This document describes a study on using Pluronic F127, a thermo-responsive polymer, as a separation matrix for microchip electrophoresis of DNA fragments. Key points: - Pluronic F127 solutions are low viscosity at low temperatures, allowing easy loading into microchips, but become more viscous and gel-like above 25°C due to formation of micelles. - Viscosity measurements showed a marked increase above 23°C as micelles form. - Separation of DNA fragments from 100 bp to 1500 bp was achieved on chips using Pluronic F127 solutions between 20-25% concentration and temperatures above 25°C. - Fluorescence microscopy showed that T

1. ORIGINAL PAPER
A viscosity-tunable polymer for DNA separation
by microchip electrophoresis
Daisuke Kuroda & Yong Zhang & Jun Wang &
Noritada Kaji & Manabu Tokeshi & Yoshinobu Baba
Received: 21 December 2007 /Revised: 12 May 2008 /Accepted: 20 May 2008 / Published online: 26 June 2008
# Springer-Verlag 2008
Abstract A thermo-responsive separation matrix, consist-
ing of Pluronic F127 tri-block copolymers of poly(ethylene
oxide) and poly(propylene oxide), was used to separate
DNA fragments by microchip electrophoresis. At low
temperature, the polymer matrix was low in viscosity and
allowed rapid loading into a microchannel under low
pressure. With increasing temperatures above 25°C, the
Pluronic F127 solution forms a liquid crystalline phase
consisting of spherical micelles with diameters of 17–19 nm.
The solution can be used to separate DNA fragments from
100 bp to 1500 bp on poly(methyl methacrylate) (PMMA)
chips. This temperature-sensitive and viscosity-tunable poly-
mer provided excellent resolution over a wide range of DNA
sizes. Separation is based on a different mechanism com-
pared with conventional matrices such as methylcellulose. To
illustrate the separation mechanism of DNA in a Pluronic
F127 solution, DNA molecular imaging was performed by
fluorescence microscopy with F127 polymer as the separa-
tion matrix in microchip electrophoresis.
Keywords Pluronic F127 . Viscosity.
Microchip electrophoresis . DNA
Introduction
Electrophoretic separation of DNA by length is generally
performed in flat gels, capillaries, or microchips [1]. Con-
ventional electrophoresis is usually conducted in polyacryl-
amide gels [2], agarose gels [3], or viscous polymer matrices
such as uncrosslinked polyacrylamide, polyethylene oxide, or
methylcellulose [1]. However, highly viscous solutions at
optimal concentrations for DNA separations often require
large back pressures in buffer loading, especially in microchip
electrophoresis (MCE).
Recent progress in nanofabrication techniques has
focused on the development of novel nanostructures as
the separation matrix for DNA analysis. The newly
developed nanostructures have great promise in the
development of high-performance separation technologies
for DNA [4–9]. Among these, nanoball technologies based
Anal Bioanal Chem (2008) 391:2543–2549
DOI 10.1007/s00216-008-2196-4
D. Kuroda :Y. Zhang (*) :N. Kaji :M. Tokeshi :Y. Baba
Department of Applied Chemistry,
Graduate School of Engineering, Nagoya University,
Furo-cho, Chikusa-ku,
Nagoya 464–8603, Japan
e-mail: yzhang@mail.apchem.nagoya-u.ac.jp
J. Wang :N. Kaji :M. Tokeshi :Y. Baba
MEXT Innovative Research Center for Preventive
Medical Engineering, Nagoya University,
Furo-cho, Chikusa-ku,
Nagoya 464–8603, Japan
Y. Baba
Health Technology Research Center,
National Institute of Advanced Industrial Science
and Technology (AIST),
Hayashi-cho 2217–14,
Takamatsu 761–0395, Japan
Y. Baba
Plasma Nanotechnology Research Center,
Nagoya University,
Furo-cho, Chikusa-ku,
Nagoya 464–8603, Japan
Y. Baba
Institute for Molecular Science,
National Institutes of Natural Sciences,
Myodaiji Nishigo-naka 38,
Okazaki 444–8585, Japan

2. on self-assembled copolymer micelles allowed us to
introduce a low-viscosity nanoball solution into a micro-
channel without difficulty and separate a wide range of
DNA molecular weights with high speed and high
resolution [8, 9].
Pluronic tri-block copolymers represent an alternative
separation matrix with its unique characteristics. At low
temperatures (≤18°C), it is a hydrated liquid even at
concentration ranges between 18% and 30% and can be
easily introduced into microchannels. With elevation of
temperature, it changes phase from a sol to a gel-like liquid
crystalline phase, due to the formation of nanometer-sized
micelles, and has potential for use as a DNA sieving matrix.
