SEARCH

SEARCH BY CITATION

Keywords:

  • microfluidic chip;
  • biomimetic hydrogel;
  • cellular behavior;
  • glioma cell;
  • hyaluronic acid

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

A biomimetic hydrogel was integrated into microfluidic chips to monitor glioma cell alignment and migration. The extracellular matrix-based biomimetic hydrogel was remodeled by matrix metalloprotease (MMP) secreted by glioma cells and the hydrogel could thus be used to assess cellular behavior. Both static and dynamic cell growth conditions (flow rate of 0.1 mL/h) were used. Cell culture medium with and without vascular endothelial growth factor (VEGF), insensitive VEGF and tissue inhibitor of metalloproteinases (TIMP) were employed to monitor cell behavior. A concentration gradient formed in the hydrogel resulted in differences in cell behavior. Glioma cell viability in the microchannel was 75–85%. Cells in the VEGF-loaded microchannels spread extensively, degrading the MMP-sensitive hydrogel, and achieved cell sizes almost fivefold larger than seen in the control medium. Our integrated system can be used as a model for the study of cellular behavior in a controlled microenvironment generated by fluidic conditions in a biomimetic matrix. © 2013 Wiley Periodicals, Inc. J Biomed Mater Res Part A: 102A: 1164–1172, 2014.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Cells in tissues are constantly exposed to microenvironments enriched in a fibrous three-dimensional (3D) extracellular matrix (ECM) and factors promoting cell proliferation, migration, and differentiation.[1, 2] Local fluid flow carrying oxygen and nutrients also influence cell activities.[3-5] Basic cell biology researchers have focused on both of these effects and have studied them extensively with conventional methods.[6-8] However, elucidation of the precise control of microenvironments has been extremely challenging to cell biologists.[9, 10] Recent microfluidic technology facilitated the control of small volumes of liquids and cells and has been broadly applied in basic cell biology, permitting insights into cell–cell interactions, cell differentiation, and cell proliferation.[11, 12] However, most of the microfluidic system was 2D and, therefore, posed limitations on studying cellular behavior in vivo in a mock environment. Recently, some researchers reported combinations of 3D matrices with microfluidic devices in cell biology.[13-16] Toh et al. developed a 3D microenvironment that incorporated both a configurable 3D matrix and fluid perfusion (in a microfluidic channel) for cell culture[17] and Lii et al. constructed a real time microfluidics system to study the growth of and communication between cells in a 3D extracellular matrix.[18] Beebe and colleagues fabricated active hydrogel components inside microfluidic channels using photopatterning,[19] and Cheng et al. used a hydrogel-based microfluidic device to study cell migration.[20] Choi et al. reported fabrication of a cellular microfluidic scaffold based on a calcium alginate hydrogel.[21] Although these 3D microfluidic chips provide more in vivo like conditions, hydrogels used in these chips have limitations in mimicking the real extracellular environment.[22] Hydrogels used in these experiments are based on biomolecules such as alginate, collagen or synthetic materials such as poly(ethylene) glycol. The drawback of using these materials is that customization of the materials is very limited, thus making it difficult to mimic the in vivo situation. For example, 3D conditions creating vasculature structures require growth factors as well as peptides guiding endothelial cell adhesion and proliferation. Most of the materials in the previous study only support the mechanical 3D structures. In addition, the design of a chip enabling the controlled diffusion of molecules through the embedded hydrogel was not fully proven in the previous studies. These two factors are very important in studying the cellular behavior in ECM.

