J. Sep. Sci.

2012, 35, 3509–3513 3509

Marko Stojkovic1 Research Article

Narasimha R. Uda1
Peter Brodmann2
Milica Popovic1 Determination of PCR products by CE with
Peter C. Hauser1
contactless conductivity detection
1 Department of Chemistry,
University of Basel, Basel, The use of CE with contactless conductivity detection for the determination of PCR products
Switzerland is demonstrated for the first time. The separation of specific length PCR products according
2 Biosafety Laboratory, State
Laboratory Basel City, Basel,
to their size could be achieved using 5% PVP as a sieving medium in a separation buffer
Switzerland consisting of 20 mM Tris and 20 mM 2-(cyclohexylamino)ethansulphonic acid (pH 8.5). A
fused silica capillary of 60 cm length and 50 ␮m id and an applied separation voltage of
–15 kV were employed and separations could be completed within 20–50 min. PCR amplified
Received August 21, 2012
Revised September 5, 2012
DNA fragments of different sizes obtained from different bacterial plasmid templates as well
Accepted September 5, 2012 as a fragment from genomic DNA of genetically modified soybeans could be successfully
identified.

Keywords: Capillary electrophoresis / Contactless conductivity / Detection /

Genetically modified organisms / PCR fragments
DOI 10.1002/jssc.201200800

1 Introduction Detection for DNA in CE is usually carried out via UV-

absorption as described previously [4–7] or for sequencing
CE has a number of advantages compared to planar gel elec- and when low detection limits are required by fluorescence
trophoresis. In case of CE, higher voltages can be applied and as described previously [8–11]. However, labeling is generally
therefore faster separation times are obtained; staining with required for the latter method.
toxic chemicals such as ethidium bromide is not required Capacitively coupled contactless conductivity detection
for detection and the CE process can easily be automated. For (C4 D) is an attractive alternative detection method. It is inher-
these reasons, CE is now used routinely for DNA sequencing. ently much simpler than optical detection and suitable for any
On the other hand, other forms of DNA analysis, such as the ionic species without requiring derivatization. The technique
determination of restriction fragment length polymorphism also enables the design of inexpensive and compact, portable,
and the simple analysis of PCR-amplified DNA fragments, battery-operated CE instruments [12–16]. The standard UV
are often still carried out by planar electrophoresis and have detection, on the other hand, is not readily implemented for
only partly been replaced by capillary methods. such instruments due to the high power consumption of con-
In both formats, planar electrophoresis and CE, DNA ventional UV lamps. For details on C4 D, recent review articles
fragments are separated by their size in a sieving matrix. are available [17–20].
Conventionally, cross-linked polyacrylamide gels have been A few instances of the detection of DNA in CE by
used very frequently in planar electrophoresis. In contrast, in conductivity measurements have been reported. Galloway
CE, entangled not cross-linked polymer solutions are now et al. [21] used conventional contact conductivity mea-
commonly used as the sieving material [1]. The polymer surements in 2002 for the detection of PCR products in
solutions should have low viscosity for easy handling, and electrochromatography. In 2005, Abad Villar et al. [22]
should have the ability to coat the capillary wall in order to demonstrated the suitability of C4 D for the detection of a DNA
suppress the EOF and prevent adsorption of analytes. Many fragment on a lab-on-chip device. Xu et al. [23] used a poten-
different polymers have been studied for the separation of tial gradient detector (a form of indirect conductometric de-
DNA fragments such as linear polyacrylamide, PVP, poly- tection) for determining DNA ladder segments and this was
N,N-dimethylacrylamid, polyvinylalcohol, polyethyleneoxide, followed by a paper using carbon nanotubes in the sieving
and various cellulose derivatives such as hydroxyethyl cel- medium and C4 D [24]. Mühlberger et al. [25] reported the
lulose, hydroxypropyl cellulose. For details see these review separation and detection of a ladder on a microfluidic chip
articles [2, 3]. using contactless conductivity measurement. Although these
studies indicated the possibilities, practical uses of CE-C4 D
Correspondence: Dr. Peter C. Hauser, Department of Chemistry, for DNA analysis have not yet been shown. In this paper, the
University of Basel, Spitalstrasse 51, 4056 Basel, Switzerland application of the determination of PCR products from plas-
E-mail: [email protected]
Fax: + 41-61-267-1013
mid DNA and also from genomic DNA is demonstrated. Dif-
ferent fragments obtained from bacterial plasmid templates
Abbreviation: C4 D, capacitively coupled contactless conduc- were first determined. This application may be useful, for ex-
tivity detection ample, for the identification of particular microorganisms in


