Full Paper
Agarose-Assisted Micro-Contact Printing for High-Quality Biomolecular Micro-Patternsa Min Jee Jang, Yoonkey Nam* Micro-contact printing has been developed to print biomolecules, such as cell adhesive molecules, proteins, or DNAs, on a substrate, which can serve as experimental platforms for investigating biological issues and engineering biosensors. Despite the popularity of this method, it has been technically challenging to use a conventional stamp made of a hydrophobic polydimethoxysilane (PDMS) elastomer that often requires surface treatments to facilitate the inking and stamping of biomolecules. In this work, we proposed a new surface modification method for a PDMS stamp using agarose hydrogel and demonstrated the applications to the design of micro-patterned substrates with biomolecules. By using a simple bench-top dip-coating method with a commercial syringe pump to steadily pull out the stamp from boiled agarose solution, we coated an agarose layer on the stamp. It consequentially enhanced the transferability of ink molecules to the target substrate and the uniformity of printed patterns compared to the traditional methods for treating stamp surface such as surfactant coating and temporary oxidation with air plasma. In addition, this microstamping method was also used to produce patterns of proteins with the preservation of bioactivity, which could guide neuronal growth. Thus, we demonstrated the applicability to the interface designs of biochips and biosensors.
Micro-contact printing (mCP) has been widely used for generating micro-scale patterns of various molecules on the surface. As this method could selectively print organic selfassembled monolayers,[1] polymers,[2] or nanomaterials, [3–5] spatial control of surface characteristics, such as wettability, has been available.[6] In addition, facile fabrication of functional substrates such as plasmonic sensors[7] or electronics with conductive materials[3–5,8] have been also recently reported using mCP. But most of all, this micropatterning technique has been highlighted M. J. Jang, Y. Nam Department of Bio and Brain Engineering, KAIST, 291 Daehak-ro, Yuseong-Gu, Daejeon 305-701, Republic of Korea E-mail: [email protected] a
Supporting Information is available from the Wiley Online Library or from the author.
ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
because of its ability to pattern biomolecules, thus enabling the design of cellular microenvironments in biological experiments and tissue engineering applications. The geometric constraints confined by micropatterns could influence cellular morphology,[9] differentiation,[10] migration,[11] proliferation,[12] and development.[13,14] In addition, mimicking biological environment between different types of cells has been also achieved to facilitate the investigation of cell–cell interactions in cultures.[15,16] Applying the micropatterning technique to scaffold materials has opened the possibility of implementing the asymmetric composition of cells or proteins in artificial tissue structures.[12,17] Despite the versatility of mCP, it has been still challenging to use a bare stamp typically made of polydimethoxysilane (PDMS) for inking biomolecules, e.g., proteins[9,18] or DNAs,[19] as the PDMS surface is hydrophobic; the poor wettability of PDMS causes insufficient inking of hydrophilic molecules, consequently resulting in the poor
wileyonlinelibrary.com1
1
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407
R
1. Introduction
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
transferability to the substrate.[20] In order to solve the problem, two main approaches have been used for modifying the stamp surface. First, treating of O2/air plasma oxidized the PDMS surface so that the binding force between hydrophilic biomolecules and the stamp surface became strong.[21] The effect of surface modification, however, was temporary due to the hydrophobic recovery.[20] Chemical immobilization of hydrophilic moieties on the PDMS surface was alternatively introduced as a semipermanent scheme of surface modification.[22] In some cases, however, the hydrophilic surface of the PDMS stamp hampered the release of ink molecules. Tan et al. reported the -CF3 treated PDMS stamp, which was more hydrophobic than a bare PDMS stamp, transferred more protein molecules to the same conditioned substrate rather than the -NH2 treated hydrophilic PDMS stamp.[22] Second, surface coating with amphiphilic surfactant was suggested as another surface modification method. Chang et al. carefully compared the transferability of PDMS stamps modified with various amphiphilic molecules, including sodium dodecyl sulfate (SDS), Triton, and centrimonium bromide (CTAB), with a bare PDMS stamp and showed that SDS-coated PDMS stamps enhanced the transfer of biomolecules to the substrate.[23] Although the surface modification with SDS was useful to deliver biomolecules and lasted much longer than plasma treatment, it was concerned about the denaturation of ink molecules in case of proteins.[23] Additionally, in our experiences, SDS did increase the inking and transferability, but non-uniform coating of SDS on the stamp surface often leads to the nonuniform ‘spotty’ prints. Our motivation was to develop another way to coat the stamp surface that could enhance both the transferability of ink molecules and uniformity of printed patterns. Here we chose the hydrophilic hydrogel as a candidate material. Among various types of hydrogel, agarose has been used as a stamp material itself in different biological or engineering applications. Agarose stamps were foremost used for multiple patterning without re-inking steps.[24] The ability of agarose to retain ink solution further enabled mammalian cell patterning to porous scaffold directly.[25] In other applications, agarose stamps were used as a reservoir of etchant for fabricating a topographical surface of metal.[26] Conductive polymer[27] or nanoparticles[28] was also capable of being patterned with agarose stamps. The advanced method that used an agarose stamp combined with electrodes was demonstrated the ability of addressable patterning.[29] Despite such successful demonstrations of agarose hydrogel as a stamp material, hydrogel stamps have been still not easy to use in practice since their low stiffness makes themselves difficult to handle. In this work, we propose a new microstamping method for printing biomolecules using a simple bench-top method of agarose coating on PDMS stamps, and quantitatively
2
2
R
2
compared the performance of our method to the conventional mCP with different surface modification schemes. By applying dip-coating method, which has been usually used to fabricate thin film on the surface, we coated agarose hydrogel on a PDMS surface. The agarose thin film with the thickness of a few micrometer enhanced the transferability of ink molecules and the uniformity of printed patterns. In addition, the bioactivity of proteins was still preserved after being patterned, and cells (neurons) grew following the underlying patterns printed with our method.
2. Materials and Methods 2.1. Materials Two types of SU8 photoresist (2002 for 2 mm-thickness and 2050 for 50 mm-thickness; Microchem, MA, USA) and (1H, 1H, 2H, 2H-perfluorooctyl) trichlorosilane (448931, Aldrich, MO, USA) were used for fabricating a silicon mold. A Sylgard 184 kit (Dow Corning, MI, USA) containing prepolymer and curing agent was used for manufacturing PDMS stamps. Acetone, 2-propanol (Junsei Chemical Co., Ltd., Japan), and phosphate buffered saline (PBS, 10010–023, Gibco, CA, USA) were used as received. Glassware, including coverslips or microscope slides, manufactured from Mrienfeld (Germany) was used. Agarose ITM (0710, Amresco, OH, USA) and SDS (L3771, Sigma, MO, USA) were used for coating PDMS stamps. As ink molecules, poly-l-lysine-conjugated FITC (PLL-FITC; P3069, Sigma, MO, USA), synthetic A chain of laminin (C6171, Sigma, MO, USA), natural mouse laminin (23017–015, Invitrogen, CA, USA), and fibronectin from human plasma (F2006, Sigma, MO, USA) were used. Antilaminin (L9393, Sigma, MO, USA), anti-fibronectin (F3648, Sigma, MO, USA), and Alexa Fluor 594 (A11012, Invitrogen, CA, USA) were used for labeling printed proteins. For cell cultures, Hank’s balanced salt solution (HBSS, Welgene, Daegu, Republic of Korea), Neurobasal medium (21103–049, Gibco, CA, USA), B27 (17504–044, Gibco, CA, USA), 2mM GlutaMAX (35050–061, Gibco, CA, USA), lglutamate (G8415, Sigma, MO, USA), and penicillin–streptomycin (15140–122, Gibco, CA, USA) were used.
2.2. Preparation of PDMS Stamps PDMS stamps were fabricated by soft-lithography process described in the previous work.[15] Briefly, a silicon mold patterned with SU8 photoresist was coated with (1H, 1H, 2H, 2H-perfluorooctyl) trichlorosilane to serve as an antistiction layer. The mixture of PDMS prepolymer and curing agent with the ratio of 10:1 was poured on the SU-8 mold and degassed in a vacuum desiccator. After curing in an oven at 60 8C for at least 2 h, cured PDMS was cut into pieces
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
with the size of 1 cm 1 cm. PDMS stamps were cleaned via ultrasonication in acetone, 2-propanol (IPA), and 2nd distilled water for 5 min each.
