Microchip isoelectric focusing with monolithic immobilized pH gradient materials for proteins separation

Research Article

Microchip isoelectric focusing withmonolithic immobilized pH gradientmaterials for proteins separation

Monolithic immobilized pH gradient (M-IPG) materials were prepared in microchannles

by photoinitiated polymerization of acrylamide, glycidylmethacrylate and Bis, followed by

the attachment of focused Ampholine onto the surface of porous monoliths via epoxide

groups. With M-IPG materials as matrix, FITC-labeled ribonuclease B, myoglobin and

a-casein were well separated by microchip isoelectric focusing (mCIEF) without carrier

amphocytes (CAs) added in the buffer. Both chemical and pressure mobilization were

applied to drive focused zones for LIF detection. Our experimental results showed that

pressure mobilization was preferable with neglectable band broadening, and good peak

shape and high detection sensitivity were obtained. All these results demonstrate that

mCIEF with M-IPG materials is not only an efficient mode for protein enrichment and

separation but also attractive to couple with other CE modes to achieve multi-dimen-

sional separation or MS for further identification, without the interference of mobile

CAs.

Keywords:

Microchip isoelectric focusing / Monolithic immobilized pH gradient materials /Photopolymerization / Proteins DOI 10.1002/elps.200900209

1 Introduction

As Hjerten and Zhu [1] transferred IEF from the conven-

tional slab gel format to CE in 1985, CIEF has been regarded

as a powerful tool for protein analysis [2–4], which has

advantages of high peak capacity, high resolution and high

sensitivity. In the past decades, with the rapid development

of m-TAS techniques, microchip isoelectric focusing (mCIEF)

has been paid much attention and shown potentials of high

throughput analysis [5–9]. Furthermore, it has been used as

the first dimension of microfluidic chip-based multi-

dimensional separation [10–15].

In mCIEF, carrier ampholytes (CAs) are usually added in

the running buffer to establish a stable pH gradient.

However, CAs are high concentration salt mixture, and their

existence in the running buffer might lead to increased

current and Joule heat during focusing. In addition, they

might also interfere with further multi-dimensional

separation and the identification by MS. Therefore, the

establishment of stable pH gradient without mobile CAs in

the running buffer has been paid much attention. Several

methods, such as diffusion of OH� and H1 between anode

and cathode [16–17], Joule heat-induced temperature gradi-

ent [18–20] and spatially varied surface electric field [21],

have been developed. However, the stability of formed pH

gradient should be further improved.

The immobilization of pH gradient on matrix is another

efficient method to generate satisfactory pH gradient for IEF

separation, which was first reported as the solid gel strip

[22], formed by casting polyacrylamide gel matrix with

covalently bonded acrylamido buffers (Immobilines) of

different pK. Sommer et al. [23] applied conventional

immobilized pH gradient gels to microchips by the diffu-

sion of Immobilines across the separation channel prior to

gel photopolymerization, and several fluorescent pI markers

and proteins were separated without the addition of CAs.

However, the microscale casting procedure is complicated.

In our previous study, monolithic immobilized pH

gradient (M-IPG) materials were prepared in capillaries to

perform CIEF without CAs added in the buffer, and good

resolution was achieved for proteins and peptides separation

[24–26]. In this paper, based on our previous work, we

Yu Liang1,2

Yongzheng Cong1,2

Zhen Liang1

Lihua Zhang1

Yukui Zhang1

1National Chromatographic R. &A. Center, Dalian Institute ofChemical Physics, ChineseAcademy of Sciences, Dalian,P. R. China

2Graduate School of the ChineseAcademy of Sciences, Beijing,P. R. China

Received March 30, 2009Revised July 29, 2009Accepted September 2, 2009

Abbreviations: AAm, acrylamide; AIBN, azobisisobutyronitrile;

CAs, carrier ampholytes; lCIEF, microchip isoelectric focusing;

GMA, glycidylmethacrylate; M-IPG, monolithic immobilizedpH gradient

Correspondence: Professor Lihua Zhang, Key Laboratory ofSeparation Science for Analytical Chemistry, National Chroma-tographic R. & A. Center, Dalian Institute of Chemical Physics,Chinese Academy of Sciences, 457 Zhongshan Road, Dalian116023, P. R. ChinaE-mail: [email protected]:186-411-84379779