Pluronic polymers are commercially available triblock
surfactants with a general formula (PEO)x(PPO)y(PEO)x,
where PEO is poly(ethylene oxide) and PPO is poly
(propylene oxide). Pluronic polymers are uncharged and
highly miscible with water. They are classified according to
different values of x and y, such as P65 [(PEO)20(PPO)30(-
PEO)20], PF80 [(PEO)73(PPO)27(PEO)73], and F127
[(PEO)106(PPO)70(PEO)106] [10].
In this study, we chose Pluronic F127 as the separation
medium for DNA analysis in microchip electrophoresis. We
investigated Pluronic F127’s viscosity changes with tem-
perature and found the optimal conditions for separation of
DNA fragments. Above the critical micelle concentration,
we studied the effect of polymer concentration on separa-
tion. For the first time, we found that T4DNA fragments
tended to take a linear path in an electric field using a
Pluronic F127 sieving matrix.
Experimental
Reagents and materials
Pluronic F127, TBE (Tris–borate–EDTA) buffer, and YOYO-
1 (1,1′-[1,3-propanediyl-bis[(dimethylmino)-3,1-propanediyl]]
bis[4-[(3-methyl-2(3H)-benzoxazolylidene)methyl]]tetra-
iodide) in DMSO were obtained from Sigma (St Louis,
MO, USA). TO-PRO-3 iodide (642/661) 1 mmol L−1
solu-
tion in DMSO was purchased from Invitrogen (Eugene,
Oregon, US). DNA ladders (25-bp and 100-bp) and PCR
DNA ladder were from Takara (Shiga, Japan). T4DNA was
from Wako Pure Chemical Industries (Osaka, Japan).
Running buffer with Pluronic F127 polymer was
prepared by adding Pluronic F127 to cool 1×TBE solution
and stirring slowly for 6 h. A fluorescent dye, 0.01% v/v
TO-PRO-3, was then mixed with the buffer and the mixture
was kept in the dark until use. T4DNA solution was
prepared by adding YOYO-1 to 100-fold diluted TBE
buffer. After mixing, the solution was kept in the dark and
cold until needed.
Viscosity measurement
The viscosities of Pluronic F127 solutions were measured
by use of the Viscolite 700 (Hydramotion, York, UK). The
Pluronic F127 solution was transferred to a beaker which
was placed in a temperature-controlled bath at the desired
temperature. In order to assure the accuracy of solution
temperature, measurements were carried out 1–2 h later
when the correct temperature had been established.
Microchip electrophoresis
Microchip electrophoresis was carried out on a Hitachi
SV1210 device. Microchip electrophoresis specifications
were as previously published by our group [11]. Briefly, the
PMMA microchips consisted of a simple cross-channel
100 μm wide and 30 μm deep. Distances from the channel
intersection to the sample, sample waste, buffer, and buffer
waste wells were 5.25, 5.25, 5.75, and 37.5 mm, respec-
tively. The effective separation channel length was 30 mm.
Buffer solution was loaded into the microchannel with a
syringe. The chips were put on ice to decrease the viscosity
of the buffer solution during loading. Microchip electro-
phoresis was then carried out under conditions specified by
the manufacturer.
Observation of T4DNA fragment migration by fluorescence
microscopy
Fluorescence microscopy (FM) detection was carried out
using an Axiovert 135T instrument (Carl Zeiss, Tokyo,
Japan), illuminated by a 100-W mercury arc lamp. Images
were captured by a CCD camera (EB-CCD, C7190–43,
Hamamatsu Photonics, Hamamatsu, Japan). The objective
magnification was 100 and numerical aperture was 1.4 NA.
The T4DNA sample labeled with YOYO-1 [12] was
introduced to the microchannel of poly(methyl methacrylate)
chips. The microchip was then placed on the FM stage. The
migration of the T4DNA fragments was monitored and
recorded by the camera. Cosmos 32 software (Library,
Tokyo, Japan) was used to process the images.
Results and discussion
Viscosity changes of a Pluronic F127 solution
Figure 1 shows the temperature-dependence of the viscosity
of 20% w/w F127 solution in 1×TBE buffer. As the
temperature was raised from 8°C to 15°C, the solution
viscosity remained low, then increased gradually, showing
a marked increase at 23°C. At temperatures below 15°C, a
Pluronic F127 solution could be easily introduced into a
2544 Anal Bioanal Chem (2008) 391:2543–2549

3. microchannel by capillary force, because viscosity was
comparable with that of water. The reason for this low
viscosity is that at low temperature either the PEO block or
the PPO block of the F127 polymer dissolved in the TBE
buffer. Thus, the F127 block copolymer chains existed as
unimers even at 20% w/w concentration and had a low
viscosity.