2.3. Surface Modification Procedures For creating agarose thin film on a PDMS stamp, we used a dip-coating method; a bare PDMS stamp attached on a microscope slide was dipped in hot agarose solution (0.5%) and steadily pulled out. The microscope slide was attached to a push block of a syringe pump (KDS-200, KDScientific, MA, USA) so that it can be pulled out from the solution at constant speed (10 cm h 1). Two different surface modification strategies (surfactant coating or oxidation by plasma) were also used; for surfactant coating, clean PDMS stamps were soaked in 10% SDS solution for 15 min. For the first 5 min, ultrasonication was used to facilitate the coating process. Then stamps were washed with distilled water once to remove any excessive SDS layer and dried with compressed air. For temporary oxidation, clean stamps were treated with air plasma (30W, Cute, Femtoscience, Korea) for 1 min. All stamps were immediately used for inking and stamping.
PDMS surfaces and patterned substrates were taken with the same imaging parameters (1 s exposure, ISO 400). For the quantitative measurement of fluorescence intensities, five field-of-views (FOVs) were randomly selected from each sample. The mean and standard deviation of fluorescence images were calculated using Matlab (Mathworks, Inc., USA). An inverted microscope (IX71, Olympus, Japan) with 10x or 20x objectives and a CCD camera (DP71, Olympus, Japan) were used for taking the fluorescence images of printed patterns.
2.5. Immunoassay To confirm the printing of human fibronectin (0.1 mg ml 1 in PBS) and laminin (0.1 mg ml 1 in PBS), primary antibodies (anti-fibronectin and anti-laminin, Sigma, MO) diluted in PBS (1:500) were treated for 1 h in a 37 8C incubator. After substrates were washed several times with PBS, secondary antibodies diluted in PBS (1:100) were also treated to each sample for 1 h in a 37 8C incubator. Samples were washed again with fresh PBS and fluorescence images of each sample were taken with an epifluorescence microscope (Olympus IX71).
2.4. Thin Film Characterization
2.6. Cell Culture
For the characterization of the surface-modified PDMS, contact angles were measured using a Phoenix 300 goniometer (Surface Electro Optics, Co., Ltd., Republic of Korea) with a drop of distilled water (7 mL). In addition, for characterizing the ability of the PDMS surface as an ink reservoir, we measured contact angles again after loading a droplet of water on each PDMS surface for 15 min. Each contact angle value was manually traced with ImageJ software. For estimating the thickness of agarose thin film on a PDMS stamp, we used z-stack imaging of the confocal microscope (Nikon C2, Japan). To image the agarose film the agarose solution was mixed with PLL-FITC (0.05 mg ml 1) and coated on a flat PDMS stamp. We then acquired the normalized fluorescence profile of the film from the orthogonal view of z-stack images and estimated thickness of the film as the full-width at half maximum of this profile.
Hippocampi were microsurgically separated from E18 Sprague-Dawley (SD) rat (Koatech, Republic of Korea). Dissected hippocampi immersed in HBSS were dissociated with pipetting and centrifuged at 1000 rpm for 2 min. Supernatant was then removed, and settled cells were resuspended in plating medium (Neurobasal medium supplemented with B27, 2 mM GlutaMAX, 12.5 mM lglutamate, and penicillin–streptomycin). After being sieved through a cell strainer (BDFalcon, NJ), cells were plated on a patterned substrate and filled with plating medium in 30 min. Half of the medium was changed twice a week with maintenance medium (plating medium without l-glutamate). All procedures were done according to approved animal use protocols of the KAIST Institutional Animal Care and Use Committee (IACUC).
3. Results
PLL-FITC (0.1 mg ml 1 dissolved in 3rd distilled water) was used as an ink material to quantify the amount of ink present on a PDMS surface and the substrate. The fluorescence intensity of PLL-FITC loaded on the stamp surface was measured after 30 min of inking and 1 min of stamping. The fluorescence images of the same position of
www.MaterialsViews.com
3.1. Fabrication and Characterization of AgaroseCoated Stamps Agarose thin film layer was created on a PDMS stamp surface by dip-coating process. We installed a commercial syringe pump in a vertical direction to implement dipcoating (Supplementary figure 1); a PDMS stamp was attached to a dummy microscope slide and dipped into a
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3 3
3
R
2.5. Fluorescence Measurement
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
fresh agarose hydrogel solution. Once the syringe pump pushed the plunger with a constant speed, it pulled out the stamp with constant speed and resulted in a thin and uniform coating of the agarose solution on the stamp. The thin layer of agarose hydrogel dried immediately after pulling. To confirm the feasibility of the dip-coating method to generate thin film of agarose hydrogel, we compared the contact angle of the agarose-coated (AC) PDMS surface with two different control groups: bare PDMS, water-coated (WC) PDMS (Figure 1Aa). WC-PDMS was prepared by a similar procedure as used for the experimental group except that agarose precursor was not added in the solvent (boiled distilled water). As shown in Figure 1Ab, the contact angle of agarose-coated (AC) PDMS was not significantly different from bare or WC-PDMS. However, the contact angle of ACPDMS apparently altered after loading a drop of water for 15 min, while that of a bare PDMS did not show any significant change (Figure 1Ac). We speculated that the thin agarose film was dried before the first measurement, which could cause the similar wettability of agarose-coated PDMS to bare one. Then for 15 min of droplet loading, the film might absorb water, thereby becoming hydrophilic. Consequently, the significant change of contact angle indicated the formation of thin agarose layer, which could convert PDMS surface to be more hydrophilic overtime. The thickness of agarose thin film formed on PDMS surface was determined by z-stack imaging with a confocal microscope (Figure 1Ba). We acquired the average intensity profile of z-axis form the orthogonal view of 3D image and estimated the thickness as the full width at half maximum of this profile (Figure 1Bb). The thickness of wet agarose thin film on PDMS surface was estimated to be 3.59 0.38 mm (mean SEM; n ¼ 3).
3.2. Agarose-Assisted mCP Process To apply the agarose dip-coating as the surface modification method of PDMS stamps to mCP, we devised a new microstamping procedure (Figure 2A); a cleaned PDMS stamp was coated with the agarose hydrogel by dip-coating and the stamp was subsequently inked with biomolecules for 30 min. The inked stamp surface was cleaned by blowing off the extra inking solution with an air-gun, and the stamp was in contact with the substrate to transfer ink molecules for 10 min. Figure 2B shows three different micropatterns of biomolecules printed by the new procedure: small dot arrays with diameter of 3 or 5 mm and spacing of 5 mm, large dot array with a diameter of 100 mm and spacing of 200 mm, and micro-grid (node diameter: 50 mm; node spacing: 200 mm; line width: 5 mm). Concerned with the distortion of patterns due to the low stiffness of agarose hydrogel, we
4
4
R
4
compared the sizes of patterns (dot arrays with the diameters of 3 mm and 100 mm) fabricated on the PDMS stamp and printed on the surface with our method (Table 1). Although the size of 3 mm- and 100 mm-dot patterns was slightly larger than those of stamps (110.1% and 100.9%, respectively), the statistical comparison was not significantly different. These results indicated that agarosecoated stamps were suitable for printing micro-scale patterns with the wide range of feature sizes with high accuracy.