C 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3510 M. Stojkovic et al. J. Sep. Sci. 2012, 35, 3509–3513

Table 1. List of primers used for PCR amplification of the DNA fragments

Primer name Primer sequence (5 –3 )

1 B1& HisDELYodASKA (Forward primer) CGATTCGTCTTTACAAACTGGCTGTTGCTTTAGGTGTC

2 B2& HisDEL-BlcASKA (Forward primer) CGCCTGCTGCCGCTGGTGGCGGC
3 R$ ASKAseq, cloning (Reverse primer) CAGGTCGACCCTTAGCGGCCGCATAGGCC
4 F-pCM655 (Forward primer) GAGCGGATAACAATTTCACACAGGAAACAG
5 F-SEQ pCM655 (Forward primer) GCGTACGTCGCGACCGCGGACATGT
6 R-SEQ pCM655 (Reverse primer) CACCGCGCTACTGCCGCCAGGCA
7 Soy-F (Forward primer) GAAGCAACCAAACATGATCCTC
8 Soy-R (Reverse primer) ATGGATCTGATAGAATTGACGTTA

cell cultures. The second application concerned the identifica- on an Eppendorf PCR Thermal Cycler (Hamburg, Germany).
tion of genetically modified soybeans. In many jurisdictions, The PCR reactions were carried out in a total volume of
food derived from genetically modified organisms must be 50 ␮L, containing 300 nM of each primer (forward and
specially labeled or is even banned completely, and therefore reverse), 4% DMSO, 1.6 mM dNTP’s (0.4 mM each), 5 U
simple means for their detection are desired. The developed Taq polymerase, 10× Thermopol reaction buffer, 100 ng
method for the analysis of PCR products is suitable for im- of template for genomic DNA, and 50 ng of plasmid DNA.
plementation on portable CE-C4 D instruments, and has the The PCR reaction consisted of an initial denaturation step
potential for field measurements. at 95⬚C for 3 min followed by 40 PCR cycles, each cycle
consisting of denaturation at 95⬚C for 3 min, annealing at
59.5⬚C for 1 min, and extension at 72⬚C for 3 min. The final
elongation was for 10 min at 72⬚C to ensure full extension
2 Materials and methods
of all amplified fragments. The PCR products were purified
on Wizard SV PCR-Gel cleanup minicolumns (Promega)
2.1 Chemicals
and eluted with 10 mM Tris-HCl (pH 7.5) containing 1 mM
EDTA. This buffer was selected as it corresponds to the
PVP (MW 1 300 000) was obtained from Sigma-Aldrich
buffer in which the DNA mass ladder was supplied.
(Buchs, Switzerland), Tris and CHES were obtained from
Fluka (Buchs, Switzerland). The DNA mass ladder was pur-
chased from Invitrogen (Carlsbad, CA, USA) that is com-
posed of an equimolar mixture of six low-mass ladder DNA 2.3 Instrumentation
fragments of 100, 200, 400, 800, 1200, and 2000 bp (base
pairs). DNA Taq polymerase with 10× Thermopol reaction Two CE instruments with contactless conductivity detectors
buffer and deoxynucleoside triphosphates (dNTP’s) were ob- were used for the measurements: a commercial as well as an
tained from New England Biolabs (Maine, USA) and oligonu- in-house constructed system. The commercial instrument
cleotide primers from Mycrosynth (Balgach, Switzerland). was a PrinCE 500 (Prince Technologies, Emmen, the Nether-
The DNA template for genetically modified organism soy- lands) and used for the work presented in Figs. 1–3. The
bean flour was obtained from the Cantonal Laboratory of unit built by us was based on the design of a portable, battery-
Basel-City and the genomic DNA was extracted by using the powered instrument described earlier [16], but utilized a high-
genomic extraction REDExtract-N-AmpTM Seed PCR Kit ob- voltage supply from Spellman (CZE 2000R, Spellman, Pul-
tained from Sigma (St. Louis, Missouri, USA). borough, UK) capable of delivering up to 30 kV and was mains
powered. It was used for the measurements presented in
Figs. 4 and 5. The detectors employed with both instruments
were also constructed in-house according to a design reported
2.2 PCR
earlier [26]. The signals were monitored and recorded with
an e-corder data acquisition system (eDAQ, Denistone East,
Plasmid DNAs (pCA24N-YodA, pCA24N-Blc, and
NSW, Australia).
pCM655Empty) were transformed and isolated from
the XL1-Blue strain by using a Wizard plus Miniprep Kit
from Promega (Madison, WI, USA) and genomic DNA of
genetically modified soybeans (soybean template Roundup 2.4 CE
Ready GM-Soybean-EX961053A) by using the REDExtract-
N-AmpTM Seed PCR Kit. These DNA sources were used Fused silica capillaries of 60 cm length, 50 ␮m id, and
as templates for the production of PCR fragments of the 365 ␮m od (Polymicro Technologies, Phoenix, AZ, USA)
specific sizes of 180, 400, 489, 555, and 674 bp by using were used for the CE measurements. New capillaries were
corresponding primers (Table 1). All PCRs were performed conditioned by rinsing with 0.1 M NaOH for 20 min and


