SEARCH

SEARCH BY CITATION

Keywords:

  • RNA aptamers;
  • circulating tumor cells;
  • human glioblastoma;
  • polydimethylsiloxane;
  • lab-on-chip;
  • nanotextured materials;
  • microscopy;
  • basement membrane

Abstract

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

BACKGROUND:

The detection of a small number of circulating tumor cells (CTCs) is important, especially in the early stages of cancer. Small numbers of CTCs are hard to detect, because very few approaches are sensitive enough to differentiate these from the pool of other cells. Improving the affinity of a selective, surface-functionalized molecule is important given the scarcity of CTCs in vivo. There are several proteins and aptamers that provide such high affinity; however, using surface nanotexturing increases this affinity even further.

METHODS:

The authors report an approach to improve the affinity of tumor cell capture by using novel aptamers against cell membrane overexpressed epidermal growth factor receptors (EGFRs) on a nanotextured polydimethylsiloxane (PDMS) substrate. Surface-immobilized aptamers were used to specifically capture tumor cells from physiologic samples.

RESULTS:

The nanotexturing of PDMS increased surface roughness at the nanoscale. This increased the effective surface area and resulted in a significantly higher degree of surface functionalization. The phenomenon resulted in increased density of immobilized EGFR-specific RNA aptamer molecules and provided significantly higher efficiency to capture cancer cells from a mixture. The data indicated that CTCs could be captured and enriched, leading to higher yield yet higher background.

CONCLUSIONS:

A comparison between glass slides, plain PDMS, and nanotextured PDMS functionalized with aptamers demonstrated that a 2-fold approach of using aptamers on nanotextured PDMS can be important for cancer cytology devices, and especially for the idea of a “lab-on-chip,” toward higher yield in capture efficiency. Cancer 2012;. © 2011 American Cancer Society.


INTRODUCTION

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

Cancer mortality can be reduced significantly by developing methods for early detection and prevention.1, 2 Several strategies for the detection and isolation of tumor cells has been reported.3-9 Detection and sorting based on affinity interactions, especially with aptamers, can yield higher efficiency and greater specificity.10 It has been demonstrated that aptamers have better affinity and greater specificity than antibodies.11 Anti-epidermal growth factor receptor (anti-EGFR) RNA aptamer substrates can specifically recognize, capture, and isolate human glioblastoma (hGBM) cells, which are known to over-express EGFR, from a mixture of fibroblasts.12 The “mean capture yield” can be increased by forcing the sample to run over the substrates multiple times. However, it takes more time and also may decrease specificity.

It is well known that the basement membrane can anchor a cancer cell to its loose, underlying connective tissue through cell adhesion molecules known as integrins.13 The natural nanostructured characteristics of the basement membrane can improve cell adhesion and growth.14, 15 For normal proliferation, cancer cells first have to attach appropriately to the matrix.14 In tissue engineering, researchers have tried to mimic the nanoscale topography of native tissues to increase cell proliferation on scaffolds.16, 17 The results indicate that nanostructured scaffolds can specifically and significantly improve the densities of certain cells.17

Polydimethylsiloxane (PDMS) is one of the most actively used polymers in biomedical devices research. It is easy to manipulate, and its stable chemical and physical properties make it vitally important in cell experiments.18

In this study, 3-dimensional, nanotextured PDMS substrates were prepared and then functionalized with anti-EGFR aptamers for cell isolation. Both of these steps increased the affinity of the surface. We demonstrate that the nanoscale topography of PDMS provides an additional factor to increase affinity for cancer cell attachment, because it provides a larger surface area for aptamer immobilization and increases the number of available aptamers on the surface for cell capture. The data presented here provide solid proof that nanotextured substrates can significantly improve cancer cell isolation and sensitivity without a significant decrease in specificity. Three-dimensional nanotexturing provides much better probability of capturing and isolating a small number of tumor cells from solution. This can be helpful in developing novel cytologic tools for circulating tumor cell (CTC) detection.

MATERIALS AND METHODS

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

All chemicals were obtained from Sigma-Aldrich (St. Louis, MO) unless noted otherwise.