3.3. Evaluation of Agarose-Assisted mCP Next, we investigated the effect of agarose thin film on the quality of printed patterns. Compared to the bare or modified PDMS stamps with different strategies (SDS coating or plasma treatment), we evaluated the ability of agarose thin film from two aspects; how much ink molecules can be transferred from the stamp to the surface (transferability) and how uniform the printed pattern is (uniformity). All procedures (inking and stamping) except the surface modification step were same. For the first parameter, we assessed the relative ink transferability of each stamp by measuring the ratio of mean fluorescence intensity of a printed substrate to that of a stamp (Figure 3Aa and Ab); the higher the ratio was, the more transferable the ink molecules were. Among the conventional methods, the SDS-assisted method showed the highest transferability, which was consistent with the previous report.[23] Our agarose-assisted method showed comparable transferability (81.3 0.05%) compared with the SDS-assisted method (88.0 0.05%). In addition, our method transferred ink molecules significantly more than other two methods (32.6 0.01% for plasma-treated and 66.4 0.05% for bare PDMS stamps). Therefore, it implied that agarose thin film was effective in achieving high efficiency in transferring ink molecules from the PDMS stamp to the substrate. Figure 3Ac shows the uniformity of stamped ink molecules on the substrates. We used flat PDMS stamps to print fluorescently labeled biomolecules (PLL-FITC) in a large area (1 mm 1 mm), and measured the fluorescence intensity from five different FOVs. Then, coefficient of variation (CV) of the intensity in each FOV was measured, and the reciprocal of this value was considered as the uniformity. As the intensity distribution of uniform patterns was less variable, the uniformity value would be large. Compared to a bare stamp (2.64 0.23), surfacemodified stamps with SDS-coating (3.65 0.25) or plasmatreatment (3.57 0.31) produced more uniform patterns. However, agarose thin film outperformed other methods and produced almost twofold more uniform patterns (7.11 0.51) than other methods. Consequently, our
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
Figure 1. Characterization of agarose-coating on PDMS stamps. (A) Contact angle of PDMS substrate with different coating. (Aa) A bare PDMS surface and three different schemes of surface coating, such as water-coated (‘‘WC’’), and agarose-coated (‘‘AC’’) were used to measure the contact angle of a modified surface. (Ab) The contact angle of all conditions (top images in (Aa)) was not significantly different. (Ac) After 15 min of water loading (bottom images in (Aa)), the contact angle of agarose-coated PDMS (‘‘AC’’) was significantly decreased, while that of bare PDMS was not changed (mean SEM; n ¼ 5 samples; ***p < 0.001, ****p < 0.0001, ns: no significance). (B) Thickness measurement of agarose thin film generated on flat PDMS surface. (Ba) 3D (left) and its orthogonal view (right) of a z-stack image (interval: 1 mm) of fluorescently visualized agarose thin film. (Bb) Average fluorescence profile of the orthogonal image (yellow line in (Ba) indicates xaxis of this profile).
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
5 5
5
R
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
Figure 2. Schematic procedure of agarose-assisted mCP and patterned substrates. (A) A PDMS stamp was dip-coated with agarose solution at first (‘‘Pulling’’), then a drop of biomolecular solution was loaded on the stamp for 30 min (‘‘Loading’’). Subsequently, the stamp was attached on the substrate for 1 min to print patterns (‘‘Contact printing’’). (B) Various designs of micropatterns composed of PLL-FITC were printed by our method: small dot arrays (‘‘3 mm dots’’ and ‘‘5 mm dots’’; spacing of 5 mm), large dot arrays (‘‘100 mm dots’’; spacing of 200 mm), and micro-grid (‘‘grid’’; node diameter: 50 mm; node spacing: 200 mm; line width: 5 mm). Scale bars indicate 5 mm in ‘‘3 mm dots’’/ ‘‘5 mm dots’’ and 100 mm in ‘‘100 mm dots’’/‘‘grid,’’ respectively.
agarose-assisted mCP was capable of not only producing micro-scale patterns that were considerably uniform in large areas, but also enhancing the number of molecules transferred to the substrate. Next, we tested whether repetitive printing was possible with the agarose film coating. Figure 3B shows the micropattern of large dot arrays. The same area of patterns repeatedly printed with the same stamp on the different substrates was shown in Figure 3Ba. The fluorescence intensity of patterns was vivid in the first stamping. In spite of bare visibility, the patterns were also successfully generated on the substrates by second to fifth stamping (80% increment of image brightness could show printed patterns). However, the mean fluorescence intensity of each stamping event indicated the steep decrease of transferred molecules relative to the first stamping (Figure 3Bb); second stamping could transfer only 35% of
Table 1. Comparison of pattern size.
Design
Stamp Printed pattern
3 mm-Diameter dot
100 mm-Diameter dot
2.7 0.0 3.1 0.2a
95.3 0.8 96.2 0.6a
*mean SEM; n ¼ 5. a The values were not significantly different under p < 0.05.
6
6
R
6
molecules compared with the first stamping. Additionally, the amount of transferred molecules gradually decreased as the stamping event increased. These results suggested that our method was capable of repetitive patterning with limitation of inconsistent transferability. To investigate the diffusion of ink molecules outside the pattern area, we measured the size of printed patterns (large dot arrays with diameter of 100 mm) with different stamping time (Figure 3C). The size of micropatterns was not correlated to stamping time, suggesting that the diffusion of ink molecules was negligible.
3.4. Patterning Proteins and Cells To demonstrate the utility of the proposed stamping method, we applied our microstamping method to create micropatterns of proteins or cell-guiding cues. We first confirmed the preservation of protein activity by immunolabeling printed patterns. Figure 4 shows the fluorescence images of two patterned extracellular matrix proteins, human fibronectin and laminin, labeled with sequential reactions of primary antibody to each protein and secondary antibody conjugated with fluorescence markers. Both proteins were labeled with red fluorescence markers, indicating that printed patterns preserved their reactivity. For more accurate corroboration, the same protein patterns printed with the original and modified procedure of agarose-assisted mCP was compared
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
Figure 3. Characterization of micropatterns printed with agarose-assisted mCP. (A) Quantitative comparison of pattern quality (ink transferability and pattern uniformity) using our method with other surface modification strategies. (Aa) Large dot array patterns (diameter: 50–100 mm) printed with the same procedure of mCP using different PDMS stamps: agarose coating (‘‘Agarose’’), SDS coating (‘‘SDS’’), plasma treatment (‘‘plasma’’), and bare (without any surface modification; ‘‘bare’’). Ink transferability (Ab) and pattern uniformity (Ac) of each method was quantitatively analyzed (mean SEM; n ¼ 15; *p < 0.05, ***p < 0.001, ns: no significance). A scale bar in ‘‘Agarose’’ of (Aa) indicates 200 mm. (B) Repetitive patterning with an agarose-coated PDMS stamp. (Ba) Same regions of patterns repeatedly printed on different substrates. All images in ‘‘Original’’ was captured with the same imaging condition. (Bb) Mean fluorescence intensity of pattern areas over the number of stamping events relative to that of patterns in first stamping. (C) Quantitative analysis of pattern sizes depending on the stamping time (mean SEM; not significantly correlated under Spearman correlation test).
www.MaterialsViews.com
fluorescence intensity (Supplementary figure 2A and 2B). Considering the similar transferability of two methods (Supplementary figure 2C), the results indicated that our method could successfully generate active protein patterns. The preservation of protein activity was also confirmed by cell cultivation. We cultured primary hippocampal
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
7 7
7
R
(Supplementary figure 2). In the modified procedure, proteins (laminin) were pre-mixed in boiled agarose solution before dip-coating so that temperature sensitive proteins might lose their reactivity. Compared to the protein patterned in harsh condition (premixed), one patterned with our method showed significantly higher
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
Figure 4. Patterning active proteins and neurons. Extracellular matrix proteins (human fibronectin and laminin) were printed with agaroseassisted mCP and subsequently labeled with fluorescence-conjugated antibodies (‘‘Protein’’). Cultured hippocampal neurons (‘Tuj1’; fixed at 4 DIV) were adhered and grew on theses patterned substrates, but only laminin patterns were useful in guiding neurons (‘‘Merged’’). The scale bar indicates 200 mm.
neurons from E18 rat on these protein-patterned substrates (middle of Figure 4). On the micro-gird patterns composed of laminin, neurons preferentially adhered on nodes and extended their neurites following thin lines at 4 days in vitro (DIV). On the other patterned substrate with the same designs but different proteins (human fibronectin), on the other hand, neurons could not recognize the underlying patterns (right of Figure 4). The results suggested the surface-printed protein patterns preserved their activity sufficient to lead neuronal guidance and selective recognition to microenvironment.