C 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2012, 35, 3509–3513 Electrodriven Separations 3511

then with deionized water for 20 min, followed by 0.1 M HCl,

and again with deionized water for 20 min. Before each run,
a capillary was rinsed with the background buffer for 3 min.
The system was then equilibrated by applying the separation
voltage until a stable baseline was obtained. For the analysis
of DNA ladder and PCR samples, hydrodynamic injection
was used. On the instrument constructed internally this was
carried out by siphoning, through lifting the injection part of
capillary up to 30 cm height, and on the commercial instru-
ment, a pressure of 13.8 kPa was applied. In both cases, the
injection time was 45 s. The background buffer consisted of
20 mM Tris, 20 mM CHES (pH 8.5), and 5% PVP (except
where stated otherwise).

3 Results and discussion

3.1 Selection of sieving polymer and choice

of the buffer

As sieving matrix a PVP-based medium reported by Xu

Figure 1. Electropherograms of low-mass DNA ladder fragments
et al. [23] and used in their work using potential gradient of (1) 100 bp, (2) 200 bp, (3) 400 bp, (4) 800 bp, (5) 1200 bp, and
detection was adopted. The polymer was employed by Xu (6) 2000 bp in different concentrations of PVP w/v. Experimental
et al. [23] for its relative low viscosity and its ability to coat conditions: fused silica capillary of 60 cm total and 55 cm effective
the inner capillary surface and thereby minimize the EOF. length with 50 ␮m id. Buffer: 20 mM Tris, 20 mM CHES (pH 8.5).
Separation voltage: –16 kV.
Important for application in CE-C4 D is also its electrical
neutrality. As background electrolyte, a buffer composed of
20 mM Tris/CHES (pH 8.5) was used. This was also adopted
from Xu et al., but used at a lower concentration to improve in Fig. 2. Clearly, variation of the high voltage is a conve-
baseline stability. nient method to achieve the desired compromise between the
Initially, the experiments were carried out with a DNA required separation (spacing of the peaks) and the analysis
mass ladder containing fragments of 100, 200, 400, 800, 1200, time.
and 2000 bp that covers the broad range of interest for the
PCR products to be detected. The effect of the concentration
of PVP for the range from 1 to 6% w/v on the separation is
shown in Fig. 1. Note that due to the responsiveness of C4 D
to all ionic species, several peaks caused by constituents of
the buffer solution in which the ladder is contained are also
observed. As expected, an increase in the concentration of
the polymer led to increased delay times for the fragments
and increasingly improved separation between the peaks. For
concentrations of PVP from 1 to 3%, the separation was not
adequate due to strong interference of the buffer matrix. For
the higher concentrations of the polymer, the ladder compo-
nents were well separated and removed from peaks due to
the matrix alone. However, in contrast to the normal tasks of
CE it is not just necessary to achieve baseline separation of
the fragments, but normally the spacing of the peaks should
be adequate to allow distinction of the PCR products of inter-
est for a task at hand. The concentration of 5% was adopted
for the subsequent work. Solutions of higher concentrations
(6% and more) were not suitable, as flushing of the capillary
was found exceedingly difficult due to the higher viscosity.
Note that, this concentration is higher than that used by Xu
et al., who reported 2%. Also tested was the influence of Figure 2. Electropherograms of low-mass ladder DNA fragments
the applied separation voltage between 12 and 21 kV using in a buffer consisting of 20 mM Tris, 20 mM CHES, and 5% PVP
the buffer containing 5% PVP and the results are shown at different separation voltages. Other conditions as for Fig. 1.