Aptamer Preparation

Purified human EGFR (R&D Systems, Minneapolis, MN) was used for anti-EGFR RNA aptamer preparation by selecting binding species.19, 20 The EGFR protein was purified from murine myeloma cells and contained the extracellular domain of human EGFR (leucine 25 to serine 645) fused to the Fc domain of human immunoglobulin G1 (proline 100 to lysine 330) through a peptide linker (IEGRMD). The anti-EGFR aptamer (Kd = 2.4 nM) and a mutant aptamer were extended with a capture sequence. The extended capture sequence did not participate in aptamer hairpin structure but was used to immobilize aptamer on the substrate through duplex formation with substrate-anchored DNA probe molecules. The sequences of the extended anti-EGFR aptamer, the extended mutant aptamer, and substrate-anchored probe were as follows: anti-EGFR aptamer, 5′-GGC GCU CCG ACC UUA GUC UCU GUG CCG CUA UAA UGC ACG GAU UUA AUC GCC GUA GAA AAG CAU GUC AAA GCC GGA ACC GUG UAG CAC AGC AGA GAA UUA AAU GCC CGC CAU GAC CAG-3′; mutant aptamer, 5′-GGC GCU CCG ACC UUA GUC UCU GUU CCC ACA UCA UGC ACA AGG ACA AUU CUG UGC AUC CAA GGA GGA GUU CUC GGA ACC GUG UAG CAC AGC AGA GAA UUA AAU GCC CGC CAU GAC CAG-3′; and substrate-anchored probe, 5′-amine-CTG GTC ATG GCG GGC ATT TAA TTC-3′ (the extended capture sequences are set in italics). The aptamer was modified by extending the DNA template at its 3′ end with a 24-nucleotide sequence tag and then hybridizing the transcribed, extended aptamer with a complementary substrate-anchored probe that was modified with an amine at its 5′ end.

Preparation of Nanotextured PDMS Substrates

Half a gram of poly(D,L-lactide-co-glycolide) (PLGA) (50/50 weight percentage; 12-16.5 × 103 molecular weight; Polysciences, Inc., Warrington, PA) was dissolved in 8 mL chloroform at 55 °C for 40 minutes.16, 21 The solution was cast into a glass Petri dish, allowed to sit overnight, and was placed in a vacuum chamber (15 in Hg) for 2 days at room temperature. The solid PLGA polymers were treated with 10 N NaOH for 1 hour to generate nanotextured surfaces16 and were further sterilized by soaking in ethanol for 24 hours followed by exposure to ultraviolet (UV) light for 1 hour. SYLGARD 184 Silicon Elastomer (Dow Corning, Midland, MI) was mixed (10:1 weight/weight) with a silicon resin curing agent. The mixture was placed in a vacuum chamber to remove all bubbles and then cast onto NaOH-treated PLGA polymer surface. Then, it was allowed to cure for 48 hours at room temperature to solidify. Finally, the PDMS was peeled from the PLGA. Before surface modification, the PDMS substrates were immersed in deionized (DI) water at 37 °C overnight to completely remove any residual PLGA.

Organic solvents like methanol and acetonitrile can cause PDMS bulk to dissolve and swell. The solubility parameters of ethanol and acetone are 12.7 cal1/2cm−3/2 and 9.9 cal1/2cm−3/2, respectively, and solvents that have a solubility parameter similar to that of PDMS (7.3 cal1/2cm−3/2) generally cause PDMS to swell more.22 The solubility parameters of methanol and acetonitrile are 14.5 cal1/2cm−3/2 and 11.9 cal1/2cm−3/2, respectively, and the swelling ratios are 1.02 and 1.01, respectively, which are less than the ratios of ethanol and acetone (1.06 and 1.04, respectively). Thus, we used methanol and acetonitrile for PDMS surface modifications. Furthermore, although methanol and acetonitrile can completely dissolve PDMS, the process takes extremely long time. Even di-isopropylamine, which has a swelling ratio as high as 2.13, still requires 1 month to dissolve PDMS completely. Thus, silanization and isothiocyanate molecule incubations were done for only 20 to 30 minutes, so that the swelling and dissolving of PDMS were insignificant.

