Biosensors and Bioelectronics 69 (2015) 272–279

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

A fluorescence in situ hybridization (FISH) microfluidic platform for detection of HER2 amplification in cancer cells$ Kai-Jie Kao a, Chien-Hsuan Tai a, Wen-Hsin Chang a, Ta-Sen Yeh d, Tse-Ching Chen e, Gwo-Bin Lee a,b,c,n a

Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan Institute of Biomedical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan c Institute of NanoEngineering and Microsystems, National Tsing Hua University, Hsinchu 300, Taiwan d Department of Surgery, Chang Gung Memorial Hospital at Linkou, Chang Gung University, Taoyuan 333, Taiwan e Department of Pathology, Chang Gung Memorial Hospital at Linkou, Chang Gung University, Taoyuan 333, Taiwan b

art ic l e i nf o

a b s t r a c t

Article history: Received 29 January 2015 Accepted 2 March 2015 Available online 3 March 2015

Over-expression/amplification of human epidermal growth factor receptors 2 (HER2) is a verified therapeutic biomarker for breast and gastric cancers. HER2 is also served as prognostic biomarker for gastric cancer because HER2 over-expression is associated with a 5–10% increase in cancer related death of gastric cancer. Cancer patients exhibiting HER2 over-expression can significantly improve their overall survival rates by taking the targeting drug Herceptin, which directly targets HER2. However, Herceptin has limited functions toward patients without HER2 over-expression and therefore it needs a highly specific and accurate detection method for diagnosis of HER2 over-expression. Currently, fluorescence in situ hybridization (FISH) technique is routinely employed to detect HER2 amplification. However, it is a labor-intensive, time-consuming hybridization process and is relatively costly. Furthermore, well-trained personnel are required to operate the delicate and complicate process. More importantly, it may take 1–2 days for well-trained personnel to perform a whole FISH assay. Given these limitations, we developed a new, integrated microfluidic FISH system capable of automating the entire FISH protocol which could be performed within a shorter period of time when compared to traditional methods. The microfluidic FISH chip consisted of a microfluidic control module for transportation of small amounts of fluids and a hybridization module to perform the hybridization of DNA probes and cells/tissue samples. With this approach, the new microfluidic chip was capable of performing the whole FISH assay within 20 h. Four cell lines, two for non-HER2 amplification and two for HER2 amplification, and two clinical tissue samples, one for non-HER2 amplification and another for HER2 amplification, were used for verifications of the developed chip. Experimental data showed that there was no significant difference between the benchtop protocol and the chip-based protocol. Furthermore, the reagent consumption was greatly reduced (∼70% reduction). Especially, only 2-μl usage for FISH deoxyribonucleic acid (DNA) probe was used, which is five-fold reduction when compared with the traditional method. It is the first time that the entire FISH assay could be automated on a single chip by using tissue samples. The microfluidic system developed herein is therefore promising for rapid, automatic diagnosis of HER2-related diseases by detecting the HER2 gene with minimal consumption of samples and reagents and has a great potential for future pharmacogenetic diagnostics and therapy. & 2015 Elsevier B.V. All rights reserved.

Keywords: Biomarker Cancer Microfluidics Fluorescence in situ hybridization (FISH) HER2 Herceptin

Abbreviations: Bio-MEMS, Bio-micro-electro-mechanical-systems; CEP17, Chromosome 17 centromere; CGMH, Chang Gung medical hospital; DNA, deoxyribonucleic acid; DAPI, 4′-diamidino-2-phenylindole; FISH, fluorescence in situ hybridization; HER2, human epidermal growth factor receptor 2; MEMS, micro-electro-mechanical-systems; NaSCN, sodium thiocyanate; PBS, phosphate buffered saline; PCR, polymerase chain reaction; PDMS, polydimethylsiloxane; PMMA, polymethylmethacrylate; RSD, relative standard deviation; SSC, sodium chloride–sodium citrate buffer; ddH2O, double-distilled water; rpm, revolutions per minute ☆ The preliminary results of this study have been presented in the 17th International Conference on Miniaturized Systems for Chemistry and Life Sciences (MicroTAS 2013). n Corresponding author at: Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan. Fax: þ 886 3 5742495. E-mail address: [email protected] (G.-B. Lee). http://dx.doi.org/10.1016/j.bios.2015.03.003 0956-5663/& 2015 Elsevier B.V. All rights reserved.