C 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3512 M. Stojkovic et al. J. Sep. Sci. 2012, 35, 3509–3513

Figure 4. Electropherogram of PCR products from bacterial plas-

mid pCM655-Empty of 180 bp and 489 bp length with the low-
mass DNA ladder. Capillary: 56 cm total length and 51 cm effec-
tive length. Separation voltage: –20 kV. Other conditions as for
Figure 3. Electropherograms of PCR products from bacterial plas- Fig. 2.
mids pCA24N-Blc, pCA24N-YodA with and without low-mass
DNA ladder. (A) PCR fragment of pCA24N-Blc bacterial plasmid
of 555 bp length with the low-mass DNA ladder, (B) PCR fragment
of 555 bp, (C) PCR fragment of pCA24N-YodA bacterial plasmid of
tively. In order to obtain a higher concentrations of the PCR
674 bp length with the low-mass DNA ladder, (D) PCR fragment products and thus to improve detection sensitivity, the num-
of 674 bp. Separation voltage: –10 kV. Other conditions as for ber of PCR cycles was increased to 40. The product was mixed
Fig. 2. with the ladder consisting of the six entities of 100, 200, 400,
800, 1200, and 2000 bp fragments, and then separated by the
CE-C4 D procedure. The resulting electropherogram is shown
3.2 Analysis of PCR samples in Fig. 4. As expected, the new tall peak corresponding to the
489 bp fragment is located nicely between the 400 and
3.2.1 Bacterial plasmid DNA 800 bp ladder peaks. This was possible with the relatively high
separation voltage of –20 kV leading to a considerably faster
The first experiments were carried out using the plasmid separation run, about 20 min duration, than for the previous
DNA isolated from Escherichia coli bacteria. Two PCR reac- separation. Under these conditions, the second PCR product
tions were carried out; one PCR reaction using primers 1 of 180 bp, with its comparatively high concentration, overlaps
and 3 and plasmid DNA, pCA24N-YodA, as template and with the smaller peak of the 200 bp fragment of the ladder.
another PCR reaction using primers 2 and 3 and plasmid The even smaller peak of the ladder fragment of 100 bp is also
DNA, pCA24N-Blc, as template yielding fragments of 674 bp found to overlap with an unidentified peak (left most peak in
and 555 bp, respectively. The PCR products were injected the electropherogram) due to a background ion in the sam-
hydrodynamically, because they contain a high salt concen- ple matrix. An improved separation of these early peaks, if
tration that prevents electrokinetic injection without prior to necessary, would be possible by a reduction of the separation
desalting of the sample. Hydrodynamic injection allows direct voltage.
analysis of PCR products without further purification of the
PCR product. The resulting electropherograms are shown in 3.2.2 Genetically modified soybeans
Fig. 3. The products of 674 and 555 bp could be clearly de-
tected and are positioned between the peaks for the fragments After successful detection of PCR product of the plasmid
of 400 and 800 bp of the low-mass DNA ladder. However, the DNA, the detection of PCR product from genomic DNA of ge-
peaks for the PCR products showed a relatively low sensitiv- netically modified soybeans was investigated as a prominent
ity. Note that a relatively low separation voltage of –10 kV was representative of an application of PCR for the identification
employed in order to achieve baseline separation, resulting of food samples. Roundup Ready GM-Soybean-EX961053A
in relatively long electropherograms. was used as a genomic DNA template. The forward Soy-F
In the second experiment, another plasmid DNA tem- and reverse Soy-R primers from Table 1 were designed in or-
plate pCM655-Empty was used as a PCR template with two der to obtain a PCR product of 400 bp length. The annealing
pairs of primers: primers 4 and 6 and primers 5 and 6 temperature of the primers was optimized to Ta = 55.5⬚C
(Table 1), to yield two fragments of 489 bp and 180 bp, respec- and the number of PCR cycles was 30. The resulting capillary


C 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2012, 35, 3509–3513 Electrodriven Separations 3513

with the PCR experiments. This work has been supported through
grant nos. 200020–126384/1 and 200020–137676/1 from the
Swiss National Science Foundation.

The authors have declared no conflict of interest.