Scanning Electron and Atomic Force Microscopy Characterization

A Zeiss Supra 55 variable pressure scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany) was used to qualitatively evaluate PDMS surface topography (Fig. 1D, inset). Samples were sputter-coated with gold at room temperature. Surface topography was evaluated quantitatively by using a Dimension 5000 atomic force microscope (AFM). The changes in surface area and root mean square surface roughness were measured. Height images of PDMS samples were captured in the ambient air with 15% to 20% humidity at a tapping frequency of approximately 300 kHz. The analyzed field measured 3 μm × 3 μm at a scan rate of 1 Hz with 256 scanning lines.

thumbnail image

Figure 1. These are images of the surface roughness of PLGA and polydimethylsiloxane (PDMS) cast on PLGA, which increased after NaOH etching. The atomic force microscope (AFM) photomicrographs (3 × 3 μm2) show (A) untreated PLGA and (B) PLGA after 10 N NaOH etching for 1 hour. The surface roughness increased from 22 nm on an untreated PLGA surface to 310 nm on a nanotextured PLGA surface. (C,D) These are AFM photomicrographs (8 × 8 μm2) of (C) a nanotextured PDMS surface (roughness, 347 nm) and (D) a nanotextured PDMS surface after aminopropyltriethoxysilane and thiophosgene modification (roughness, 289 nm). The inset in D shows a scanning electron microscope photomicrograph of an NaOH-treated PLGA surface. Scale bar = 100 μm in inset.

Download figure to PowerPoint

Attachment of Anti-EGFR Aptamer on PDMS and Glass Substrates

The attachment method was adopted from earlier descriptions.20, 23-25 The PDMS substrates and the glass slides were cut into 5 × 5 mm2 pieces and cleaned with UV-ozone plasma and piranha solution (H2O2:H2SO4, 1:3 ratio) for 30 minutes and 10 minutes, respectively. After rinsing with DI water and drying in a nitrogen flow, the PDMS and glass substrates were immersed in 2% (volume/volume) of 3-aminopropyltriethoxysilane (APTES) in methanol for 30 minutes at room temperature. Then, the substrates were rinsed sequentially with methanol and DI water. The amino groups on PDMS and glass substrates were converted to the isothiocyanate groups by introducing a 0.5% (volume/volume) thiophosgene solution in acetonitrile for 20 minutes at 40 °C. The substrates were then washed with DI water and dried in a stream of nitrogen. The amino-modified DNA capture probes were prepared at 30 μM concentration in 5 mM Tris buffer with 50 mM NaCl. A volume of 5 μL of DNA solution was placed on each substrate, and it was allowed to incubate in a humidified chamber at 37 °C overnight. Each substrate was then washed with DI water. Salmon sperm DNA was used for prehybridization to reduce RNA physical adsorption. A volume of 5 μL anti-EGFR RNA aptamer at 1 μM concentration was placed on each substrate in 1 × annealing buffer (10 mM Tris, pH 8.0; 1 mM ethylenediamine tetraacetic acid, pH 8.0; and 100 mM NaCl). After 1 hour of hybridization at 37 °C, the substrates were washed with 1 × annealing buffer and DI water for 5 minutes. The negative control devices were hybridized with mutant aptamer using the same protocol. The substrates were placed in 1 × phosphate-buffered saline (PBS) with 5 mM magnesium chloride, pH 7.5, and were used immediately.

Contact Angle Measurements

Contact angles were measured on isothiocyanate groups of modified PDMS (with/without nanotexturing), unmodified PDMS (with/without nanotexturing), and glass (with/without isothiocyanate group modification). A droplet of DI water was placed on the surface of the substrate at room temperature; and, after 30 seconds, the contact angle was measured using a contact angle goniometer (NRL-100; Rame-Hart, Washington, DC). On average, 5 measurements were calculated for each run.

Fluorescence Measurements of Fluorescamine

Surface modification was confirmed further by fluorescence measurements of fluorescamine. The density of surface-grafted amino groups from APTES was measured by fluorogenic derivatization reaction with fluorescamine.26 A mixture of 900 μL of 0.1% (weight/volume) fluorescamine dissolved in acetone, 150 μL of 0.1 M borate buffer, and 1.91 mL DI water was made. After APTES modification, glass, PDMS, and nanotextured PDMS samples were immersed into the fluorescamine mixture solution for 5 minutes at room temperature. All samples were rinsed with acetone to remove the excessive reagents. The fluorescence measurements were taken at 390 nm wavelength using Zeiss confocal microscope. The fluorescence intensities were analyzed with ImageJ software (National Institutes of Health, Bethesda, MD).

