`G to T transversion by means of DHPLC on a monolithic capillary column under
`completely denaturing conditions. The chromatograms in Fig. 6a and b illustrate two
`homozygous samples, where the two completely denatured single strands of the same
`chain length are completely separated by IP-RP-HPLC. In the case of a heterozygote
`sample, three peaks are observed corresponding to two coeluting and two separated
`single strands (Fig. 6c). As mentioned above, the single base change affects the retention
`of the single-stranded components sufficiently to allow the discrimination of mutated
`from wild-type DNA in at least one pair of corresponding DNA strands. Moreover, the
`separation on the monolithic column is very similar to that reported recently for this
`polymorphic locus on a microparticulate 4.6 mm i.d. column (compare Fig. 6f in Ref.
`[13]).
`One distinct advantage of capillary IP-RP-HPLC is the feasibility of coupling this
`separation system on-line to mass spectrometry. This will allow positive confirmation of
`the identity of the resolved single-stranded components and the unambiguous genotyping
`of the amplified PCR fragments. One very crucial step in mass spectrometric analysis of
`nucleic acids is the proper desalting and removal of low-molecular-weight impurities in
`the samples in order to obtain mass spectra of good quality [31]. The on-line purification
`of nucleic acids by HPLC prior to electrospray ionization mass spectrometry (ESI-MS)
`is very attractive because it not only removes cations from nucleic acid samples, but it
`also allows fractionation of nucleic acid mixtures that are otherwise too complex for
`direct infusion ESI-MS. Moreover, the identification even of heterozygous alleles eluting
`
`11
`
`

`

`16
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`Fig. 5. Influence of the nucleotide position, the number of mismatches, and column temperature on the
`chromatographic separation profiles obtained for one homozygous control and six heterozygotes under partially
`denaturing conditions using a conventional microparticulate 4.6 mm i.d. column. Column, 5034.6 mm i.d.
`stainless steel column packed with 2 mm PS /DVB-C18 particles (DNASepE, Transgenomic); mobile phase,
`(A) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, (B) 100 mM TEAA, 0.1 mM Na EDTA, pH 7.0, 25%
`4
`4
`acetonitrile; linear gradient, 50–52% B in 0.5 min, 52–59% B in 3.5 min; flow rate, 0.9 ml/min; column
`temperature, 59 and 558C; injection volume, 9 ml; sample, sequence 2. See Table 1 for the nature and location
`of mismatches underlying the chromatographic profiles.
`
`as one single chromatographic peak should become feasible because ESI-MS can
`directly analyze and deconvolute simple mixtures of nucleic acids [32].
`The mobile phase components usually applied in IP-RP-HPLC, namely water,
`acetonitrile, triethylamine, and acetic acid, are volatile, and, therefore, suitable for direct
`interfacing with ESI-MS. Recently, it has been demonstrated that the solvent com-
`position has a tremendous effect on the mass spectrometric detectability of nucleic acids
`[16,33]. A reduction in the eluent conductivity through a decrease in ion-pair reagent
`concentration from 100 to 25 mM, and the exchange of acetic acid with carbonic acid
`resulted in considerable gains in mass spectrometric signal intensities at the cost of only
`a small decrease in chromatographic separation efficiency. Additionally, an increase in
`the proportion of organic solvent
`in the electrosprayed solution upon post-column
`addition of acetonitrile as sheath liquid through the triaxial electrospray probe was found
`to greatly enhance the detectability of nucleic acids in ESI-MS. Under such optimized
`conditions, purification, separation, detection, and mass characterization of low fem-
`
`12
`
`

`

`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`17
`
`Fig. 6. Direct allelic discrimination of two alleles based on their different retention under completely
`denaturing conditions using a monolithic capillary column. Column, monolithic PS /DVB, 5030.2 mm i.d.;
`mobile phase, (A) 100 mM TEAA, pH 7.0, (B) 100 mM TEAA, pH 7.0, 20% acetonitrile; linear gradient,
`36–48% B in 10.0 min; flow rate, 3.0 ml /min; temperature, 758C; detection, UV, 254 nm; injection volume,
`500 nl; sample, sequence 3, (a) homozygous G, (b) homozygous T, (c) heterozygous G/T.
`
`13
`
`

`

`18
`
`C.G. Huber et al. / J. Biochem. Biophys. Methods 47(2001)5–19
`
`tomole amounts of nucleic acids in real matrices such as PCR reactions can be realized
`in a time frame of a few minutes.
`
`4. Conclusions
`
`It has been demonstrated that monolithic 200 mm i.d. columns are highly suited for
`the detection of mutations by capillary denaturing high-performance liquid chromatog-
`raphy. Miniaturization of the separation channel did not affect the resolving power of the
`system. Moreover, it allowed a reduction in solvent consumption by a factor of 3 and
`reduced the amount of sample required for analysis to a few hundred nanoliters. Future
`developments utilizing the developed miniaturized column technology will certainly
`include the bundling of capillaries into arrays coupled with multi-color fluorescence
`detection and hyphenation of the separation system with mass spectrometry, which will
`significantly improve the productivity of DHPLC for high-throughput mass screening for
`known and unknown mutations.
`
`Acknowledgements
`
`This study was supported by the Austrian Science Fund (P-14133-PHY) and NIH
`grant HG01932.
`
`References
`
`[1] Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T. Detection of polymorphisms of human DNA by
`gel
`electrophoresis
`as
`single-strand confirmation polymorphisms. Proc Natl Acad Sci USA
`1989;86:2766–70.
`[2] Myers RM, Lumelsky N, Lerman LS, Maniatis T. Detection of single base substitutions in total genomic
`DNA. Nature 1985;313:495–8.
`[3] Tee MK, Moran C, Nicholas FW. Temperature gradient gel electrophoresis: detection of a single base
`substitution in the cattle beta-lactoglobulin gene. Anim Genet 1992;23:431–5.
`[4] Gibbs RA, Caskey CT. Identification and localization of mutations at
`the Lesch-Nyhan locus by
`ribonuclease A cleavage. Science 1987;236:303–5.
`[5] Cotton RG, Rodrigues NR, Campbell RD. Reactivity of cytosine and thymine in single-base-pair
`mismatches with hydroxylamine and osmium tetroxide and its application to the study of mutations. Proc
`Natl Acad Sci USA 1988;85:4397–401.
`[6] Rosenthal A, Charmock JDS. New protocols for DNA sequencing with dye