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2009, 30, 4034–40394034

proposed a simple method to prepare M-IPG materials in

microchannels by photoinitiated polymerization of acrylamide

(AAm), glycidylmethacrylate (GMA) and Bis, followed by the

attachment of focused Ampholines onto porous monolith

surface via epoxide groups. With such a matrix, several FITC-

labeled proteins were separated by mCIEF with high resolu-

tion, high detection sensitivity and high throughput.

2 Materials and methods

2.1 Reagents and instrumentation

DMSO was obtained from Shenyang Chemical Reagent

Plant (Shenyang, China). AIBN was purchased from The

Fourth Shanghai Reagent Plant (Shanghai, China) and

recrystallized in our laboratory. AAm and Bis were ordered

from Acros Organics (NJ, USA). 1,4-Butanediol, dodecanol

and GMA were from Fluka (St. Gallen, Switzerland).

3-methacryloxypropyltrimethoxysiliane (98%), Ampholine

(pH 3.5–10.0), ribonuclease B (from bovine pancreas),

myoglobin (from equine) and a-casein (from bovine milk,

70%) were purchased from Sigma (St. Louis, MO, USA);

SephacrylTM

S-200 was bought from Amersham Phamacia

Biotech (Uppsala, Sweden).

XX-15A/F UV lamps were purchased from Spectronics

(NY, USA). Type SG2506 glass plates were purchased from

Shaoguang Microelectronics (Changsha, China). Fused-silica

capillaries (75 mm id, 365 mm od) were purchased from

Ruifeng Chromatographic Device (Yongnian, China). Intel-

ligent eight-path-high-voltage electric device was purchased

form Shandong Normal University (Jinan, China). Microchip

electrophoresis system with an LIF detector was ordered from

Zhejiang University (Hangzhou, China).

2.2 Fluorescent labeling of proteins

Each kind of protein (about 1 mg) was, respectively,

dissolved in 300 mL of 0.02 M phosphate buffer (pH 8.0),

and then 5 mL of 20 mg/mL FITC dissolved in DMSO was

added. The mixture was incubated overnight at room

temperature with continuous stirring. The labeled proteins

were purified by SephacrylTM

S-200 before analysis.

2.3 Microchip fabrication

A simple straight microchannel, 100 mm wide and 4.5 cm

long, as shown in Fig. 1A, was fabricated with Type SG2506

glass plates by photolithography, wet chemical etching and

room-temperature bonding, as described by Fang et al [27].

Briefly, a design on a photomask with microchannels was

transferred onto a glass substrate with chromium and

AZ1805 photoresist coating by UV exposure. Then the

microchannels were etched in a well-stirred bath containing

dilute HF/HNO3. After reservoirs were drilled, etched

substrate and cover plate were washed sequentially with

acetone, detergent and water with high flow rate. Subse-

quently, the plates were dried by heating, followed by being

soaked in concentrated sulfuric acid for 8–12 h. Finally, the

plates were washed with water, bonded and dried at room

temperature

2.4 Pre-treatment of microchannels

Before photopolymerization, the inner wall of microchannels

was vinylized to ensure the covalent attachment of monolith.

Then the channel was washed by HCl (0.1 mol/L) for 30 min,

water for 30 min, NaOH (0.1 mol/L) for 2 h, water for 30 min

Figure 1. Photograph of microchip with M-IPG materials (A) andSEM of M-IPG materials in microchannels (B).

Electrophoresis 2009, 30, 4034–4039 Microfluidics and Miniaturization 4035


and methanol for 30 min, respectively, followed by drying

with N2 in an oven at 701C for 2 h. Subsequently, the channel

was filled with 50% v/v 3-methacryloxypropyltrimethoxysi-

lane in methanol, and kept at room temperature in the dark

for 24 h. Finally, the vinylized channel was washed with

methanol and dried by N2.