Human Glioblastoma and Meninges-Derived Primary Fibroblast Cell Culture

The hGBM cells were cultured in chemically defined, serum-free Dulbecco's Modified Eagle's Medium (DMEM)/F-12 medium supplemented with 20 ng/mL mouse EGF and 20 ng/mL of basic fibroblast growth factor (both from Peprotech, Rocky Hill, NJ), 1 × B27 supplement (Invitrogen, Carlsbad, CA), 1 × insulin-transferrin-selenium (Invitrogen), and penicillin:streptomycin 100 U/mL:100 μg/mL (HyClone, Wilmington, DE) and plated at a density of 3 × 106 live cells per 60 mm plate. The hGBM cells were stably transduced with a lentivirus expressing mCherry fluorescent protein. The primary rat meninges-derived fibroblasts were plated in T-75 tissue culture flasks in DMEM/F-12 medium containing 10% fetal bovine serum.

Tumor Cell Capture on Substrates

The cell suspensions were centrifuged, the supernatants were removed, and sterilized 1 × PBS solution (with 5 mM MgCl2) was added to dilute the cells. Approximately 500 μL of cell suspension in 1 × PBS were placed on each substrate surface. The substrates were incubated for 30 minutes at 37 °C and then washed with sterilized 1 × PBS on a shaker at 90 revolutions per minute for 15 minutes.27 For tumor-specific isolation studies, the hGBM cells were mixed with fibroblasts at a 1:1 ratio.

RESULTS AND DISCUSSION

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

Surface Topography of Nanotextured Substrates

The effects of NaOH concentration and etching times have been characterized previously.16, 17 PLGA bulk surface was etched with 10 N NaOH for 1 hour to generate nanotextured surface. First, the average surface roughness of PLGA substrates was analyzed quantitatively with AFM (Fig. 1A,B). The surface roughness increased from 22 nm on untreated PLGA to 310 nm on nanotextured PLGA after NaOH etching. The photomicrographs of PDMS cast on these 2 substrates are provided in Figure 1C,D. The nanotextured surfaces created on PDMS before and after chemical modification had a roughness of 347 nm and 289 nm, respectively.

Contact Angle Measurements

Contact angle data from a water droplet provide a measure of the hydrophobicity of a surface.25 We measured the contact angles of each substrate. The average contact angles (n = 10) and standard deviations are listed in Table 1. After APTES and isothiocyanate modification, all 3 types of substrates had hydrophilic surfaces. The aptamer immobilization would further decrease the contact angle and make these substrates more hydrophilic. It is known that the hydrophilic surfaces have lower protein and cell physical adsorption. Moreover, the roughness of a surface can significantly affect the contact angle.28 In other words, the contact angle decreases on a nanotextured hydrophilic surface, whereas it increases on a nanotextured hydrophobic surface. Even with more hydrophilic nanotextured surface, we observed tumor cell isolation.

Table 1. Contact Angles Measured Immediately After Ultraviolet-Ozone Treatment of the Substrates and Chemical Activation with Phenylene Di-Isothiocyanate for Each of the Substrates Employed in The Current Study
 Contact Angle±SD, Degrees
Substrate TypeBase SubstrateAfter PDITC Treatment Resulting in N[DOUBLE BOND]C[DOUBLE BOND]S Groups on the Surface
  1. Abbreviations: PDITC, phenylene di-isothiocyanate; PDMS, polydimethylsiloxane; SD, standard deviation.

Glass46±151±1
PDMS without nanotexturing115±259±2
PDMS with nanotexturing144±446±3

PDMS initially has methyl groups on both sides of the backbone; however, after UV-ozone treatment, the methyl groups are substituted with hydroxyl groups. The residual methyl groups on the PDMS surface still contribute to hydrophobicity; consequently, the PDMS surface contact angle was higher than that for the glass surface even after APTES and isothiocyanate group modification (Table 1). Because of increased surface roughness, the nanotextured PDMS substrate had the largest contact angle,29 but this decreased to the lowest number after APTES and isothiocyanate modification (from 144° to 46°). The etching of NaOH created an anisotropic surface on PLGA, which, in turn, produced the same effect on PDMS when it was cast on PLGA. The strong anisotropy of the nanotextured PDMS surface caused the contact angle to vary; thus, the standard deviation of nanotextured PDMS also was higher than that for the other groups.30