In this article, we fabricated a biomimetic[23-25] 3D culture system combining a hydrogel matrix with a microfluidic system which can provide growth factors in a controllable way. MMP-sensitive hyaluronic acid (HA) hydrogel was used as the backbone hydrogel matrix. MMP-sensitive HA hydrogels were developed and used to study cell differentiation and tissue regeneration.[26, 27] The advantage of using this hydrogel system is that more in vivo like microenvironments can be created by simply adding growth factors as well as peptide fragments. For example, chips integrating this biomimetic hydrogel can create the predetermined neural cell differentiation microenvironments by incorporating brain derived growth factors with axonal adhesion peptides such as IKVAV. An MMP-sensitive peptide was used as a crosslinker. This allowed the biomimetic scaffolds to be degraded by the MMPs secreted from the cells in the chips. The activity of MMPs secreted in the tumor cells can modulate the remodeling rates of this biomimetic hydrogel by cleaving MMP sensitive crosslinkers.[28] In this chip, we cultured the glioma cell, one of the human brain cancer cell lines, by changing the fluid flow through the polyurethane nanofiber membrane that was prepared by electrospinning. The role of the nanofiber membrane was not only to support the wall to prevent the collapse of the hydrogel into the microfluidic channel but also to act as a selective porous membrane to allow the diffusion of media and growth factors into the hydrogel.

Diffusion of medium through a hydrogel was investigated by the fluorescence recovery after photobleaching (FRAP) technique. We changed the composition of fluids by adding growth factors such as VEGF and TIMP. Cells in the VEGF samples spread extensively compared to those in the control samples, and the orientation of the cell spreading correlated with the fluid conditions. Integration of biomimetic hydrogels into the microfluidic chips is one of the fascinating techniques for creating biomimetic in vivo microenvironments for studying tissue perfusion as well as cell behavior in the 3D matrix.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Materials

Polyurethane (PU, Pellethane 2363-80AE) and Polydimethylsiloxane (PDMS, Sylgard 184 Silicone kit) were purchased from Dow Corning (Midland, MI). Hyaluronic acid (HA) (MW 50,000 Da) was purchased from Lifecore Biomedical Co. (Chaska, MN). 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), adipic acid dihydrazide (ADH), formamide, and triethanolamine (TEA) were purchased from Sigma-Aldrich, Inc. (St. Louis, MO). 1-hydroxybenzotriazole hydrate (HOBT) was the product of Fluka Chemical Co. (Buchs, Switzerland). N-acryloxysuccinimide (NAS) was purchased from Acros Organics (Morris Plains, NJ). MMP-sensitive and RGD peptides were purchased from Anygen (Gwang-ju, Korea). Dulbecco's Modified Eagle's Medium (DMEM), fetal bovine serum (FBS), penicillin/streptomycin (10,000 U/mL), and trypsin/EDTA solution were purchased from GIBCO BRL (Carlsbad, CA). Alexa Fluor 594 phalloidin, Image-iT™ FX signal enhancer and Live/Dead assay kit were purchased from Invitrogen (Eugene, OR). Recombinant human VEGF and insensitive VEFG and TIMP were purchased from PeproTech (Rocky Hill, NJ). And FITC-dextran was purchased from Sigma (70 kDa, St. Louis, MO).

Preparation of electrospun nanofiber membrane

To form a biomimetic hydrogel integrated into a microfluidic channel, we used an electrospun nanofiber membrane that provided not only hydrogel support within the microfluidic chip but also a fluid diffusion path. We used PU, which has suitable properties for electrospinning.[29] The electrospinning protocol has been previously published.[30, 31] In brief, a 12% (w/w) PU solution and the Chungpa EMT electrospinning apparatus (Seoul, Korea) were employed. For easy detachment of the PU nanofiber membrane from the collecting drum, we used oxygen plasma to make the cover glass hydrophilic and affixed the treated cover glass to the collecting drum. The electrospun PU nanofiber membrane was dried and residual solvent removed in a vacuum oven at 65°C over 3 days.