5 References

[1] Karger, B. L., Chu, Y. H., Foret, F., Annu. Rev. Biophys.
Biomol. Struct. 1995, 24, 579–610.
[2] Barbier, V., Viovy, J.-L., Curr. Opin. Biotechnol. 2003, 14,
51–57.
[3] Xu, F., Baba, Y., Electrophoresis 2004, 25, 2332–2345.
[4] Beckmann, A., Gebhardt, F., Brandt, B. H., J. Chromatogr.
B 1998, 710, 75–80.
[5] Giovannoli, C., Anfossi, L., Tozzi, C., Giraudi, G., Vanni,
A., J. Sep. Sci. 2004, 27, 1551–1556.
[6] Stellwagen, N. C., Gelfi, C., Righetti, P. G., Biopolymers
1997, 42, 687–703.
[7] Wang, Q., Xu, X., Chin. Chem. Lett. 2003, 14, 1278–1280.
Figure 5. Electropherograms of the PCR product from Roundup
[8] Akbari, A., Marthinsen, G., Lifjeld, J. T., Albregtsen, F.,
Ready GM-Soybean-EX961053A of 400 bp length (A) and low-
Wennerberg, L., Stenseth, N. C., Jakobsen, K. S., Elec-
mass ladder (B). Separation voltage: –15 kV. Other conditions as
trophoresis 2008, 29, 1273–1285.
for Fig. 2. The inset is a picture of the ethidium bromide stained
agarose gel also showing both separations. [9] Garcı́a-Cañas, V., González, R., Cifuentes, A., J. Agric.
Food Chem. 2002, 50, 4497–4502.
[10] Mátyás, G., Giunta, C., Steinmann, B., Hossle, J. P., Hell-
electropherograms of PCR product and ladder are shown in wig, R., Hum. Mutat. 2002, 19, 58–68.
Fig. 5 along with a picture of a conventional 1% agarose [11] Skeidsvoll, J., Magne, U. P., Anal. Biochem. 1995, 231,
slab gel electrophoresis. Note that this conventional agarose 359–365.
gel electrophoresis separation required approximately 2 h, [12] Blanco, G. A., Nai, Y. H., Hilder, E. F., Shellie, R. A., Di-
whereas the separation by CE could be completed in less than cinoski, G. W., Haddad, P. R., Breadmore, M. C., Anal.
30 min using optimized conditions. The left most peaks in Chem. 2011, 83, 9068–9075.
the capillary electropherogram are again due to unidentified [13] Kubáň, P., Seiman, A., Makarotseva, N., Vaher, M., Kalju-
matrix elements from the sample. rand, M., J. Chromatogr. A 2011, 1218, 2618–2625.
[14] Kumar, A., Burns, J., Hoffmann, W., Demattio, H., Malik,
A. K., Matysik, F. M., Electrophoresis 2011, 32, 920–925.
[15] Ryvolova, M., Preisler, J., Brabazon, D., Macka, M., TrAC,
4 Concluding remarks Trends Anal. Chem. 2010, 29, 938–938.
[16] Kubáň, P., Nguyen, H. T. A., Macka, M., Haddad, P. R.,
It has been demonstrated that CE-C4 D is suitable for the de- Hauser, P. C., Electroanalysis 2007, 19, 2059–2065.
termination of PCR products and may be employed in the
[17] Kubáň, P., Hauser, P. C., Electrophoresis 2011, 32, 30–42.
detection of DNA fragments amplified from varied sources.
[18] Kubáň, P., Hauser, P. C., Electrophoresis 2009, 30, 176–
The method is more easily automated than the currently still
188.
widely used slab gel separations. The separation could be
[19] Trojanowicz, M., Anal. Chim. Acta 2009, 653, 36–58.
achieved in a much shorter period of time, and further opti-
mization in this regard is possible. CE-C4 D does not require [20] Matysik, F. M., Microchim. Acta 2008, 160, 1–14.
sample preprocessing and there is no need for labeling. A [21] Galloway, M., Soper, S. A., Electrophoresis 2002, 23,
further advantage is the simplicity of the method and the 3760–3768.
potential for implementation of the method in portable in- [22] Abad-Villar, E. M., Kubáň, P., Hauser, P. C., Electrophore-
struments. This is demonstrated by the results presented in sis 2005, 26, 3609–3614.
the last two figures, which were obtained on the quasifield [23] Xu, Y., Qin, W., Li, S. F. Y., Electrophoresis 2005, 26, 517–
portable instrument built in house, and are undistinguish- 523.
able from the measurements carried out on the conventional [24] Xu, Y., Li, S. F. Y., Electrophoresis 2006, 27, 4025–4028.
commercial instrument. [25] Mühlberger, H., Hwang, W., Guber, A. E., Saile, V.,
Hoffmann, W., IEEE Sens. J. 2008, 8, 572–579.
The authors would like to thank Zhang Ling, and Gong Xiao [26] Zhang, L., Khaloo, S. S., Kubáň, P., Hauser, P. C., Meas.
Yang for some preliminary tests and Marc Creus for assistance Sci. Technol. 2006, 17, 3317–3322.


C 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com