With elevated temperature, the PPO block dehydrates
and becomes hydrophobic, resulting in a smaller core size.
An increase in the temperature led to the formation of a gel-
like viscous liquid-crystalline phase, consisting of spherical
micelles with diameters of 17–18 nm which pack with local
cubic symmetry [10]. This specific property caused the
change of viscosity, which could be tuned by adjusting the
temperature. Formation of the micelle may be due to
dehydration, because large positive entropy values (heat)
accompanied the process [13].
0 10 20 30 40 50
1.5
2.0
2.5
3.0
3.5
4.0Log/viscosity(cP)
Temperature (o
C)
Fig. 1 Temperature-dependence of the viscosity of 20% w/w Pluronic
F127 solution in 1×TBE buffer. The dotted line approximates the
resulting curve
500 600 700 800 900 1000 1100 1200 1300
0
2000
4000
6000
8000
25%
Migration time (s)
FluorescenceInt.
200 400 600 800 1000 1200
0
1000
2000
3000
4000
5000
6000
7000
8000
23%
FluorescenceInt.
Migration time(s)
400 600 800
0
5000
10000
20%
Migration time (s)
FluorescenceInt.
Fig. 2 Electropherograms obtained from a 25-bp DNA ladder in different concentrations of Pluronic F127 solution (20%, 23%, and 25% w/w).
Experimental conditions: Esep=177 V cm−1
, 40°C
Anal Bioanal Chem (2008) 391:2543–2549 2545

4. Effects of Pluronic F127 concentration on sieving structure
No liquid crystalline phase formed when the Pluronic F127
concentration was below 18% [14] and above this critical
concentration the solution could rapidly transform into a
gel-like phase above 20°C, as shown in Fig. 1. The sieving
effect of Pluronic F127 was thus investigated at 20%, 23%,
and 25% w/w in the viscous liquid crystalline phase region.
The separation of a 25-bp DNA ladder with a PMMA chip
is shown in Fig. 2. The DNA fragments migrated faster and
with higher resolution as the concentration of Pluronic
F127 was reduced. We speculate that a higher concentration
of Pluronic F127 made sieving size smaller and led to
slower migration of the DNA fragments. This result was in
agreement with data reported by others [15, 16]. Affinity of
the PPO core of Pluronic F127 at the higher concentration
may be responsible for poor resolution. Hence, we chose
20% Pluronic F127 in the buffer for further analysis.
Effects of temperature on sieving structure
The viscosity of the F127 block copolymer solution was
temperature-sensitive. To investigate the effects of temper-
ature on MCE performance, 100-bp DNA ladder fragments
were separated over a temperature range from 15°C to 40°C
using 20% w/w Fluronic F127 solution as the separation
medium. Although not in the gel state, the 20% w/w F127
polymer solution at 15°C was able to separate the 100-bp
DNA ladder (Fig. 2). At this temperature, the 20% w/
w F127 polymer solution has a relatively low viscosity, less
than 50 cP (Fig. 1). This result indicated that factors other
than viscosity contributed to DNA separation. The small
micelles may coexist with unimers resulting in a weak
network hindering the movement of DNA fragments.
Compared with the results obtained at higher temperatures,
the faster migration of all DNA fragments is consistent with
this hypothesis.
In addition to the finding that small micelles may play a
role in the separation of DNA fragments, viscosity showed
a pivotal effect on separation. The viscosity of Pluronic
F127 solution was about 2,000 cP at 25°C, about 4,000 cP
at 30°C, and about 3,000 cP at 40°C. At 30°C all the DNA
fragments from 100 bp to 1,500 bp were resolved. At 25°C
and 40°C the 200-bp and 300-bp fragments were only
partly resolved at lower viscosities (Fig. 3). Thus, the
effects of temperature on DNA separation were related to
micelle formation and the viscosity of the solution.
The separation required more than 10 min which seemed
exceedingly long for microchip electrophoresis on a 3-cm-
long channel. Considering the negative charge of the
pristine PMMA channel, EOF effect may dramatically
affect the migration speed of DNA fragments.
400 600 800 1000 1200
1000
2000
3000
4000
5000
6000
7000
15
o
C
FluorescenceInt.
Migration time (s)
400 600 800 1000 1200
1500
2000
2500
3000
3500
4000
4500
25
o
C
FluorescenceInt.