2.5 Preparation of M-IPG materials in microchannels

M-IPG materials were prepared in microchannels by

photoinitiated polymerization with various reaction solution

compositions, as shown in Table 1. The mixture was first

purged with N2 for 1 min to remove dissolved O2. Then the

microchannel was completely filled with the polymerization

mixture and sealed with sealing tapes. The microchip was

subsequently covered with a mask, with only the channel

exposed to UV light for 20 min in a black box, equipped with

two 365 nm, 15 W UV lamps with an overall intensity of

3790 mW/cm2 at a reaction distance of 10 cm. The black box

was kept at room temperature, and to avoid undesired

thermal polymerization, a fan was equipped to dissipate

heat.

After polymerization, the monolith was first washed

with ethanol and water, and then filled with 10% w/v

Ampholine. Simultaneously, the anodic and cathodic

reservoirs were filled with 0.020 mol/L H3PO4 and NaOH,

respectively. With 1600 V voltage applied, Ampholines were

migrated in microchannel to form a stable pH gradient by

electric focusing. Until the current was kept stable, buffers

in both reservoirs were changed to water, and sealed with

sealing tapes before kept in an oven at 601C for 24 h to

immobilize CAs. Finally, M-IPG materials in microchannels

were washed by water for 1 h, and ready for separation, as

shown in the amplified zone in Fig. 1A.

2.6 Operation of lCIEF

After the microchannel with M-IPG materials was filled

with FITC-labeled proteins solution, the anodic and cathodic

reservoirs were filled with 0.020 mol/L H3PO4 and NaOH,

respectively. With 1600 V voltage applied, the focusing

was performed until the current was kept stable. The

focused sample zones were subsequently mobilized

either by chemical or pressure mobilization, and then

detected by an LIF detector. Chemical mobilization was

achieved by replacing NaOH (catholyte) with H3PO4

(anolyte), and 2400 V voltage was applied. Hydrodynamic

mobilization was performed by the manual pressure

pump, which was connected to the anodic reservoir via a

75 mm id fused-silica capillary, fixed by a home-made

interface.

3 Results and discussion

3.1 Optimization of polymerization solution

Photopolymerization technique has been widely used to

prepare monoliths in microchips [28–29], since the reaction

can be finished within a very short time even at room

temperature. Furthermore, the monoliths can be prepared

within any required position by using a photo mask.

In this study, to prepare monolithic matrix, AAm, GMA

and Bis were chosen as monomers, and binary porogens,

including dodecanol and 1,4-butanediol were selected, and

the reaction procedure is shown in Fig. 2A However, the

optimal composition for polymerization initiated by heat in

capillaries [14] was found not suitable to prepare M-IPG

materials in microchannels. Therefore, the optimization of

polymerization solution was performed.

The ratio of monomers has effects on both morphology

and functionality of monolithic matrix. The increase of

GMA is favorable to immobilize more Ampholines via epoxy

groups to establish stable pH gradient, while AAm was

helpful to improve the hydrophilicity of matrix, thus to

reduce the non-specific absorption of samples. In addition,

to improve the matrix rigidity, there should be enough Bis

in the polymer solution. After the systematic optimization,

the final ratio of GMA, AAm and Bis was chosen as

24.76:21.90:53.33 w/w.

The types and ratios of porogens have great effects on

the porous structure of monoliths. In our experiments,

1,4-butanediol was chosen to improve the homogeneity, and

a long-chain aliphatic alcohol, dodecanol, was selected to

improve the penetrability of monolithic matrix. Further-

more, DMSO was used to improve the solubility of mono-

mers and porogens. The effect of the ratio of 1,4-butanediol

to dodecanol on the porous structure of monoliths

was investigated. With the ratio shown in column a of

Table 1, translucent monoliths with poor permeability were

obtained. However, with more dodecanol added, the

permeability of monoliths was improved, but the homo-

geneity and rigidity of matrix became poor, as shown in

columns c and d. Therefore, the optimized weight ratio of

1,4-butanediol to dodecanol was chosen as 1:1 w/w, shown

in column b.