Fabrication of microfluidic chip

We fabricated a Polydimethylsiloxane (PDMS)-based microfluidic platform using a conventional soft lithographic process,[32] We prepared two PDMS layers on which patterns of upper and lower channels were engraved. The previously reported optimized method has four steps.[14] In brief, (i) inlet and outlet port holes were created in the lower layer using a polished 18 gauge needle, (ii) PDMS layer surfaces were exposed to oxygen plasma for 15 s to improve adhesion and bond, (iii) the nanofiber membrane on the grooved edge, to which a half-cured PDMS solution was glued for bonding, was cured and used to cover the upper layer, and (iv) tubing lines were connected and the assembled chip placed on a hotplate at 80°C for 4 h for thermal curing. Figure 1(a) illustrates the components of chips and Figure 1(b) shows the synthesis of the biomimetic hydrogel structure.

image

Figure 1. (a) Schematic of the microfluidic chip containing a glioma cell-loaded biomimetic hydrogel and an electrospun nanofiber membrane, (b) Synthesis of biomimetic hydrogel structure, and (c) SEM image of electrospun PU nanofiber membrane, (d) Micrograph of the chip fabricated microfluidic chip. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Synthesis of biomimetic hydrogel with acrylation of hyaluronic acid

We previously described the acrylation of hyaluronic acid. HA (0.1 g, 0.25 mmol) was dissolved in 10 mL of triple-distilled water and EDC (0.24 g, 1.25 mmol) was added, as were HOBT (0.17 g, 1.25 mmol) and ADH (1.1 g, 6.25 mmol).[27] The mixture was incubated for 2 h at a pH of 4.8. The same amounts of EDC and HOBT were again added and incubation continued for 2 h at a pH of 7.0. The HA-ADH solution was dialyzed against 100 mM NaCl for 2 days and distilled water for 1 day using a dialysis membrane (MWCO 14,000; SpectraPor, Rancho Dominguez, CA). 20 mL of NAS (0.25 g, 1.5 mmol) in formamide was added to the HA-ADH solution. The reaction continued for 12 h at 37°C. The HA-ADH-NAS solution was dialyzed against 100 mM NaCl for 2 days and distilled water for 1 day using a dialysis membrane. Finally, the reaction mixture was lyophilized for 4 days to obtain an acrylated HA (HA-AC) powder.

Cell preparation

We used the glioma cell line A-172 (American Type Culture Collection [ATCC], Rockville, MD); cells were maintained in DMEM supplemented with 10% (v/v) FBS and 1% (v/v) of the commercial penicillin/streptomycin mixture in a humidified atmosphere containing 5% (v/v) CO2 at 37°C. Cells were harvested by trypsin/EDTA treatment.

Gel preparation and integration into the microchip

For gel preparation, acrylated HA was dissolved in TEA-buffered solution (0.3 M, pH 8.0). As a crosslinker, MMP-sensitive peptides (GCRDGPQGIWGQDRCG) and RGD peptides (GCGYGRGDSPG) were added with the same molar ratios of acryl and thiol groups. The HA-based hydrogel was formed via a Michael-type addition reaction.[33] This hydrogel (10%, w/w) was mixed with an A-172 cell suspension (5 × 104 cells) to yield a final volume of 50 μL. Finally, the hydrogels were incubated for 30 min at 37°C for gelation within a glass reservoir (inner diameter 3 mm and length 35 mm); this aids in stable gelation and facilitates microfluidic chip embedding.

Diffusion in the biomimetic hydrogel

Diffusion is a dominant factor guiding cell behavior.[34, 35] In order to measure the diffusion coefficient of our biomimetic hydrogels, the FRAP technique was performed. The hydrogel samples were prepared using three different concentration ratios: 5, 7, and 10% (w/w) to optimize the suitable concentration of hydrogel. The samples were loaded with FITC-dextran (70 kDa, 50 mg/mL) for 1 day. The fluorescent samples were photobleached by an Ar-ion laser (35LAP43, CVI-Melles Griot, Carlsbad, CA) at 488 nm with 23 mW. The exposed time was controlled at 257 ms using a mechanical shutter (Uniblitz LS3, NY). After the shutter was closed, the images of the fluorescence recovery for three different samples were captured using an ICCD camera (Dicam-Pro, Cooke Corp., MI) and AQM6 S/W (Kinetic Imaging, Nottingham, UK). The fluorescence recovery was measured by an inverted fluorescence microscope (Olympus, IX71, Tokyo, Japan) and monitored in real-time. Plotting the recovery curve allowed us to calculate diffusion coefficients by using Equation (1).[36] From our observation of TEM00 Gaussian beam profile, we assumed that the photobleaching beam was uniformly circular. Then, the diffusion coefficient was expressed as follows:

  • display math(1)

where inline image is the half-recovery time, w is the measured initial spot radius, and D is the diffusion coefficient.

To observe the diffusion in biomimeteic hydrogel integrated microchip, we introduced the blue-dye solution into the chip using a syringe pump. We recorded the diffusion during 36 h using a digital camcorder (Panasonic MX-500, Japan) with 0.1 mL/h of flow rate. Similarly, to investigate the diffusion profile biomimetic hydrogel, the BSA-FITC was introduced into microchip with same method.

Mimicking a cellular microenvironment in a microfluidic chip

Use of a 3D nanoporous membrane and MMP-sensitive peptides in hydrogel-integrated microfluidic chips may offer several benefits. The microfluidic channels provide effective transport and exchange of cell culture medium for hydrogel remodeling, which guides the growth and migration of the glioma cell. Biomimetic hydrogels loaded with glioma cells were placed in the upper area of the microchannels. As shown in Figure 1(a), a nanofiber membrane was placed in the hole of the channel to separate the hydrogel from the microfluidic channel. We varied flow condition (static and dynamic [0.1 mL/h]), and medium composition (DMEM, DMEM with 10% [v/v] FBS, DMEM with 10% [v/v] FBS, TIMP [100 ng/mL], insensitive VEGF [100 ng/mL], and VEGF [100 ng/mL]). We cultured cells in the microchip in a CO2 incubator for 7 days. Precisely, to analyze the modulated cell's morphology, a thin window frame-shaped PDMS holder (thickness 3 mm) was prepared, by which we could distinguish hydrogel sections (top, middle, and bottom). Cells in each section were investigated by actin filament staining and the Live/Dead assay. Cells were stained with Alexa Fluor 568 phalloidin fluorescent dye for morphological analysis. To improve the signal-to-noise ratio of fluorescently labeled cells, Image-iT™ FX signal enhancer was applied and background noise was blocked by preventing nonspecific staining. Fluorescent images were taken and processed using an Axiovert 200 (Carl Zeiss, Darmstadt, Germany) inverted microscope. To measure viability, cells were treated with Live/Dead assay reagents and incubated for 10 min at room temperature. Viability was evaluated using a fluorescence microscope.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Biomimetic hydrogel microfluidic chip

PDMS-based microfluidic chips containing a biomimetic hydrogel were fabricated and glioma cells were successfully loaded as in Figure 1. As shown in Figure 1(c), extracellular matrix-like electrospun PU nanofiber membrane was used as the interface between hydrogel and microchannels. To prevent the collapse of HA hydrogel into microfluidic channel, especially, we bonded the PU nanofiber membrane onto the lower layer. That could successfully support the HA hydrogel in the upper layer and perform a buffer function of allowed inflow fluids for 7 days. Beebe and colleagues have also integrated hydrogels into microfluidic chips and explored cellular behavior in such systems.[37] The unique feature of the biomimetic hydrogel system is that it allows one to create unique microenvironments by simply changing the crosslinker, cell adhesion signal peptides, and growth factors. We used MMP-sensitive peptides as crosslinkers and RGD peptides as cell adhesion signals. By combining a biomimetic hydrogel with a microfluidic system, we could monitor cellular matrix invasion under five different microfluidic flow conditions.