Migration time (s)
400 600 800 1000 1200
1000
1500
2000
2500
3000
3500
4000
4500
1500 bp
1000
900
800
700
600
500
400300
200
100
30
o
C
Migration time (s)
FluorescenceInt.
400 600 800 1000 1200
1000
2000
3000
4000
5000
6000
Migration time (s)
FluorescenceInt.
40 o
C
Fig. 3 Electropherograms obtained from 100-bp DNA ladders at 15, 25, 30, and 40°C. Experimental conditions: 20% w/w Pluronic F127 solution
in 1×TBE buffer, Esep=177 V cm−1
2546 Anal Bioanal Chem (2008) 391:2543–2549

5. Comparison of separation power of Pluronic F127
with methylcellulose
Figure 4 compares the resolving power of a Pluronic F127
solution and a methylcellulose (MC) solution for separation
of a 25-bp DNA ladder. For this purpose, polymer solutions
of Pluronic F127 and MC were adjusted to similar
viscosities in loading by changing concentration. Methyl-
cellulose is a conventional matrix for separation of DNA by
microchip electrophoresis [1]. The resolving powers for
both polymers were almost comparable with smaller DNA
fragments, but Pluronic F127 polymer solution gave better
resolution for the larger DNA fragments indicated by the
dotted rectangle in Fig. 4a. Figure 4b shows the relation
between electrophoretic mobility and DNA size for a 100-
bp DNA ladder in the Pluronic F127 and the methylcellu-
lose solutions. The figure indicates that the slope of the
plots for the larger DNA fragments above 500 bp was
steeper for Pluronic F127 than methylcellulose. This unique
property was similar to the character of a nanoball solution
[8, 9] and promised to separate a wide size range of DNA
within a short time period. This result indicated that the
separation mechanism was fundamentally different from
that of conventional polymers.
The separation mechanism of a Pluronic F127 solution
Several mechanisms have been proposed for the behavior
of DNA during electrophoretic migration in a conventional
polymer. These have been based on the electric field, the
size of the molecule, and the concentration of the polymer
used. Ogston proposed a sieving mechanism such that the
DNA molecules passed through a random network with an
average characteristic pore size as an undeformed spherical
particle [17]. The reptation model assumes that the
migration of a DNA molecule in a polymer network
occurred in a snake-like, head-first movement [18, 19].
The series of connected pores that house the fragment form
an effective “tube” in which the fragment was trapped; no
lateral motion was allowed in the tube. The bias reptation
model suggested that the number of pores housing a given
fragment did not change with time or with electric field
strength. The passage of the fragments through the gel
network led to nonrandom-walk molecular conformations
[20–22]. However, the behavior of DNA electrophoretic
migration in F127 block copolymer did not obey any of the
existing mechanisms [23].
In order to reveal the separation mechanism in Pluronic
F127 polymer, the electrophoretic migration of T4DNA
(165.6-kbp) molecules was monitored with fluorescent
microscopy (Fig. 5). The results showed that T4DNA
migrated through interstitial spaces between micelles and
hydrated PEO strands, squeezing out random-coiled confor-
mations. The data also suggested that T4DNA fragments
tended to take a linear path in an electric field using a
Pluronic F127 sieving matrix. as shown in Fig. 5. DNA
molecular imaging in different separation matrices has been
reported previously, including a crosslinked gel [20–22], an
uncrosslinked polymer [25], nanostructured entropy trapping
[5, 24], nanopillar [12], and nanoball [8]. However, the
0 50 100150200250300350400450500550600650700750800850
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Mobility(10
-4
.cm
2
.s
-1
.V
-1
)
DNA size (bp)
1
2
3
4
(A)
5
6
7
8
9
10
11,12
1
2
3
4
5
6
7
8
9
10
11
12
88
99
0011010
11,12,11,1
88
99
101010
11111
121212
(B)
Fig. 4 (A) Electropherograms obtained from separation of 25-bp
DNA ladders using a 0.2% w/v methylcellulose solution (top) and a
20% v/v Pluronic F127 solution (bottom) on a PMMA chip. (B) Plots
of relative mobility versus DNA size in a Pluronic solution (triangles)
and a methylcellulose solution (circles)
Anal Bioanal Chem (2008) 391:2543–2549 2547

6. movement of a single DNA molecule in a Pluronic F127
polymer solution as shown in Fig. 5 was completely different
from any others.
Since the Pluronic F127 liquid crystalline phase was a
face-centered cubic lattice, revealed by small-angle X-ray
scattering [15], there were at least four different domains in
the Pluronic F127 copolymer. The micellar core was a PPO
block with a diameter of 9 nm, and a minimum gap
between cores of 9 nm. The gap was occupied by hydrated
PEO chains that extended from the micelle core surface.