Table 1. Compositions of the polymerization solution and

characteristic of monolithic matrix

a b c d

Compositions (g) GMA 0.0230 0.0230 0.0230 0.0230

AAm 0.0260 0.0260 0.0260 0.0260

Bis 0.0560 0.0560 0.0560 0.0560

DMSO 0.4585 0.4585 0.4585 0.4585

1,4-Butanediol 0.3059 0.2292 0.2192 0.2092

Dodecanol 0.1523 0.2293 0.2393 0.2492

AIBN 0.0010 0.0010 0.0010 0.0010

Characteristics Permeability Poor Good Good Good

Homogeneity Good Good Good Poor

Rigidity Good Good Poor Poor

Electrophoresis 2009, 30, 4034–40394036 Y. Liang et al.


3.2 Immobilization of Ampholine

According to our previous work, with monolithic poly

(GMA-co-AAm-co-Bis) as matrix, EOF generated from epoxy

groups could be ignored under electric field [14]. Therefore,

Ampholines were focused under electric field in micro-

channels with M-IPG materials to form a steady pH

gradient before immobilization. The focusing time of

Ampholines might affect the formation of pH gradient.

On the one hand, insufficient time might affect the linearity

of formed pH gradient; on the other hand, overlong time

might result in the obvious drift of pH gradient. In our

study, after ca. 5 min, the focusing current for Ampholines

stopped decreasing, and 5 min was chosen as CAs focusing

time.

As Ampholines were a complex mixtures of oligoamino

and oligacarboxylic acids, the focused Ampholines could be

bonded onto the monolith directly through the reaction

between the epoxy and amino or carboxyl groups, as shown

in Fig. 2B. Although high concentration of Ampholines

could also fasten the reaction and increase the immobilized

amount of Ampholines, it might also result in increased

current and Joule heat during focusing. Consequently,

10% w/v Ampholine was chosen to generate the immobi-

lized pH gradient via epoxy groups on monolithic matrix.

Although high temperature could accelerate the immobili-

zation, the diffusion of Ampholines happened simulta-

neously, which might degrade the formed pH gradient.

Therefore, the optimized temperature for immobilization

was selected as 601C. After the reaction for 24 h, the mate-

rials in microchannels were washed by water for 1 h to flush

out the unbound Ampholines.

Figure 1B shows the SEM image of internal morphol-

ogies of M-IPG materials prepared in microchannels, which

are tightly attached to the inner wall of microchip via the

covalent interaction between the polymer and the vinylized

surface. In addition, good homogeneity and large porous

structure are observed, which ensure uniformed pH gradi-

ent and decreased non-specific interaction between samples

and matrix during focusing.

3.3 Protein separation by lCIEF with M-IPG

materials

To perform mCIEF with M-IPG materials, the addition of

CAs in the running buffer is not necessary, as Ampholines

are attached onto the monolith surface, and the long chains

of Ampholines have the properties of ‘‘free’’ oligoamino and

oligacarboxylic acids within a small region around the

immobilized points, acting in the same role as free

Ampholines added in running buffer.

According to the theory of CIEF, high voltage could

improve the separation resolution. However, too high

voltage could lead to increased Joule heating, resulting in

bubbles formed in channels of mCIEF. After systematic

study, 1600 V was selected as the optimal voltage for the IEF

separation of FITC-labeled ribonuclease B, myoglobin and

a-casein.

After the microchannel with M-IPG materials was filled

FITC-labeled proteins, 1600 V focusing voltage was applied,

and proteins were focused according to their pIs. Within

2 min, the current stopped dropping, which indicated that the

isoelectric focusing of FITC-labeled proteins was completed.

Figure 2. Procedure for M-IPGmaterials preparation.



Both chemical and pressure mobilization were applied

to mobilize the focused protein zones to the LIF detection

point. For chemical mobilization, after replacing NaOH

(catholyte) with H3PO4 (anolyte), the focused a-casein

migrated toward the cathodic reservoir under the voltage of

2400 V. When the detection point was set just behind the

focused zone, a sharp peak was detected by LIF (Fig. 3A).

However, when the detection point was set far behind the

focused zone, the detected peak was broad (Fig. 3B), which

indicates that, by chemical immobilization, the focused

zones were broadened seriously in microchannels with

M-IPG materials due to the relatively long migration time.