4. Discussion In this work, we proposed a new surface modification of PDMS stamps with agarose hydrogel for advanced mCP. Our dip-coating method using a commercial syringe pump was simple and straightforward to execute. Adding the coating step to conventional mCP, we also demonstrated the ability of agarose thin film to improve the patterning quality (transferability and uniformity) and to implement active protein patterns especially for designs of the neuronal microenvironment. For past decades, several strategies for the surface modification of PDMS stamp have been introduced to improve the quality of biomolecular patterns. However, it was difficult to satisfy two essential requirements: patterning efficiency and pattern uniformity. The plasma
8
8
R
8
treatment of PDMS stamps for temporal oxidization has been the standard technique for converting hydrophobic stamp surface into hydrophilic surface for loading and transferring biomolecular inks.[21] According to our result this method produced more uniform patterns, but the amount of ink molecules delivered to the substrate was less than that from a bare stamp. The oxidized PDMS surface might be too strong to release ink molecules due to stronger wettability than glass substrates.[20,31] Chang et al. suggested another strategy with SDS-coating on PDMS stamps, which had advantages over efficient transfer of ink molecules.[23] We also obtained a similar result using the SDS-coated, while it was observed that the printed patterns had sporadic hot spots where fluorescence intensity was exceptionally higher than surrounding areas owing to the aggregation of ink molecules, which might result in less uniformity. Compared with the conventional strategies, our agarose-coating method achieved both high transferability and uniformity at the same time. Our agarosecoated stamps were quite different from the agarose stamp that is made of agarose hydrogel. Mayer et al. suggested the several advantages of agarose hydrogel stamps for biomolecular printing, especially the possibility of consecutive printing without re-inking.[24] As our method could not repeatedly produce the patterns with the same quality, our agarose thin film might be too thin to serve as a reservoir for the ink solution. The uniformity of surface patterns is important to generate reliable conditions in a variety of experimental
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
5. Conclusion Here we proposed the agarose-assisted mCP method for patterning biomolecules. A thin agarose layer was generated on hydrophobic PDMS surface by means of the dipcoating method; it promoted the transfer of ink molecules to the substrate and rendered the printed pattern considerably uniform in a large area. As a consequence, the simple modification of a PDMS stamp with agarose hydrogel improved the quality of printed biomolecular patterns, thereby contributing to execute reliable experimentations and reduce the opportunity costs in biochip applications.
Acknowledgement: This research was supported by the National Research Foundation of Korea (NRF-2011–0019213, NRF2012R1A2A1A01007327), funded by the Ministry of Science, ICT & Future Planning (MSIP).
Received: September 12, 2014; Revised: November 14, 2014; Published online:DOI: 10.1002/mabi.201400407 Keywords: agarose hydrogel; cell chip; DNA chip; micropatterning; protein chip; soft-lithography
[1] A. Kumar, G. M. Whitesides, Appl. Phys. Lett. 1993, 63. [2] P. Xiao, J. Gu, J. Chen, D. Han, J. Zhang, H. Cao, R. Xing, Y. Han, W. Wang, T. Chen, Chem. Commun. 2013, 49, 11167. [3] G. Liu, W. Guo, D. Song, Biosens Bioelectron. 2014, 52, 360. [4] H.-S. Shin, H.-J. Yun, K.-H. Baek, T.-H. Ham, K.-S. Park, D.-P. Kim, G.-W. Lee, H.-D. Lee, K. Lee, L. M. Do, J. Nanosci, Nanotech 2012, 12, 5325.
www.MaterialsViews.com
[5] Y.-M. Chien, F. Lefevre, I. Shih, R. Izquierdo, Nanotechnology 2010, 21, 134020. [6] M. Ghezzi, S. C. Thickett, C. Neto, Langmuir 2012, 28, 10147. [7] K. Bhatnagar, A. Pathak, D. Menke, P. V. Cornish, K. Gangopadhyay, V. Korampally, S. Gangopadhyay, Nanotechnology 2012, 23, 495201. [8] H.-J. Yun, Y.-H. Ham, H.-S. Shin, K.-S. Jeong, J.-G. Park, D.-S. Choi, G.W. Lee, J. Nanosci. Nanotech. 2011, 11, 5640. [9] N. Gadhari, M. Charnley, M. Marelli, J. Brugger, M. Chiquet, Biochim. Biophys. Acta 2013, 1833, 3415. [10] P. Bajaj, B. Reddy, L. Millet, C. Wei, P. Zorlutuna, G. Bao, R. Bashir, Integr. Biol. 2011, 3, 897. [11] M. C. Leong, S. R. K. Vedula, C. T. Lim, B. Ladoux, Commun. Integr. Biol. 2013, 6, e23197. [12] E. T. Baran, K. Tuzlakolu, A. Salgado, R. L. Reis, J. Tissue Eng. Regen. Med. 2011, 5, e108. [13] M. J. Jang, Y. Nam, J. Neural Eng. 2012, 9, 046019. [14] W. R. Kim, M. J. Jang, S. Joo, W. Sun, Y. Nam, Lab Chip 2014, 14, 799. [15] M. E. Boggs, W. R. Thompson, M. C. Farach-Carson, R. L. Duncan, T. P. Beebe Jr., Biointerphases 2011, 6, 200. [16] D. Liazoghli, A. D. Roth, P. Thostrup, D. R. Colman, ACS Chem Neurosci. 2012, 3, 90. [17] X. Tang, M. Y. Ali, M. T. A. Saif, Soft Matter. 2012, 8, 7197. [18] A. Bernard, J. P. Renault, B. Michel, H. R. Bosshard, E. Delamarche, Adv. Mater. 2000, 12, 1067. [19] S. A. Lange, V. Benes, D. P. Kern, J. K. H. Hºrber, A. Bernard, Anal. Chem. 2004, 76, 1641. [20] T. Kaufmann, B. J. Ravoo, Polym. Chem. 2010, 1, 371. [21] C. D. James, R. C. Davis, L. Kam, H. G. Craighead, J. N. Turner, W. Shain, Langmuir 1998, 14, 741. [22] J. L. Tan, J. Tien, C. S. Chen, Langmuir 2002, 18, 519. [23] J. C. Chang, G. J. Brewer, B. C. Wheeler, Biomaterials 2003, 24, 2863. [24] M. Mayer, J. Yang, I. Gitlin, D. H. Gracias, G. M. Whitesides, Proteomics 2004, 4, 2366. [25] M. M. Stevens, M. Mayer, D. G. Anderson, D. B. Weibel, G. M. Whitesides, R. Langer, Biomaterials 2005, 26, 7636. [26] S. K. Smoukov, K. J. M. Bishop, R. Klajn, C. J. Campbell, B. A. Grzybowski, Adv. Mater. 2005, 17, 1361. [27] T. S. Hansen, K. West, O. Hassager, N. B. Larsen, Adv. Mater. 2007, 19, 3261. [28] Y.-C. Kao, F. C.-N. Hong, Nanotechnology 2011, 22, 185303. [29] S. Sekine, S. Nakanishi, T. Miyake, K. Nagamine, H. Kaji, M. Nishizawa, Langmuir 2010, 26, 11526. [30] M. J. Jang, Y. Nam, BioChip J 2012, 6, 107. [31] G. Csucs, T. Kuºnzler, K. Feldman, F. Robin, N.?D. Spencer, Langmuir 2003, 19, 6104. [32] S. K. W. Dertinger, X. Jiang, Z. Li, V. N. Murthy, G. M. Whitesides, Proc. Natl. Acad. Sci. USA 2002, 99, 12542. [33] A. C. Philipsborn von, S. Lang, J. Loeschinger, A. Bernard, C. David, D. Lehnert, F. Bonhoeffer, M. Bastmeyer, Development 2006, 133, 2487. [34] R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, T. A. Witten, Nature 1997, 389, 827. [35] J. M. Moran-Mirabal, H. G. Craighead, C. P. Tan, R. N. Orth, E. O. Williams, D. M. Lin, Anal. Chem. 2007, 79, 1109.
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
9 9
9
R
designs. For example, the different distribution of molecules within or across the patterns could influence cellular behavior in different ways as cells could recognize the concentration gradient of surface-bound molecules.[32,33] In case of microarrays, the fabrication of uniform surface spots has been technically challenging as the printed spots usually formed ‘‘coffee rings’’ due to the capillary flow;[34] as the amount of target molecules bound to the complementary sequences printed on the surface could be varied, the microarray data could be noisy.[35] In such circumstances, agarose-assisted mCP technique could reduce experimental errors or batch effects. The improved transferability of a stamp would be beneficial to reduce the quantity of biomolecules required for creating patterns since most of the biomolecules including proteins or antibodies were expensive.