The PEO chains entangled together to form overlapped
micellar shells. The water-rich gaps were among micelles
[10]. Because Pluronic F127 was not like a crosslinked gel,
such as polyacrylamide, T4DNA fragments traveled around
regularly arranged spheres in a different way. During
movement, the main resistance that T4DNA fragments
experienced came from the PEO brush. The resistance was
much weaker than that seen in crosslinked gel. Thus, it was
not necessary for T4DNA fragments to change the direction
of movement. At the same time, the existence of the PEO
brush forced T4DNA fragments to move in a limited space.
T4DNA thus moved in a linear path.
Conclusion
A thermo-responsive Pluronic F127 solution has been
developed as a viscosity-tunable separation matrix for
microchip electrophoresis of DNA. Separation performance
over a wide range of DNA sizes and easy introduction into
a microchannel are suitable for future microchip-based
separation techniques. While further investigations of
separation mechanisms and a shorter analysis time are still
required, the unique properties and various advantages of
Pluronic F127 over conventional polymers provide new
opportunities in microchip electrophoresis.
Fig. 5 Successive images of a T4DNA fragment in an electric field: a, 0 s; b, 41 s; c, 91 s; d, 141 s; e, 191 s; f, 241 s; g, 291 s; h, 341 s; i, 391 s,
E=41.7 V cm−1
, the bar scale is 10 μm. Arrow indicates direction of migration of the T4DNA fragment
2548 Anal Bioanal Chem (2008) 391:2543–2549

7. References
1. Xu F, Baba Y (2004) Electrophoresis 25:2332–2345
2. Chiari M, Nesi M, Fazio M, Righetti PG (1992) Electrophoresis
13:690–697
3. Aldroubi A, Unser M, Tietz D, Trus B (1991) Electrophoresis
12:39–46
4. Mohamadi MR, Mahmoudian L, Kaji N, Tokeshi M, Chuman H,
Baba Y (2006) Nanotoday 1:38–45
5. Han J, Craighead HG (2000) Science 288:1026–1029
6. Doyle PS, Bibette J, Bancaud A, Viovy JL (2002) Science
295:2237
7. Huang LR, Tegenfeldt JO, Kraeft JJ, Sturum JC, Austin RH, Cox
EC (2002) Nature Biotech 20:1048–1051
8. Tabuchi M, Ueda M, Kaji N, Yamasaki N, Nagasaki Y,
Yoshikawa K, Kataoka K, Baba Y (2004) Nature Biotech
22:337–340
9. Tabuchi M, Katsuyama Y, Nogami K, Nagata H, Wakuda K,
Fujimoto M, Nagasaki Y, Yoshikawa K, Kataoka K, Baba Y
(2005) Lab Chip 5:199–204
10. Rill RL, Liu Y, Winkle DH, Locke BR (1998) J Chromatogr A
817:287–295
11. Zhang Y, Ping G, Kaji N, Tokeshi M, Baba Y (2007)
Electrophoresis 28:3308–3314
12. Kaji N, Tezuka Y, Takamura Y, Ueda M, Nishimoto T, Nakanishi
H, Horiike Y, Baba Y (2004) Anal Chem 76:15–22
13. Wanka G, Hoffmann H, Ulbricht W (1994) Macromolecules
27:4145–4159
14. Liu Y, Locke BR, Winkle DH, Rill RL (1998) J Chromatogr A
817:367–375
15. Wu C, Liu T, Chu B (1997) Macromolecules 30:4574–4583
16. Wu C, Liu T, Chu B (1998) Electrophoresis 19:231–241
17. Ogston AG (1958) Trans Faraday Soc 54:1754–1758
18. Slater GW, Rousseau J, Noolandi J, Turmel C, Lalande M (1988)
Biopolymers 27:509–524
19. Slater GW, Noolandi J (1989) Biopolymers 28:1781–1791
20. Smith SB, Aldridge PK, Callis JB (1989) Science 243:203–206
21. Schwartz DC, Koval M (1989) Nature 338:520–522
22. Gurrieri S, Smith SB, Bustamante C (1999) Proc Natl Acad Sci
USA 96:453–458
23. Rill RL, Locke BR, Liu Y, Winkle DH (1998) Proc Natl Acad Sci
USA 95:1534–1539
24. Ueda M, Oana H, Baba Y, Doi M, Yoshikawa K (1998) Biophys
Chem 71:111–122
25. Han J, Turner SW, Craighead HG (1999) Phys Rev Lett 83:1688–1691
Anal Bioanal Chem (2008) 391:2543–2549 2549