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

1. Introduction Detections of the HER2 over-expression become more and more important recently especially when HER2 over-expression is found to be associated with a 5–10% increase in cancer related death of gastric cancer (Hsu et al., 2011). Previous study indicated that cancer patients demonstrating HER2 over-expression can significantly improve their overall survival rates by taking the targeting drug Herceptin, which directly targets HER2 (Hsu et al., 2011). Since Herceptin is relatively costly and it has limited functions toward patients without HER2 over-expression, a highly specific and accurate detection method is needed to prevent unnecessary medical cost and drug abuse. An abnormality in an individual's genome due to, for instance, exposure to a carcinogenic substance, can lead to a variety of genetic disorders or cancers. Traditionally, polymerase chain reaction (PCR) and reverse transcription polymerase chain reaction have been commonly used to detect human genetic disorders (Bruneau et al., 1990). However, these methods are inefficient for analysis of numerous genes in parallel. Alternatively, fluorescence in situ (FISH) hybridization is an important technique with high specificity for localizing mutant genes, or their respective mRNAs, within individual cells (Lichter et al., 1990; Pinkel et al., 1988; Vanneste et al., 2009; Langer-Safer et al., 1982), as well as cancers (Fox et al., 1995; Leversha et al., 2009; Nath and Johnson, 2000). FISH can allow for observations of genetic alterations, such as rearrangements and translocations, which are crucial indicators for certain cancers. FISH technique is routinely used to detect HER2 amplification (Young et al., 1995). However, it needs a labor-intensive, time-consuming process and is relatively costly. Furthermore, it also needs well-trained personnel to operate the delicate and relatively complicate process. Recently, detection of cancer biomarkers by FISH has become increasingly important for the early diagnosis, prognosis, and treatment of cancer (Mitri et al., 2012). As mentioned above, genes such as HER2 have been associated with many types of cancers including breast cancer and gastric cancer. Furthermore, survival rates of patients demonstrating HER2/HER2 amplification/overexpression can be enhanced by Herceptin targeting therapeutics. However, FISH-based HER2 over-expression detection analyses require expensive reagents and skilled personnel. Also, in spite of long hybridization times and tedious washing steps, the technician may inadvertently contaminate the samples, thus compromising the accuracy of the analysis. Therefore, there is an urgent need to develop methods that can rapidly perform detection assays for over-expression of HER2 in an automated manner. Micro-electro-mechanical-systems (MEMS) would be a solution for abovementioned issues. MEMS is a technique used to manufacture miniature systems that consist of microsensors, microactuator and micro-circuits (Li, 2002). Miniaturized systems have several advantages over their conventional counterparts, including size compactness, low power consumption, enhanced performance and reliability (Courtois and Blanton, 1999). Recently, a technology known as bio-micro-electro-mechanical-systems (Bio-MEMS) has been exploited for use in biosensing, drug delivery, biomedical engineering, and a variety of biomedical and chemical applications (Velten et al., 2005). Bio-MEMS technology is capable of producing microdevices comprised of low-cost materials that are associated with lower degrees of contamination due to their usage of lower reagent volumes and their disposability to a single use. Moreover, the miniaturized microfluidic systems developed by Bio-MEMS have demonstrated similar, or even superior, performance to their large-scale counterparts. Therefore, with Bio-MEMS technology, diverse processes and modules can be integrated into a single chip in order to reduce the amounts of samples and reagents, and power consumption. Furthermore,

273

reaction times and overall costs could be dramatically reduced while simultaneously increasing sensitivity, throughput, portability, and potential for integration and automation (Byun et al., 2008). Bio-MEMS and microfluidic-based technologies could meet these diagnostic demands. Recently, two Bio-MEMS and microfluidic-based systems have been reported for FISH assays. For instance, Mayer et al. employed an OncoCEETM microchannel technology to a pilot clinical trial for FISH detection of HER2 (Mayer et al., 2011). The adequate sensitivity and specificity of the OncoCEETM microchannel were demonstrated in this work. However, the sample should be subjected to complicated manual pretreatment step before applied to the OncoCEETM microchannel. Another Bio-MEMS and microfluidic-based system was reported by our group previously (Tai et al., 2013).This system integrated all manual steps into a microfluidic system so that a total automated fashion of FISH can be realized. However, the samples used in this study were cell lines only. In the hospitals, most of tissue samples of gastric cancer are paraffin sections which have undergone a series of dedicated steps in order to be used for FISH assays. Therefore, in this work, a new integrated FISH microfluidic platform aimed to automate the entire FISH assay for the detection of HER2 by using tissue samples is proposed. The integrated microfluidic system which is capable of performing the entire FISH protocol from tissue sample pre-treatment to DAPI staining to diagnose HER2 over-expression is demonstrated. It integrated several functional devices, including microvalves, micropumps, and reaction chambers that were capable of transporting, mixing, incubating, and storing multiple reagents and samples, on a single chip. Specifically, a fluid control module to control fluids transport, a temperature control system to regulate the temperature and a hybridization module to enhance the hybridization process were integrated to perform the whole FISH protocol. More importantly, the reagent consumption was significantly reduced (∼70% reduction). To the best of our knowledge, this is the first work which could automate the entire FISH assay by using tissue samples.