Diffusion coefficients of Hydrogel

To provide an adequate supply of media or growth factors to cell via diffusion, the diffusion coefficients must be necessary. Diffusion coefficients for the different concentration of HA-hydrogel were quantified using a FRAP technique and plotted in Figure 2(a). Different diffusion coefficients at the different HA concentration implies that HA definitely influences diffusion in cells. The literature survey shows that the diffusion coefficients ranging from 1.1–1.4 × 10−6 cm2/s for our HA-hydrogel were similar to those for the following materials. The diffusion coefficient of fluorescein-G-actin having 43 kDa of molecular weight in vitro buffer was measured to be 0.920 × 10−6 cm2/s. In water solution, Ribonuclease A protein with a molecular weight of 14 kDa has a diffusion coefficient of 1.360 × 10−6 cm2/s.[38] The diffusion coefficient for Adenosine 5′-triphosphate (ATP) in cytoplasm was reported to be 1.500 × 10−6 cm2/s.[39] The diffusion coefficient for Cyanmetmyoglobin (MbCN) was measured to be 1.018 × 10−6 cm2/s.[40] From the point of view of molecular diffusion in cells, nutrients in our HA samples are expected to diffuse in a similar way to the above materials.

image

Figure 2. (a) Diffusion coefficient calculation by FRAP technique with 70 kDa FITC-dextran in HA-hydrogel; the loaded fluorescence was fully recovered after photobleaching, (b) Blue dye fluids diffusion through biomimetic smart hydrogel integrated microfluidic chip, and (c) diffusion profliles of an FITC-BSA was introduced in biomimetic smart hydrogel for 36 h. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Monitoring diffusion and flows in the hydrogels assembled in the microchannels are important features to apply our chip-based biomimetic 3D cell culture system, it was necessary to control of diffusion for controllable microenvironment. As shown in Figure 2(a), we introduced diluted blue dye-solution into the fabricated microchip without cells. By tracing the dye solution through the channel, the diffusion of the dye into the integrated hydrogel worked well without leakage or disintegration of the system. We could estimate the diffusion rates by measuring the time and diffused area in the hydrogel. The calculated diffusivity was about 0.22 mm/h at 10 wt % HA-based hydrogel when the flow rate was set to 0.1 mL/h. Figure 2(b) shows that the diffusion of the dye was not saturated even after 36 h when comparing the dye intensity by sectioning top, middle, and bottom areas. Therefore cells in the top areas of the hydrogel could not be provided the sufficient nutrients until 36 h.

Measuring cell compatibility of fluidic chips

To mimic a cellular microenvironment, we immobilized cancer glioma cells into HA hydrogel then embedded in microfluidic chip with various flow condition. During the 7 days, the cells were continuously exposed to the different microenvironment culture condition. To compare the responsibility of the cell to mimicked microenvironment, cell viability was evaluated. Figure 3 shows cell survival rates under static conditions and four different dynamic flow conditions. Overall, the 7-day survival rate was 80 ± 5%; no significant differences were seen even when media fluidic conditions (static and 4-dynamic flow condition [pure media, with TIMP, with insensitive VEGF and with VEGF]) were altered. In a previous study, cell viability in hydrogel-containing tissue culture wells was over 90%. However, our system may nonetheless be used to monitor cellular events because cell viability remains relatively high, and we offer a biomimetic 3D matrix with controllable flows.

image

Figure 3. Survival rates of glioma cells in different flow conditions (one static condition and four dynamic flow conditions [no-VEGF medium, TIMP-containing medium, VEGF-containing medium and insensitive VEGF-containing medium, each flowing at 0.1 mL/h, n = 5, * p < 0.05, and ** p < 0.01]). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Remodeling of HA-hydrogel