In addition, pressure mobilization, performed by the

manual pressure pump via a home-made interface, was

applied as well to drive the focused zones in microchannels

with M-IPG materials. As shown in Fig. 4, ribonuclease B,

myoglobin and a-casein were well separated by mCIEF with

M-IPG materials. The sharp peaks indicated that peak

broadening could be suppressed with the existence of

monolithic matrix, and pressure mobilization is preferable

for such a separation mode. The reproducibility was also

studied. As the focused sample zones were driven by

manual pressure pump, and it was difficult to apply repea-

table pressure, the reproducibility of the migration time of

each peak was not quite good. However, the peak shape was

of good reproducibility, and the resolution of proteins was

repeatable. In two consecutive runs, the resolution for

ribonuclease B and myoglobin was, respectively, 5.7 and 6.5,

and that for myoglobin and a-casein were, respectively, 7.0

and 7.5.

Compared with the traditional IEF, CIEF or mCIEF,

mCIEF with M-IPG materials has several advantages for

proteins separation. Besides high throughput, the avoidance

of moving CAs in buffer could decrease Joule heat generated

in microchannels during focusing. In addition, the mono-

lithic matrix in microchannels could prevent anolyte or

catholyte buffer from flowing into the channel, which could

keep a stable pH gradient and prevent gradient compres-

sion. Furthermore, the diffusion and the drift of focused

zones during mobilization could also be minimized due to

the existence of monolithic matrix. Most importantly, mCIEF

with M-IPG materials could be used as the first-dimensional

separation in chip-based multi-dimensional analysis or

mCIEF-MS, since compared with the traditional mCIEF with

CAs added in the buffer, the immobilized CAs could not

interfere with further separation and detection. When stored

in water, M-IPG materials prepared in microchannels could

Figure 3. Effect of detection point on chemical immobilization offocused FITC-labeled a-casein (0.026 mg/mL) by mCIEF experi-mental conditions: IEF conditions: catholyte: NaOH (20 mmol/L),anolyte: H3PO4 (20 mmol/L); separation voltage: 1.6 kV; focusingtime: 2 min; Chemical mobilization conditions: catholyte: H3PO4(20 mmol/L), anolyte: H3PO4 (20 mmol/L); chemical mobilizationvoltage: 2.4 kV; the distance of detection point and cathodereservoir: (A) 11 mm and (B) 4 mm.

Figure 4. Separation of ribonuclease B (0.0075 mg/mL), myoglo-bin (0.022 mg/mL) and a-casein (0.026 mg/mL) by mCIEF withpressure mobilization. Experimental conditions: catholyte:NaOH (20 mmol/L), anolyte: H3PO4 (20 mmol/L); separationvoltage: 1.6 kV; focusing time: 2 min; the distance of detectionpoint and cathodic reservoir was 4 mm. Pressure mobilizationwas performed by the manual pressure pump via a home-madeinterface (described in Section 2.6).

Electrophoresis 2009, 30, 4034–40394038 Y. Liang et al.


be used for at least 20 days, and applied for the separation of

polar compounds for over 25 runs.

4 Concluding remarks

M-IPG materials were prepared in microchannles by photo

initiation, and successfully applied for microfluidic chip-

based CIEF without CAs in buffer. The experimental results

indicated that with pressure mobilization, proteins could be

separated with high resolution, high sensitivity and high

throughput. Without mobile CAs in buffer, such a separa-

tion mode is promising in further application for microchip-

based multi-dimensional CE separation and even the

hyphenation with MS.

The authors are grateful for the financial support fromNational Natural Science Foundation (20775080), NationalBasic Research Program of China (2007CB714503 and2007CB914100) and Knowledge Innovation Program ofChinese Academy of Sciences (KJCX2YW.H09).

The authors have declared no conflict of interest.

5 References

[1] Hjerten, S., Zhu, M. D., J. Chromatogr. 1985, 346,265–270.

[2] Zhang, M. Q., Rassi, Z. E., J. Proteome Res. 2006, 5,2001–2008.