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing for High-Quality Biomolecular Micro-Patternsa Min Jee Jang, Yoonkey Nam* Micro-contact printing has been developed to print biomolecules, such as cell adhesive molecules, proteins, or DNAs, on a substrate, which can serve as experimental platforms for investigating biological issues and engineering biosensors. Despite the popularity of this method, it has been technically challenging to use a conventional stamp made of a hydrophobic polydimethoxysilane (PDMS) elastomer that often requires surface treatments to facilitate the inking and stamping of biomolecules. In this work, we proposed a new surface modification method for a PDMS stamp using agarose hydrogel and demonstrated the applications to the design of micro-patterned substrates with biomolecules. By using a simple bench-top dip-coating method with a commercial syringe pump to steadily pull out the stamp from boiled agarose solution, we coated an agarose layer on the stamp. It consequentially enhanced the transferability of ink molecules to the target substrate and the uniformity of printed patterns compared to the traditional methods for treating stamp surface such as surfactant coating and temporary oxidation with air plasma. In addition, this microstamping method was also used to produce patterns of proteins with the preservation of bioactivity, which could guide neuronal growth. Thus, we demonstrated the applicability to the interface designs of biochips and biosensors.
Micro-contact printing (mCP) has been widely used for generating micro-scale patterns of various molecules on the surface. As this method could selectively print organic selfassembled monolayers,[1] polymers,[2] or nanomaterials, [3–5] spatial control of surface characteristics, such as wettability, has been available.[6] In addition, facile fabrication of functional substrates such as plasmonic sensors[7] or electronics with conductive materials[3–5,8] have been also recently reported using mCP. But most of all, this micropatterning technique has been highlighted M. J. Jang, Y. Nam Department of Bio and Brain Engineering, KAIST, 291 Daehak-ro, Yuseong-Gu, Daejeon 305-701, Republic of Korea E-mail: [email protected] a
Supporting Information is available from the Wiley Online Library or from the author.
ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
because of its ability to pattern biomolecules, thus enabling the design of cellular microenvironments in biological experiments and tissue engineering applications. The geometric constraints confined by micropatterns could influence cellular morphology,[9] differentiation,[10] migration,[11] proliferation,[12] and development.[13,14] In addition, mimicking biological environment between different types of cells has been also achieved to facilitate the investigation of cell–cell interactions in cultures.[15,16] Applying the micropatterning technique to scaffold materials has opened the possibility of implementing the asymmetric composition of cells or proteins in artificial tissue structures.[12,17] Despite the versatility of mCP, it has been still challenging to use a bare stamp typically made of polydimethoxysilane (PDMS) for inking biomolecules, e.g., proteins[9,18] or DNAs,[19] as the PDMS surface is hydrophobic; the poor wettability of PDMS causes insufficient inking of hydrophilic molecules, consequently resulting in the poor
wileyonlinelibrary.com1
1
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407
R
1. Introduction
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
transferability to the substrate.[20] In order to solve the problem, two main approaches have been used for modifying the stamp surface. First, treating of O2/air plasma oxidized the PDMS surface so that the binding force between hydrophilic biomolecules and the stamp surface became strong.[21] The effect of surface modification, however, was temporary due to the hydrophobic recovery.[20] Chemical immobilization of hydrophilic moieties on the PDMS surface was alternatively introduced as a semipermanent scheme of surface modification.[22] In some cases, however, the hydrophilic surface of the PDMS stamp hampered the release of ink molecules. Tan et al. reported the -CF3 treated PDMS stamp, which was more hydrophobic than a bare PDMS stamp, transferred more protein molecules to the same conditioned substrate rather than the -NH2 treated hydrophilic PDMS stamp.[22] Second, surface coating with amphiphilic surfactant was suggested as another surface modification method. Chang et al. carefully compared the transferability of PDMS stamps modified with various amphiphilic molecules, including sodium dodecyl sulfate (SDS), Triton, and centrimonium bromide (CTAB), with a bare PDMS stamp and showed that SDS-coated PDMS stamps enhanced the transfer of biomolecules to the substrate.[23] Although the surface modification with SDS was useful to deliver biomolecules and lasted much longer than plasma treatment, it was concerned about the denaturation of ink molecules in case of proteins.[23] Additionally, in our experiences, SDS did increase the inking and transferability, but non-uniform coating of SDS on the stamp surface often leads to the nonuniform ‘spotty’ prints. Our motivation was to develop another way to coat the stamp surface that could enhance both the transferability of ink molecules and uniformity of printed patterns. Here we chose the hydrophilic hydrogel as a candidate material. Among various types of hydrogel, agarose has been used as a stamp material itself in different biological or engineering applications. Agarose stamps were foremost used for multiple patterning without re-inking steps.[24] The ability of agarose to retain ink solution further enabled mammalian cell patterning to porous scaffold directly.[25] In other applications, agarose stamps were used as a reservoir of etchant for fabricating a topographical surface of metal.[26] Conductive polymer[27] or nanoparticles[28] was also capable of being patterned with agarose stamps. The advanced method that used an agarose stamp combined with electrodes was demonstrated the ability of addressable patterning.[29] Despite such successful demonstrations of agarose hydrogel as a stamp material, hydrogel stamps have been still not easy to use in practice since their low stiffness makes themselves difficult to handle. In this work, we propose a new microstamping method for printing biomolecules using a simple bench-top method of agarose coating on PDMS stamps, and quantitatively
2
2
R
2
compared the performance of our method to the conventional mCP with different surface modification schemes. By applying dip-coating method, which has been usually used to fabricate thin film on the surface, we coated agarose hydrogel on a PDMS surface. The agarose thin film with the thickness of a few micrometer enhanced the transferability of ink molecules and the uniformity of printed patterns. In addition, the bioactivity of proteins was still preserved after being patterned, and cells (neurons) grew following the underlying patterns printed with our method.
2. Materials and Methods 2.1. Materials Two types of SU8 photoresist (2002 for 2 mm-thickness and 2050 for 50 mm-thickness; Microchem, MA, USA) and (1H, 1H, 2H, 2H-perfluorooctyl) trichlorosilane (448931, Aldrich, MO, USA) were used for fabricating a silicon mold. A Sylgard 184 kit (Dow Corning, MI, USA) containing prepolymer and curing agent was used for manufacturing PDMS stamps. Acetone, 2-propanol (Junsei Chemical Co., Ltd., Japan), and phosphate buffered saline (PBS, 10010–023, Gibco, CA, USA) were used as received. Glassware, including coverslips or microscope slides, manufactured from Mrienfeld (Germany) was used. Agarose ITM (0710, Amresco, OH, USA) and SDS (L3771, Sigma, MO, USA) were used for coating PDMS stamps. As ink molecules, poly-l-lysine-conjugated FITC (PLL-FITC; P3069, Sigma, MO, USA), synthetic A chain of laminin (C6171, Sigma, MO, USA), natural mouse laminin (23017–015, Invitrogen, CA, USA), and fibronectin from human plasma (F2006, Sigma, MO, USA) were used. Antilaminin (L9393, Sigma, MO, USA), anti-fibronectin (F3648, Sigma, MO, USA), and Alexa Fluor 594 (A11012, Invitrogen, CA, USA) were used for labeling printed proteins. For cell cultures, Hank’s balanced salt solution (HBSS, Welgene, Daegu, Republic of Korea), Neurobasal medium (21103–049, Gibco, CA, USA), B27 (17504–044, Gibco, CA, USA), 2mM GlutaMAX (35050–061, Gibco, CA, USA), lglutamate (G8415, Sigma, MO, USA), and penicillin–streptomycin (15140–122, Gibco, CA, USA) were used.
2.2. Preparation of PDMS Stamps PDMS stamps were fabricated by soft-lithography process described in the previous work.[15] Briefly, a silicon mold patterned with SU8 photoresist was coated with (1H, 1H, 2H, 2H-perfluorooctyl) trichlorosilane to serve as an antistiction layer. The mixture of PDMS prepolymer and curing agent with the ratio of 10:1 was poured on the SU-8 mold and degassed in a vacuum desiccator. After curing in an oven at 60 8C for at least 2 h, cured PDMS was cut into pieces
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
with the size of 1 cm 1 cm. PDMS stamps were cleaned via ultrasonication in acetone, 2-propanol (IPA), and 2nd distilled water for 5 min each.