A unique feature of our microfluidic system is that the hydrogel can be customized by varying the hydrogel composition. In the present work, we used RGD peptides to mediate cell adhesion and MMP-sensitive peptides to monitor matrix remodeling resulting from cell migration and alignment. We investigated morphological changes in glioma cells using actin staining. As shown in Figure 4(a), cells grown under static conditions mostly remained round and showed no remodeling of the hydrogel matrix. Cells grown under dynamic conditions, in contrast, developed elongated shapes. Cells began to change shape when supplied with nutrients by flow. Cells receiving VEGF-loaded flow were fivefold larger than the normal medium controls. Cells grown under static conditions remained round regardless of their hydrogel position whereas those grown under dynamic conditions showed elongated shapes. In particular, cells exposed to VEGF-containing medium showed longer extensions than did control cells. Notably, cellular orientation varied with cell position within the hydrogel [Fig. 4(b)]. Cell alignments caused by different flow rates showed various features. Fibroblast cells exposed to a slow flow, such as that of interstitial fluids, aligned perpendicularly to the flow direction.[41] Cells exposed to higher flow rates (such as 0.1 mL/h) aligned with the flow direction. In our case, glioma cells indirectly exposed to 0.1 mL/h flow aligned with the flow direction. Therefore, matrix diffusion driven by flow through the microchannel also affected cell alignment. Our system mimics cancer tissues interfaced with microvessels showing low flow rates. As shown in Figure 4(c), we have divided the hydrogel with three sections (top, middle, and bottom) and observed the cellular morphology and orientation. Cells in static condition remain round regardless of the position in the hydrogel. Cells in the dynamic condition showed the elongated shape. Especially, cells exposed to VEGF containing medium extend larger compared to normal medium. Interestingly, cellular orientation showed the tendency depending on the position of cells in the hydrogel. Cells in top area of hydrogel align and possible migrate toward to the bottom area and showing more than 80° compared to the flow direction in all the dynamic conditions. Cells in bottom area showed the 40° degree. This result reflects that cells on the top start migrate to the bottom area with higher concentration of nutrient. However, cells in the bottom area of hydrogel enriched with nutrients aligned toward the flow direction. This result shows that cellular remodeling was affected by the flows in the bottom area of hydrogels and cells start aligning and migrating according to the direction of flows. The alignment and possible migration of glioma cells toward flow in the biomimetic matrix may explain cell behavior in cancer tissues. Our basic study can be applied to an understanding of cancer cell biology, especially cell migration and metastasis in cancer tissues.

image

Figure 4. (a) Comparison of the actin filament-based cytoskeleton of different hydrogel regions (top, middle, and bottom) using different media and flow conditions, (b) different-sized cellular distribution was analyzed using image processing software using different flow conditions (n= 10, * p < 0.05), (c) cellular orientation was analyzed image processing software (LSM, Image Examiner) with divided the hydrogel with three section (top, middle and bottom), different fluids composition and flow condition (n = 10, * p < 0.05). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES

Remodeling of extracellular matrix by surrounding cells is an important issue in biological as well as biomedical sciences. Metastasis, cell migration through the extracellular matrix, is one of the unique features of tumor cells and requires fundamental studies using more defined technologies. Cell adhesion and migration in the biomimetic scaffolds are monitored by changing the fluidic conditions using microfluidic chips. Cell in the MMP-sensitive hydrogels can degrade the hydrogels by secreting MMPs and the degradation rate of the hydrogel can be modulated by the activity of MMPs secreted from the cells.[27, 42] The driving force of cell migration is achieved by adding growth factors, such as VEGF, in the microfluidic chips. Cells in the dynamic conditions changed the cell orientations by remodeling biomimetic hydrogel degraded by MMPs. This well-defined system can be a powerful tool for analyzing major determinants of cell spreading and migration in the 3D biomimetic conditions. Moreover, 3D biomimetic conditions can be easily modulated by incorporating growth factors and peptides immobilized in the hydrogels. For instance, YIGSR peptide can be immobilized for facilitating vascular formation with VEGF. The combination of the biomimetic 3D system with the microfluidic system can open a new window for controlling diffusion and flow rates as well as biomimetic microenvironments for creating tissue like environments for studying cell migration, spreading, and differentiation.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS AND DISCUSSION
  6. CONCLUSION
  7. REFERENCES