2. Materials and methods To verify the performance of the FISH chip, four breast cancer cell lines and two clinical tissue samples, including non-HER2 amplification cases as negative controls and HER2 amplification cases as positive controls demonstrating HER2 over-expression, were tested. The protocols for the cultured cells and tissue samples differed slightly, but both types of samples could be used with the chip designed herein. Therefore, the FISH microfluidic chip could be used with dual functions, which has never been demonstrated in literature. 2.1. Cell lines Preparation In order to investigate whether HER2 genes could be detected on the FISH microfluidic chip, the breast cancer cell lines obtained from Chang Gung Medical Hospital [CGMH]-MDA-MB-468 (nonHER2 amplification case), MCF-7 (non-HER2 amplification case), SK-BR-3 (HER2 amplification case), and MDA-MB-361 (HER2 amplification case) were used. The cells were cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum and maintained in a 37 °C incubator. After trypsinization, suspended cells at a concentration of 5  106 cells/ml were washed twice with phosphate buffered saline (PBS) and spun at 1200 revolutions per minute (rpm) for 5 min, and the culture media supernatant was decanted. Then, cells were re-suspended in 1–2 ml Carnoy's fixative prior to application on the glass slide. Fixed cells (10 μl) were deposited on the glass slide and incubated at room

274

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

temperature for several seconds to allow the fixative to evaporate. The glass slides were then frozen for 30 min at  20 °C before performing on-bench and on-chip FISH analyses. 2.2. Tissue samples All clinical tissue samples were obtained from CGMH (Institutional Review Board (IRB) number 101-3307A3). 2.3. FISH analyses For FISH analyses of cell lines, the HER2 DNA Probe Kit was purchased from Abbott (Illinois, USA), and the manufacturer's instructions were followed. Briefly, 2  sodium chloride-sodium citrate buffer (SSC) (pH 5.3) heated to 73 °C was first injected into the reaction chamber with micropumps (as shown in Fig. 1(a)) to simulate the slightly acidic conditions in the stomach. Treatment with 0.0025% pepsin was then administered to digest membrane proteins. After PBS washing and fixation in 1% formaldehyde, the cells were then sequentially dehydrated for 1 min in 70%, 85%, and 100% ethanol. Then, a 2-μl probe solution containing an orangefluorescent probe specific for the HER2 gene and a green-fluorescent probe specific for the chromosome 17 centromere (CEP17) was transferred to the reaction chamber. DNAs were denatured at 75 °C for 5–10 min and hybridized for 16 h at 37 °C. After washing

away unbound probes, 4′, 6′-diamidino-2-phenylindole dihydrochloride (DAPI) solution was used to stain the nuclei. Finally, fluorescent images were acquired with a fluorescence microscopy (Leica DM2500, Leica Microsystems, Singapore). Table 1(a) shows the detailed procedure of the cultured cell line protocol. FISH analyses of tissue samples, a paraffin-embedded tissue with a dimension of 5 mm  5 mm was first sliced (2.5-μm thickness) and placed on a glass slide below the reaction area. In general, the protocol was similar to that used with the cultured cells, though sodium thiocyanate (NaSCN) was used instead of SSC, and double-distilled water (ddH2O) was used instead of PBS. NaSCN was found to better extract cross-linked DNA and protein molecules from the formaldehyde, and ddH2O was less likely to cause artifacts resulting from osmotic pressure changes. When the background noise of the fluorescent signal was too high, the NaSCN incubation was prolonged. Also, the length of the pepsin incubation varied from sample to sample as a function of the extent of the intactness of the extracellular protein matrix surrounding the target cells. Table 1(b) lists the detailed protocol for FISH analysis of HER2 over-expression in tissue samples. Additional differences between the two protocols can be found in Fig. 1 (b) and Table 1.