On the PDMS surface, the hydroxyl groups that were created after the surface oxidation would gradually change back to methyl groups, sometimes in very short period.31, 32 Thus, it was important to perform the subsequent chemical modification relatively quickly after UV-ozone treatment of PDMS. In addition, it is worth noting that long UV-ozone treatment (>90 minutes) can make the PDMS stiffer and can create lots of tiny cracks on the surface.33, 34 In these experiments, the PDMS surfaces were treated with UV-ozone for just 30 minutes to minimize such textural aberrations.35

Fluorescence Measurements

Proper oxidization of the PDMS surface can significantly increase the number of hydroxyl groups, can further increase the number of available amino groups from APTES, and, finally, can improve the total number of immobilized aptamers. The relative number of amino groups on different samples was determined by comparing the relative fluorescence intensities of fluorescamine on each sample. More amount of available on the surface improved the total number of immobilized aptamers, which favored tumor cell isolation. Fluorescamine is intrinsically nonfluorescent, but its reaction with amino groups results in highly fluorescent derivatives. The glass substrates already had hydroxyl groups on the surface, so these underwent only the amine treatment.

The average fluorescence intensities of 3 types of samples are provided in Table 2. Nanotextured PDMS had the highest intensity (83.9 ± 14.1 arbitrary units). The nanotextured surface increased the effective surface area. Consequently, the nanotextured surface generated a significantly higher number of hydroxyl groups compared with the number generated on the plain PDMS or glass surfaces; thus, more amino groups were introduced on the surface after silanization. It has been demonstrated that the amino group concentration can reach 4 × 10−8 mol/cm2.36

Table 2. Fluorescence-Intensity Data for Glass, Polydimethylsiloxane, and Nanotextured Polydimethylsiloxane Substrates After 3-Aminopropyltriethoxysilane and Fluorescamine Modification
Substrate TypeFluorescence Intensity: Mean±SD, AU
  1. Abbreviations: AU, arbitrary units; PDMS, polydimethylsiloxane; SD, standard deviation.

Glass4.7±1.5
PDMS without nanotexturing52.7±6.3
PDMS with nanotexturing83.9±14.1

The increased number of available aptamers on the surface was favorable for tumor cell isolation. On planar substrates, the density of the anchored probe DNA can be approximately 1 per 4 nm2 or 5 nm2.37 The packing density is a function of the radius of gyration of the probe molecules, which defines the footprint of a probe molecule and, thus, how densely the molecules can be packed.38 On the nanotextured surface, there are 2 advantages. First, the distance is reduced between the immobilized ends of the probes, because the free ends have more room on a curvaceous surface and, thus, require a smaller footprint than that required on a plane surface for the same radius of gyration. Thus, the probe density would be far greater on curvaceous surfaces. Second, on the nanotextured substrate, there is more effective area than the area available on a plane surface of same cross-sectional dimension. A reduced distance between adjacent probes and a larger effective area provide very high probe packing density. The nonspecific adsorption of aptamers on the surface would occur from Van der Waals forces only if aptamers can find their way to the surface. The negative charges from the tightly packed DNA probes on the surface repel the aptamers from inserting into the spaces between adjacent probes and, thus, impede the aptamers from reaching the surface of the substrates, where they can bind nonspecifically. However, on nanotextured PDMS surfaces, the upper space between probes is widened because of the effectively curvaceous surface created by nanotexturing, and the significantly larger effective surface area increases the total number of probes, reducing nonspecific aptamer adsorption but increasing more densely packed aptamers.

Isolation of hGBM Cells

Figure 2A-F depicts representative images of hGBM cells captured on glass, PDMS, and nanotextured PDMS substrates with anti-EGFR or mutant aptamers. The average density of cells (±standard deviation) on substrates before washing was 400.9 ± 43.3 per mm2. All substrates were washed with 1 × PBS at 90 revolutions per minute for 15 minutes. Fluorescent images of cells on 10 substrates of each type were taken. The quantitative analysis results are provided in Figure 2G. On average, 149.6 ± 12.2 hGBM cells were captured per mm2 on anti-EGFR aptamer-modified, nanotextured PDMS substrate. Conversely, 79.3 ± 11.5 cells per mm2 and 37.4 ± 10.1 cells per mm2 were captured on anti-EGFR aptamer-modified glass and PDMS substrates, respectively. Four major factors influenced the cell capture: the available number of anti-EGFR aptamer molecules on the substrate, the EGFR density on the cell membrane, the affinity between EGFR and the aptamer, and the surface quality of the substrate. We deduce that the available number of aptamer molecules is a direct function of surface nanotexturing. Cell isolation efficacy can be improved by increasing the affinity between surface bound aptamer and the overexpressed EGFR. In this case, the higher affinity came from nanotexturing, which increased the quantity of aptamers on the surface.