2.3. Surface Modification Procedures For creating agarose thin film on a PDMS stamp, we used a dip-coating method; a bare PDMS stamp attached on a microscope slide was dipped in hot agarose solution (0.5%) and steadily pulled out. The microscope slide was attached to a push block of a syringe pump (KDS-200, KDScientific, MA, USA) so that it can be pulled out from the solution at constant speed (10 cm h 1). Two different surface modification strategies (surfactant coating or oxidation by plasma) were also used; for surfactant coating, clean PDMS stamps were soaked in 10% SDS solution for 15 min. For the first 5 min, ultrasonication was used to facilitate the coating process. Then stamps were washed with distilled water once to remove any excessive SDS layer and dried with compressed air. For temporary oxidation, clean stamps were treated with air plasma (30W, Cute, Femtoscience, Korea) for 1 min. All stamps were immediately used for inking and stamping.
PDMS surfaces and patterned substrates were taken with the same imaging parameters (1 s exposure, ISO 400). For the quantitative measurement of fluorescence intensities, five field-of-views (FOVs) were randomly selected from each sample. The mean and standard deviation of fluorescence images were calculated using Matlab (Mathworks, Inc., USA). An inverted microscope (IX71, Olympus, Japan) with 10x or 20x objectives and a CCD camera (DP71, Olympus, Japan) were used for taking the fluorescence images of printed patterns.
2.5. Immunoassay To confirm the printing of human fibronectin (0.1 mg ml 1 in PBS) and laminin (0.1 mg ml 1 in PBS), primary antibodies (anti-fibronectin and anti-laminin, Sigma, MO) diluted in PBS (1:500) were treated for 1 h in a 37 8C incubator. After substrates were washed several times with PBS, secondary antibodies diluted in PBS (1:100) were also treated to each sample for 1 h in a 37 8C incubator. Samples were washed again with fresh PBS and fluorescence images of each sample were taken with an epifluorescence microscope (Olympus IX71).
2.4. Thin Film Characterization
2.6. Cell Culture
For the characterization of the surface-modified PDMS, contact angles were measured using a Phoenix 300 goniometer (Surface Electro Optics, Co., Ltd., Republic of Korea) with a drop of distilled water (7 mL). In addition, for characterizing the ability of the PDMS surface as an ink reservoir, we measured contact angles again after loading a droplet of water on each PDMS surface for 15 min. Each contact angle value was manually traced with ImageJ software. For estimating the thickness of agarose thin film on a PDMS stamp, we used z-stack imaging of the confocal microscope (Nikon C2, Japan). To image the agarose film the agarose solution was mixed with PLL-FITC (0.05 mg ml 1) and coated on a flat PDMS stamp. We then acquired the normalized fluorescence profile of the film from the orthogonal view of z-stack images and estimated thickness of the film as the full-width at half maximum of this profile.
Hippocampi were microsurgically separated from E18 Sprague-Dawley (SD) rat (Koatech, Republic of Korea). Dissected hippocampi immersed in HBSS were dissociated with pipetting and centrifuged at 1000 rpm for 2 min. Supernatant was then removed, and settled cells were resuspended in plating medium (Neurobasal medium supplemented with B27, 2 mM GlutaMAX, 12.5 mM lglutamate, and penicillin–streptomycin). After being sieved through a cell strainer (BDFalcon, NJ), cells were plated on a patterned substrate and filled with plating medium in 30 min. Half of the medium was changed twice a week with maintenance medium (plating medium without l-glutamate). All procedures were done according to approved animal use protocols of the KAIST Institutional Animal Care and Use Committee (IACUC).
3. Results
PLL-FITC (0.1 mg ml 1 dissolved in 3rd distilled water) was used as an ink material to quantify the amount of ink present on a PDMS surface and the substrate. The fluorescence intensity of PLL-FITC loaded on the stamp surface was measured after 30 min of inking and 1 min of stamping. The fluorescence images of the same position of
www.MaterialsViews.com
3.1. Fabrication and Characterization of AgaroseCoated Stamps Agarose thin film layer was created on a PDMS stamp surface by dip-coating process. We installed a commercial syringe pump in a vertical direction to implement dipcoating (Supplementary figure 1); a PDMS stamp was attached to a dummy microscope slide and dipped into a
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3 3
3
R
2.5. Fluorescence Measurement
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
fresh agarose hydrogel solution. Once the syringe pump pushed the plunger with a constant speed, it pulled out the stamp with constant speed and resulted in a thin and uniform coating of the agarose solution on the stamp. The thin layer of agarose hydrogel dried immediately after pulling. To confirm the feasibility of the dip-coating method to generate thin film of agarose hydrogel, we compared the contact angle of the agarose-coated (AC) PDMS surface with two different control groups: bare PDMS, water-coated (WC) PDMS (Figure 1Aa). WC-PDMS was prepared by a similar procedure as used for the experimental group except that agarose precursor was not added in the solvent (boiled distilled water). As shown in Figure 1Ab, the contact angle of agarose-coated (AC) PDMS was not significantly different from bare or WC-PDMS. However, the contact angle of ACPDMS apparently altered after loading a drop of water for 15 min, while that of a bare PDMS did not show any significant change (Figure 1Ac). We speculated that the thin agarose film was dried before the first measurement, which could cause the similar wettability of agarose-coated PDMS to bare one. Then for 15 min of droplet loading, the film might absorb water, thereby becoming hydrophilic. Consequently, the significant change of contact angle indicated the formation of thin agarose layer, which could convert PDMS surface to be more hydrophilic overtime. The thickness of agarose thin film formed on PDMS surface was determined by z-stack imaging with a confocal microscope (Figure 1Ba). We acquired the average intensity profile of z-axis form the orthogonal view of 3D image and estimated the thickness as the full width at half maximum of this profile (Figure 1Bb). The thickness of wet agarose thin film on PDMS surface was estimated to be 3.59 0.38 mm (mean SEM; n ¼ 3).
3.2. Agarose-Assisted mCP Process To apply the agarose dip-coating as the surface modification method of PDMS stamps to mCP, we devised a new microstamping procedure (Figure 2A); a cleaned PDMS stamp was coated with the agarose hydrogel by dip-coating and the stamp was subsequently inked with biomolecules for 30 min. The inked stamp surface was cleaned by blowing off the extra inking solution with an air-gun, and the stamp was in contact with the substrate to transfer ink molecules for 10 min. Figure 2B shows three different micropatterns of biomolecules printed by the new procedure: small dot arrays with diameter of 3 or 5 mm and spacing of 5 mm, large dot array with a diameter of 100 mm and spacing of 200 mm, and micro-grid (node diameter: 50 mm; node spacing: 200 mm; line width: 5 mm). Concerned with the distortion of patterns due to the low stiffness of agarose hydrogel, we
4
4
R
4
compared the sizes of patterns (dot arrays with the diameters of 3 mm and 100 mm) fabricated on the PDMS stamp and printed on the surface with our method (Table 1). Although the size of 3 mm- and 100 mm-dot patterns was slightly larger than those of stamps (110.1% and 100.9%, respectively), the statistical comparison was not significantly different. These results indicated that agarosecoated stamps were suitable for printing micro-scale patterns with the wide range of feature sizes with high accuracy.