[3] Zhou, F., Hanson, T. E., Johnston, M. V., Anal. Chem.2007, 79, 7145–7153.

[4] Ramsay, L. M., Dickerson, J. A., Dovichi, N. J., Electro-phoresis 2009, 30, 297–302.

[5] Hofmann, O., Che, D. Q., Cruickshank, K. A., Muller,U. R., Anal. Chem. 1999, 71, 678–686.

[6] Cui, H. C., Horiuchi, K., Dutta, P., Ivory, C. F., Anal.Chem. 2005, 77, 1303–1309.

[7] Cui, H. C., Horiuchi, K., Dutta, P., Ivory, C. F., Anal.Chem. 2005, 77,7878–7886.

[8] Yao, B., Yang, H. H., Liang, Q. L., Luo, G. A., Wang, L. D.,Ren, K. N., Gao, Y. D., Wang, Y. M., Anal. Chem. 2006,78, 5845–5850.

[9] Ou, J., Glawdel, T., Samy, R., Wang, S., Liu, Z., Ren,C. L., Rawliszyn, J., Anal Chem. 2008, 80, 7401–7407.

[10] Herr, A. E., Molho, J. I., Drouvalakis, K. A., Mikkelsen,J. C., Utz, P. T., Santiago, J. G., Kenny, T. W., Anal.Chem. 2003, 75, 1180–1187.

[11] Wang, Y. C., Choi, M. H., Han, J. Y., Anal. Chem. 2004,76, 4426–4431.

[12] Li, Y., Buch, J. S., Rosenberger, F., DeVoe, D. L., Lee,C. S., Anal. Chem. 2004, 76, 742–748.

[13] Emrich, C. A., Medintz, I. L., Chu, W. K., Mathies, R. A.,Anal Chem. 2007, 79, 7360–7366.

[14] Das, C., Zhang, J., Denslow, N. D., Fan, Z. H., Lab Chip2007, 7, 1806–1812.

[15] Yang, S., Liu, J., Lee, C. S., Devoe, D. L., Lab Chip 2009,9, 592–599.

[16] Macounova, K. I., Catherine, R. C., Yager, P., Anal.Chem. 2001, 73, 1627–1633.

[17] Catherine, R. C., Finlayson, B., Yager, P., Anal. Chem.2001, 73, 658–666.

[18] Huang, T., Pawliszyn, J., Electrophoresis 2002, 23,3504–3510.

[19] Kates, B., Ren, C. L., Electrophoresis 2006, 27,1967–1976.

[20] Tang, G., Yang, C., Electrophoresis 2008, 29, 1006–1012.

[21] May, E. L., Hillier, A. C., Anal. Chem. 2005, 77,6487–6493.

[22] Bjellqvist, B., Ek, K., Righetti, P. G., Gianazza, E.,Westermeier, R., Postel, W. J., J. Biochem. Biophys.Meth. 1982, 6, 317–339.

[23] Sommer, G. J., Singh, A. K., Hatch, A. V., Anal. Chem.2008, 80, 3327–3333.

[24] Yang, C., Zhu, G. J., Zhang, L. H., Zhang, W. B., Zhang,Y. K., Electrophoresis 2004, 25, 1729–1734.

[25] Zhu, G. J., Yang, C., Zhang, L. H., Liang, Z., Zhang,W. B., Zhang, Y. K., Talanta 2006, 70, 2–6.

[26] Zhu, G. J., Yuan, H. M., Zhao, P., Zhang, L. H., Liang, Z.,Zhang, W. B., Zhang, Y. K., Electrophoresis 2006, 27,3578–3583.

[27] Jia, Z. J., Fang, Q., Fang, Z. L., Anal. Chem. 2004, 76,5597–5602.

[28] Cong, Y., Davey, M. H., Svec, F., Frechet, J. M. J., Anal.Chem. 2001, 73, 5088–5096.

[29] Hatch, A. V., Herr, A. E., Throckmorton, D. J., Brennan,J. S., Singh, A. K., Anal. Chem. 2006, 78, 4976–4984.



Documents

Microchip isoelectric focusing with monolithic immobilized pH gradient materials for proteins separation