thumbnail image

Figure 2. Human glioblastoma (hGBM) cells on the anti-epidermal growth factor receptor (anti-EGFR) and mutant aptamer-modified glass, polydimethylsiloxane (PDMS), and nanotextured PDMS (N-PDMS) substrates. Substrates were incubated with hGBM cells and washed with phosphate-buffered saline. The hGBM cell densities (number of cells per mm2) on the anti-EGFR aptamer-modified (A) glass, (C) PDMS, and (E) N-PDMS substrates (±standard deviation) are 79.3 ± 11.5 cells per mm2, 37.4 ± 10.1 cells per mm2, and 149.6 ± 12.2 cells per mm2, respectively; and the cell densities on the mutant aptamer-modified (B) glass, (D) PDMS, and (F) N-PDMS substrates are 2.2 ± 1.2 cells per mm2, 0.6 ± 0.8 cells per mm2, and 25.6 ± 6.2 cells per mm2, respectively (P < .05). (G) The plot illustrates the average hGBM cell density on each type of substrate. The table in the inset depicts the actual numbers used in the plot. Scale bar = 100 μm in B (also applies to A and C-F).

Download figure to PowerPoint

Fluorescamine analysis demonstrated that the nanotextured PDMS could generate more hydroxyl groups after oxidization; therefore, more amino groups from APTES could be attached on the surface after silanization, ultimately resulting in increased numbers of available anti-EGFR aptamer molecules. Thus, the density of immobilized anti-EGFR aptamer increased. In addition, the nanotextured surface mimicked the basement membrane, facilitating cell attachment. Therefore, the number of captured cells on the nanotextured PDMS substrate was higher than the numbers in the other 2 groups. A flat PDMS surface also can generate more hydroxyl groups after oxidization, as discussed above, and a few nanometers of rough texturing can be achieved with long UV-ozone treatment.39 However, on a flat PDMS surface, even after chemical functionalization, it is still a major challenge to maintain cells on the surface, especially in long-term cell culture on PDMS, because a stable, cell-adhesive layer is not easy to form.40 Moreover, the generated hydroxyl groups undergo dehydration reaction and reform silicon-oxygen-silicon (Si-O-Si) bonds, and the high chain mobility pulls the hydrophobic methyl groups to the surface. These 2 factors can prohibit the formation of a stable, cell-adhesive layer. Thus, the number of captured cells on a PDMS surface is lower than that on a glass and nanotextured PDMS surface. This also can happen on nanotextured PDMS substrate; however, the nanotexturing itself provides a tradeoff by improving cell attachment and isolation. The data indicate that nanotextured surfaces produce improved cancer cell isolation, but nonspecific cell attachment also increases. Increased surface area provides more sites not only for protein adsorption but also for focal contact adhesion sites, which are used by cells to attach to the surfaces. Therefore, the nanotextured surfaces are better suited for overall cell adhesion goals. In the control group, cell density on mutant aptamer-modified, nanotextured PDMS substrate was 25.6 ± 6.2 cells per mm2, which was almost 12 times higher than the cell density on a glass substrate. Obviously, the higher physical absorption decreases isolation specificity, but it also significantly improves detection sensitivity. In practical applications, the selection of material and surface texture depends on the competing goals of isolation sensitivity and specificity. The photomicrographs in Figure 3 reveal the captured cells on PDMS, nanotextured PDMS, and glass surfaces. After 30 minutes of incubation, hGBM cells formed pseudopods, indicating that cells could firmly attach to nanotextured PDMS surfaces. This phenomenon was not observed on smooth PDMS surfaces. Conversely, the morphology of hGBM cells on nanotextured PDMS surfaces was flatter than that on smooth glass and PDMS surfaces. Cells had a globular shape from strong repulsion by the hydrophobic smooth PDMS surface. Although cells on a glass surface also could form pseudopods and had semielliptical shapes, whole cells did not spread as well as they did on a nanotextured PDMS surface. In short, the phenomenon of spreading was more pronounced on nanotextured PDMS than that on smooth PDMS or glass. The topography of substrate resulted in differential cell spreading that indicated much more distinct behavior on the nanotextured PDMS. Conversely, cells on smooth PDMS and glass surfaces still maintained a round or semielliptical shape. This serves as a novel and important cytologic behavior that can help distinguish between CTCs and other captured cells.