3.3. Evaluation of Agarose-Assisted mCP Next, we investigated the effect of agarose thin film on the quality of printed patterns. Compared to the bare or modified PDMS stamps with different strategies (SDS coating or plasma treatment), we evaluated the ability of agarose thin film from two aspects; how much ink molecules can be transferred from the stamp to the surface (transferability) and how uniform the printed pattern is (uniformity). All procedures (inking and stamping) except the surface modification step were same. For the first parameter, we assessed the relative ink transferability of each stamp by measuring the ratio of mean fluorescence intensity of a printed substrate to that of a stamp (Figure 3Aa and Ab); the higher the ratio was, the more transferable the ink molecules were. Among the conventional methods, the SDS-assisted method showed the highest transferability, which was consistent with the previous report.[23] Our agarose-assisted method showed comparable transferability (81.3 0.05%) compared with the SDS-assisted method (88.0 0.05%). In addition, our method transferred ink molecules significantly more than other two methods (32.6 0.01% for plasma-treated and 66.4 0.05% for bare PDMS stamps). Therefore, it implied that agarose thin film was effective in achieving high efficiency in transferring ink molecules from the PDMS stamp to the substrate. Figure 3Ac shows the uniformity of stamped ink molecules on the substrates. We used flat PDMS stamps to print fluorescently labeled biomolecules (PLL-FITC) in a large area (1 mm 1 mm), and measured the fluorescence intensity from five different FOVs. Then, coefficient of variation (CV) of the intensity in each FOV was measured, and the reciprocal of this value was considered as the uniformity. As the intensity distribution of uniform patterns was less variable, the uniformity value would be large. Compared to a bare stamp (2.64 0.23), surfacemodified stamps with SDS-coating (3.65 0.25) or plasmatreatment (3.57 0.31) produced more uniform patterns. However, agarose thin film outperformed other methods and produced almost twofold more uniform patterns (7.11 0.51) than other methods. Consequently, our
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
Figure 1. Characterization of agarose-coating on PDMS stamps. (A) Contact angle of PDMS substrate with different coating. (Aa) A bare PDMS surface and three different schemes of surface coating, such as water-coated (‘‘WC’’), and agarose-coated (‘‘AC’’) were used to measure the contact angle of a modified surface. (Ab) The contact angle of all conditions (top images in (Aa)) was not significantly different. (Ac) After 15 min of water loading (bottom images in (Aa)), the contact angle of agarose-coated PDMS (‘‘AC’’) was significantly decreased, while that of bare PDMS was not changed (mean SEM; n ¼ 5 samples; ***p < 0.001, ****p < 0.0001, ns: no significance). (B) Thickness measurement of agarose thin film generated on flat PDMS surface. (Ba) 3D (left) and its orthogonal view (right) of a z-stack image (interval: 1 mm) of fluorescently visualized agarose thin film. (Bb) Average fluorescence profile of the orthogonal image (yellow line in (Ba) indicates xaxis of this profile).
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
5 5
5
R
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
Figure 2. Schematic procedure of agarose-assisted mCP and patterned substrates. (A) A PDMS stamp was dip-coated with agarose solution at first (‘‘Pulling’’), then a drop of biomolecular solution was loaded on the stamp for 30 min (‘‘Loading’’). Subsequently, the stamp was attached on the substrate for 1 min to print patterns (‘‘Contact printing’’). (B) Various designs of micropatterns composed of PLL-FITC were printed by our method: small dot arrays (‘‘3 mm dots’’ and ‘‘5 mm dots’’; spacing of 5 mm), large dot arrays (‘‘100 mm dots’’; spacing of 200 mm), and micro-grid (‘‘grid’’; node diameter: 50 mm; node spacing: 200 mm; line width: 5 mm). Scale bars indicate 5 mm in ‘‘3 mm dots’’/ ‘‘5 mm dots’’ and 100 mm in ‘‘100 mm dots’’/‘‘grid,’’ respectively.
agarose-assisted mCP was capable of not only producing micro-scale patterns that were considerably uniform in large areas, but also enhancing the number of molecules transferred to the substrate. Next, we tested whether repetitive printing was possible with the agarose film coating. Figure 3B shows the micropattern of large dot arrays. The same area of patterns repeatedly printed with the same stamp on the different substrates was shown in Figure 3Ba. The fluorescence intensity of patterns was vivid in the first stamping. In spite of bare visibility, the patterns were also successfully generated on the substrates by second to fifth stamping (80% increment of image brightness could show printed patterns). However, the mean fluorescence intensity of each stamping event indicated the steep decrease of transferred molecules relative to the first stamping (Figure 3Bb); second stamping could transfer only 35% of
Table 1. Comparison of pattern size.
Design
Stamp Printed pattern
3 mm-Diameter dot
100 mm-Diameter dot
2.7 0.0 3.1 0.2a
95.3 0.8 96.2 0.6a
*mean SEM; n ¼ 5. a The values were not significantly different under p < 0.05.
6
6
R
6
molecules compared with the first stamping. Additionally, the amount of transferred molecules gradually decreased as the stamping event increased. These results suggested that our method was capable of repetitive patterning with limitation of inconsistent transferability. To investigate the diffusion of ink molecules outside the pattern area, we measured the size of printed patterns (large dot arrays with diameter of 100 mm) with different stamping time (Figure 3C). The size of micropatterns was not correlated to stamping time, suggesting that the diffusion of ink molecules was negligible.
3.4. Patterning Proteins and Cells To demonstrate the utility of the proposed stamping method, we applied our microstamping method to create micropatterns of proteins or cell-guiding cues. We first confirmed the preservation of protein activity by immunolabeling printed patterns. Figure 4 shows the fluorescence images of two patterned extracellular matrix proteins, human fibronectin and laminin, labeled with sequential reactions of primary antibody to each protein and secondary antibody conjugated with fluorescence markers. Both proteins were labeled with red fluorescence markers, indicating that printed patterns preserved their reactivity. For more accurate corroboration, the same protein patterns printed with the original and modified procedure of agarose-assisted mCP was compared
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
Figure 3. Characterization of micropatterns printed with agarose-assisted mCP. (A) Quantitative comparison of pattern quality (ink transferability and pattern uniformity) using our method with other surface modification strategies. (Aa) Large dot array patterns (diameter: 50–100 mm) printed with the same procedure of mCP using different PDMS stamps: agarose coating (‘‘Agarose’’), SDS coating (‘‘SDS’’), plasma treatment (‘‘plasma’’), and bare (without any surface modification; ‘‘bare’’). Ink transferability (Ab) and pattern uniformity (Ac) of each method was quantitatively analyzed (mean SEM; n ¼ 15; *p < 0.05, ***p < 0.001, ns: no significance). A scale bar in ‘‘Agarose’’ of (Aa) indicates 200 mm. (B) Repetitive patterning with an agarose-coated PDMS stamp. (Ba) Same regions of patterns repeatedly printed on different substrates. All images in ‘‘Original’’ was captured with the same imaging condition. (Bb) Mean fluorescence intensity of pattern areas over the number of stamping events relative to that of patterns in first stamping. (C) Quantitative analysis of pattern sizes depending on the stamping time (mean SEM; not significantly correlated under Spearman correlation test).
www.MaterialsViews.com
fluorescence intensity (Supplementary figure 2A and 2B). Considering the similar transferability of two methods (Supplementary figure 2C), the results indicated that our method could successfully generate active protein patterns. The preservation of protein activity was also confirmed by cell cultivation. We cultured primary hippocampal
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
7 7
7
R
(Supplementary figure 2). In the modified procedure, proteins (laminin) were pre-mixed in boiled agarose solution before dip-coating so that temperature sensitive proteins might lose their reactivity. Compared to the protein patterned in harsh condition (premixed), one patterned with our method showed significantly higher
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
M. J. Jang and Y. Nam www.mbs-journal.de
Figure 4. Patterning active proteins and neurons. Extracellular matrix proteins (human fibronectin and laminin) were printed with agaroseassisted mCP and subsequently labeled with fluorescence-conjugated antibodies (‘‘Protein’’). Cultured hippocampal neurons (‘Tuj1’; fixed at 4 DIV) were adhered and grew on theses patterned substrates, but only laminin patterns were useful in guiding neurons (‘‘Merged’’). The scale bar indicates 200 mm.
neurons from E18 rat on these protein-patterned substrates (middle of Figure 4). On the micro-gird patterns composed of laminin, neurons preferentially adhered on nodes and extended their neurites following thin lines at 4 days in vitro (DIV). On the other patterned substrate with the same designs but different proteins (human fibronectin), on the other hand, neurons could not recognize the underlying patterns (right of Figure 4). The results suggested the surface-printed protein patterns preserved their activity sufficient to lead neuronal guidance and selective recognition to microenvironment.