Fig. 1. (a) Schematic diagram of the FISH chip for use with cultured cells. (b) Schematic diagram of the FISH chip for use with clinical tissue samples. (c) An exploded view of the FISH microfluidic chip composed of (a) an air layer, (b) a liquid chamber layer, and (c) a glass slide. (d) A photograph of the microfluidic chip with dimensions of the 2.2 cm  5.7 cm. The blue color indicated the liquid layer, and the red color indicated the air layer. Ø: diameter. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

Table 1 (a) The detailed protocol for FISH analysis of HER2 over-expression in cell lines. Summary of FISH chip protocol steps (cell line)

Time (min)

Volume (ml)/ benchtop

Volume (ml)/ chip

2  SSC (pH 5.3; 73 °C) 0.0025% pepsin (37 °C) 1  PBS (room temperature) 1% formaldehyde (room temperature) 1  PBS (room temperature) 70, 85, 100% ethanol (room temperature) Hybridization (37 °C) 2  SSC with 0.3% NP-40 (pH 7.0–7.5; 75°C) DAPI (room temperature) Chip storage (  20°C)

3 15 5 5

30 10 30 5

35  10  3 35  10  3 35  10  3 35  10  3

5 3

30 90

35  10  3 35  10  3

960 5

1  10  3 30

2  10  3 35  10  3

1 30

2  10  3

2  10  3

(b) The detailed protocol for FISH analysis of HER2 over-expression in tissue samples Summary of FISH chip protocol Time Volume (ml)/ Volume (ml)/ (tissue samples) (min) benchtop chip 1 M NaSCN (80 °C) 40 30 35  10  3 ddH2O (room temperature) 3 30 35  10  3 0.0025% pepsin (37 °C) 5 10 35  10  3 ddH2O (room temperature) 3 30 35  10  3 70, 85, 100% ethanol (room 3 90 35  10  3 temperature) Hybridization (37 °C) 960 1  10  3 2  10  3 2  SSC with 0.3% NP-40 (pH 5 30 35  10  3 7.0–7.5; 75 °C) DAPI (room temperature) 1 2  10  3 2  10  3 Chip storage (  20 °C) 30

2.4. FISH microfluidic chip design Previous studies reported FISH chips that could be used to detect certain genes in merely cell lines without full automation (Vedarethinam et al., 2010; Tai et al., 2013). Furthermore, detections of clinical tissue samples are quite crucial in practical applications. Therefore, a new FISH microfluidic chip (as shown in Fig. 1(a) and (b)) was designed to deliver both samples and reagents and readily hybridize probes to both cultured cells and clinical tissues in an automatic format. Note that the chip design is quite different from our previous work (Tai et al., 2013) since we performed different FISH assays and used tissue samples. There were two major micro-components of the integrated chip, which contained two micropumps and several normally-closed microvalves. When the air chamber above the micropump was applied with a negative gauge pressure (vacuum), the thin membrane of the micropump above the liquid channel could be deflected upwards to form a local pressure drop. This pressure difference allows the reagent fluids to be transported to the reaction chamber below the micropump. For FISH analyses, 2  SSC (pH 5.3) heated to 73 °C was first injected into the reaction chamber while activating micropump. Treatment with 0.0025% pepsin was then administered to digest membrane proteins. After PBS washing and fixation in 1% formaldehyde, the cells were then sequentially dehydrated for 1 min in 70%, 85%, and 100% ethanol. Then, 2-μl probe were transferred to the reaction chamber by activating the micropump. DNAs were then denatured at 75 °C for 5–10 min and hybridized for 16 h at 37 °C at the reaction chamber. After washing away unbound probes, DAPI solution was used to stain the nuclei. Detailed differences between the cell and the tissue protocols can be found in Fig. 1(a) and (b) and Table 1. After the FISH assays, both on-bench and on-chip slides were scanned by a high-resolution scanner (Axio Imager, ZEISS, Germany) with a software (Metasystems, Metacyte, Germany) to obtain the fluorescent images of the cells.