thumbnail image

Figure 3. These are scanning electron microscope photomicrographs of captured tumor cells on (A) polydimethylsiloxane (PDMS), (B) nanotextured PDMS, and (C) glass substrates. The photomicrographs reveal that cells firmly attached to the rough surface, which mimics the basement membrane structure. Cells were fixed in 4% paraformaldehyde for 3 hours; then, the substrates were immersed into 20%, 30%, 50%, 70%, 85%, 95%, and 100% (volume/volume) ethanol concentration gradient solutions (15 minutes in each solution). All substrates were lyophilized overnight. Scale bar = 1 μm in A (also applies to B and C).

Download figure to PowerPoint

Isolation of hGBM Cells From Cell Mixture

A mixture of hGBM cells and fibroblasts was prepared at a ratio of 1:1. The average density of plated cells on the surface was 332.3 ± 23.6 cells per mm2. The aptamer-functionalized nanotextured PDMS substrates were incubated in the cell mixture, washed, and imaged. Both differential interference contrast images and fluorescent images were obtained. The photomicrographs in Figure 4A,B) were obtained from a random area of the nanotextured PDMS substrate. Approximately 18.4% ± 9.1% hGBM cells from 10 substrates did not have any fluorescence, although they did express mCherry fluorescent protein. Data from the mixture group revealed no fluorescence in approximately 31% of cells from 10 substrates (846 of 2729 cells in the mixture of hGBM cells and fibroblast cells; average, 58.8 ± 21.4 cells per mm2). The cells that were not revealed in the fluorescence images included captured nonfluorescent hGBM cells (18.4% of total hGBM cells) and nonspecifically bound fibroblast cells; therefore, 2307.6 hGBM cells and 431.4 fibroblasts cells were counted. On average, approximately 15.4% of the captured cells were fibroblasts. Thus, the aptamer-functionalized, nanotextured PDMS was capable of selectively isolating and enriching a 1:1 mixture suspension of fibroblasts and cancer cells to 1:5.5 on the surface. Compared with previously published work on smooth glass substrate,20 the ratio of hGBM cells and fibroblasts decreased from 1:8.24 (for glass) to 1:5.5 (for nanotextured PDMS). These results indicate that the nanotextured PDMS substrates also led to the attachment of fibroblasts, and the increased number of fibroblasts could be attributable to the nanotexturing and the higher number of available aptamers on the surface that also would bind to the EGFR on the surface of fibroblasts. Although the nanotextured PDMS substrate increased the attachment of fibroblasts, in any case, it still could specifically capture hGBM cells and improve the ratio from 1:1 to 1:5.5. The increased sensitivity decreases its specificity, but the tradeoff is the advantage of isolating as many of the small number of cancer cells as possible. In practical applications, the selection of material and surface structure depends on the goals of isolation sensitivity or specificity.

thumbnail image

Figure 4. Human glioblastoma (hGBM) cells and fibroblasts are observed on nanotextured polydimethylsiloxane (PDMS) substrates. Substrates were incubated with a mixture of hGBM cells and fibroblasts and were washed with phosphate-buffered saline. Shown are (A) a differential interference contrast image and (B) a fluorescent image from the same position on the substrate. The circles in A indicate a few hGBM cells that did not fluoresce and the fibroblasts that were captured and cannot be seen in B. Scale bar = 100 μm.

Download figure to PowerPoint

In conclusion, the current study demonstrated that anti-EGFR RNA aptamer-modified, nanotextured PDMS substrates can capture more hGBM cells compared with traditional smooth-glass substrates; moreover, the nanotextured PDMS substrate still can specifically recognize, capture, and isolate hGBM cells from a mixture of fibroblasts. The nanotextured surface simulates the basement membrane structure and can facilitate tumor cell isolation. This can have important implications for chip-based cancer cell isolation substrate selection.

FUNDING SOURCES

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

This work was supported by a National Science Foundation CAREER grant (ECCS-0845669 to S.M.I.). Y.T.K. was supported by the UTA Nano-Bio Cluster Program. A.D.E. was supported by the Welch Foundation (grant F-1654) and by a National Cancer Institute award (5R01CA119388-05).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

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