4. Discussion In this work, we proposed a new surface modification of PDMS stamps with agarose hydrogel for advanced mCP. Our dip-coating method using a commercial syringe pump was simple and straightforward to execute. Adding the coating step to conventional mCP, we also demonstrated the ability of agarose thin film to improve the patterning quality (transferability and uniformity) and to implement active protein patterns especially for designs of the neuronal microenvironment. For past decades, several strategies for the surface modification of PDMS stamp have been introduced to improve the quality of biomolecular patterns. However, it was difficult to satisfy two essential requirements: patterning efficiency and pattern uniformity. The plasma
8
8
R
8
treatment of PDMS stamps for temporal oxidization has been the standard technique for converting hydrophobic stamp surface into hydrophilic surface for loading and transferring biomolecular inks.[21] According to our result this method produced more uniform patterns, but the amount of ink molecules delivered to the substrate was less than that from a bare stamp. The oxidized PDMS surface might be too strong to release ink molecules due to stronger wettability than glass substrates.[20,31] Chang et al. suggested another strategy with SDS-coating on PDMS stamps, which had advantages over efficient transfer of ink molecules.[23] We also obtained a similar result using the SDS-coated, while it was observed that the printed patterns had sporadic hot spots where fluorescence intensity was exceptionally higher than surrounding areas owing to the aggregation of ink molecules, which might result in less uniformity. Compared with the conventional strategies, our agarose-coating method achieved both high transferability and uniformity at the same time. Our agarosecoated stamps were quite different from the agarose stamp that is made of agarose hydrogel. Mayer et al. suggested the several advantages of agarose hydrogel stamps for biomolecular printing, especially the possibility of consecutive printing without re-inking.[24] As our method could not repeatedly produce the patterns with the same quality, our agarose thin film might be too thin to serve as a reservoir for the ink solution. The uniformity of surface patterns is important to generate reliable conditions in a variety of experimental
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
Agarose-Assisted Micro-Contact Printing www.mbs-journal.de
5. Conclusion Here we proposed the agarose-assisted mCP method for patterning biomolecules. A thin agarose layer was generated on hydrophobic PDMS surface by means of the dipcoating method; it promoted the transfer of ink molecules to the substrate and rendered the printed pattern considerably uniform in a large area. As a consequence, the simple modification of a PDMS stamp with agarose hydrogel improved the quality of printed biomolecular patterns, thereby contributing to execute reliable experimentations and reduce the opportunity costs in biochip applications.
Acknowledgement: This research was supported by the National Research Foundation of Korea (NRF-2011–0019213, NRF2012R1A2A1A01007327), funded by the Ministry of Science, ICT & Future Planning (MSIP).
Received: September 12, 2014; Revised: November 14, 2014; Published online:DOI: 10.1002/mabi.201400407 Keywords: agarose hydrogel; cell chip; DNA chip; micropatterning; protein chip; soft-lithography
[1] A. Kumar, G. M. Whitesides, Appl. Phys. Lett. 1993, 63. [2] P. Xiao, J. Gu, J. Chen, D. Han, J. Zhang, H. Cao, R. Xing, Y. Han, W. Wang, T. Chen, Chem. Commun. 2013, 49, 11167. [3] G. Liu, W. Guo, D. Song, Biosens Bioelectron. 2014, 52, 360. [4] H.-S. Shin, H.-J. Yun, K.-H. Baek, T.-H. Ham, K.-S. Park, D.-P. Kim, G.-W. Lee, H.-D. Lee, K. Lee, L. M. Do, J. Nanosci, Nanotech 2012, 12, 5325.
www.MaterialsViews.com
[5] Y.-M. Chien, F. Lefevre, I. Shih, R. Izquierdo, Nanotechnology 2010, 21, 134020. [6] M. Ghezzi, S. C. Thickett, C. Neto, Langmuir 2012, 28, 10147. [7] K. Bhatnagar, A. Pathak, D. Menke, P. V. Cornish, K. Gangopadhyay, V. Korampally, S. Gangopadhyay, Nanotechnology 2012, 23, 495201. [8] H.-J. Yun, Y.-H. Ham, H.-S. Shin, K.-S. Jeong, J.-G. Park, D.-S. Choi, G.W. Lee, J. Nanosci. Nanotech. 2011, 11, 5640. [9] N. Gadhari, M. Charnley, M. Marelli, J. Brugger, M. Chiquet, Biochim. Biophys. Acta 2013, 1833, 3415. [10] P. Bajaj, B. Reddy, L. Millet, C. Wei, P. Zorlutuna, G. Bao, R. Bashir, Integr. Biol. 2011, 3, 897. [11] M. C. Leong, S. R. K. Vedula, C. T. Lim, B. Ladoux, Commun. Integr. Biol. 2013, 6, e23197. [12] E. T. Baran, K. Tuzlakolu, A. Salgado, R. L. Reis, J. Tissue Eng. Regen. Med. 2011, 5, e108. [13] M. J. Jang, Y. Nam, J. Neural Eng. 2012, 9, 046019. [14] W. R. Kim, M. J. Jang, S. Joo, W. Sun, Y. Nam, Lab Chip 2014, 14, 799. [15] M. E. Boggs, W. R. Thompson, M. C. Farach-Carson, R. L. Duncan, T. P. Beebe Jr., Biointerphases 2011, 6, 200. [16] D. Liazoghli, A. D. Roth, P. Thostrup, D. R. Colman, ACS Chem Neurosci. 2012, 3, 90. [17] X. Tang, M. Y. Ali, M. T. A. Saif, Soft Matter. 2012, 8, 7197. [18] A. Bernard, J. P. Renault, B. Michel, H. R. Bosshard, E. Delamarche, Adv. Mater. 2000, 12, 1067. [19] S. A. Lange, V. Benes, D. P. Kern, J. K. H. Hºrber, A. Bernard, Anal. Chem. 2004, 76, 1641. [20] T. Kaufmann, B. J. Ravoo, Polym. Chem. 2010, 1, 371. [21] C. D. James, R. C. Davis, L. Kam, H. G. Craighead, J. N. Turner, W. Shain, Langmuir 1998, 14, 741. [22] J. L. Tan, J. Tien, C. S. Chen, Langmuir 2002, 18, 519. [23] J. C. Chang, G. J. Brewer, B. C. Wheeler, Biomaterials 2003, 24, 2863. [24] M. Mayer, J. Yang, I. Gitlin, D. H. Gracias, G. M. Whitesides, Proteomics 2004, 4, 2366. [25] M. M. Stevens, M. Mayer, D. G. Anderson, D. B. Weibel, G. M. Whitesides, R. Langer, Biomaterials 2005, 26, 7636. [26] S. K. Smoukov, K. J. M. Bishop, R. Klajn, C. J. Campbell, B. A. Grzybowski, Adv. Mater. 2005, 17, 1361. [27] T. S. Hansen, K. West, O. Hassager, N. B. Larsen, Adv. Mater. 2007, 19, 3261. [28] Y.-C. Kao, F. C.-N. Hong, Nanotechnology 2011, 22, 185303. [29] S. Sekine, S. Nakanishi, T. Miyake, K. Nagamine, H. Kaji, M. Nishizawa, Langmuir 2010, 26, 11526. [30] M. J. Jang, Y. Nam, BioChip J 2012, 6, 107. [31] G. Csucs, T. Kuºnzler, K. Feldman, F. Robin, N.?D. Spencer, Langmuir 2003, 19, 6104. [32] S. K. W. Dertinger, X. Jiang, Z. Li, V. N. Murthy, G. M. Whitesides, Proc. Natl. Acad. Sci. USA 2002, 99, 12542. [33] A. C. Philipsborn von, S. Lang, J. Loeschinger, A. Bernard, C. David, D. Lehnert, F. Bonhoeffer, M. Bastmeyer, Development 2006, 133, 2487. [34] R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, T. A. Witten, Nature 1997, 389, 827. [35] J. M. Moran-Mirabal, H. G. Craighead, C. P. Tan, R. N. Orth, E. O. Williams, D. M. Lin, Anal. Chem. 2007, 79, 1109.
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201400407 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
9 9
9
R
designs. For example, the different distribution of molecules within or across the patterns could influence cellular behavior in different ways as cells could recognize the concentration gradient of surface-bound molecules.[32,33] In case of microarrays, the fabrication of uniform surface spots has been technically challenging as the printed spots usually formed ‘‘coffee rings’’ due to the capillary flow;[34] as the amount of target molecules bound to the complementary sequences printed on the surface could be varied, the microarray data could be noisy.[35] In such circumstances, agarose-assisted mCP technique could reduce experimental errors or batch effects. The improved transferability of a stamp would be beneficial to reduce the quantity of biomolecules required for creating patterns since most of the biomolecules including proteins or antibodies were expensive.
Early View Publication; these are NOT the final page numbers, use DOI for citation !!