275

Fig. 1(c) shows an exploded view of the FISH microfluidic chip composed of an air layer, a liquid chamber layer, and a glass slide. The air layer and the liquid chamber layer were made from polydimethylsiloxane (PDMS) (Sylgad 184A/B, Dow Corning Corp., USA). Fig. 1(d) shows a photograph of the microfluidic chip with dimensions of the 2.2 cm  5.7 cm. The radii of the reaction chamber and the bigger micropump were 2.5 and 5.0 mm, respectively. The main reaction chamber was located below the smaller micropump. Note that, due to its small size, the smaller micropump generated insufficient suction force to move reagents. Therefore, a bigger micropump was added to aid in reagent transport. Microvalves were responsible for opening or closing the microchannels. Electromagnetic valves were used to apply air pressure to either open or close the appropriate microstructures mentioned above. The microfluidic FISH system was equipped with a dual temperature control module comprising two thermoelectric (TE) coolers which could be controlled independently without interfering each other (Fig. S2) (Tai et al., 2013, Hung et al., 2014). Heating regions on two sides of the microchip worked in parallel to rapidly establish the two temperatures needed for the FISH reactions. The heating area provided the high temperature needed to denature the DNA prior to probe binding, and the cooling area prevented reagents from high temperature-induced degradation. Note that the PDMS structures and the glass slide were assembled together (Fig. S2) using a polymethymethacrylate (PMMA) cover and several screws. A hybridization module, which was indeed a micropump with a small volume around 2 μl, was used in this study to facilitate the hybridization of the DNA probe and the cells. Before the reagent injection, a negative gauge pressure was applied at the waste chamber to form a vacuum environment inside. Afterwards, the DNA probe was transported to the reaction chamber by activating the hybridization module (low-volume micropump). Not only did the lower volume of the reaction chamber reduce the reagent consumption, but it also shortened the distance between the suspended DNA probes and the cells/tissue attached below. With this hybridization module, the critical FISH hybridization could be stably performed with 2-μl probes within a shorter heating time. 2.5. Chip fabrication The FISH microfluidic chip consisted of two PDMS layers, including an air layer and a liquid chamber layer. Both layers were produced by using the PMMA and PDMS molding technique (Yang et al., 2010). Briefly, molds with inverse microfluidic structures on PMMA were manufactured by a computer-numerical-control machining process (EGX-400, Roland Inc., Japan) which was equipped with a 550-μm drill bit. The rotational speed and the moving rate of the drill bit were set to be 28,000 rpm and 12 mm min  1, respectively. Detailed information could be found in the Supplemental Information (Fig. S1). PDMS was then used to replicate the inverse microstructures of the micropumps, the microvalves, the microchambers and the microchannels on PMMA molds. After PDMS curing, the air layer and the liquid chamber layer were mechanically peeled off and then bonded together by using an oxygen plasma treatment (CUTE-MPR, UVOTECH Systems Inc., USA). Finally, the PDMS structures were clamped with a glass slide containing either cultured cells or tissue samples on the reaction chamber to form the FISH microfluidic chip. Note that no bonding between the PDMS layers and the glass slide was performed such that one could detach the chips afterwards for optical scannig of FISH chips.

276

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

2.6. Suction-type and compression-type micropumps In this work, a suction-type micropump was adopted to drive rapid fluid motion in order to, for instance, transport samples and reagents (Weng et al., 2011). Briefly, the thin membrane above the liquid channel could be deflected upwards to create a vacuum while a negative gauge pressure was applied. This pressure difference allows the reagent fluids to be transported to the reaction chamber below the micropump. Note that the suction-type micropump can be also used to mix fluids located in different chambers while operated at different modes (Weng et al., 2011). Similar in function but operating under different working principles, the compression-type micropump has also been used to rapidly move samples/reagents in a controlled setting (Huang et al., 2006). Briefly, the membrane above the liquid channel could be deflected downwards to build a “compression” condition because of the force driven by applying a positive gauge pressure. Also, the reagent fluids can be transported to the reaction chamber below the micropump due to this pressure difference. Note that the compression-type micropump can be also used in conjunction with a suction-type micropump to achieve efficient mixing between two chambers.

Fig. 2. The relationship between the pumping rate and the applied gauge pressure at different cycle durations. Error bars represented standard deviation (n¼3).

Leakage could be a serious issue on fluid transportation and bubbles could hinder the transportation of reagents and samples. Therefore, the clamp was used to provide the FISH chip a normal force to seal chambers and channels. Fig. S3(a) shows that no leakge and bubbles were observed when transporting fluids.

2.7. Statistical analysis

3.2. FISH analysis using cell lines and tissue samples

The relative standard deviation (RSD) and the two-tailed Student's t-test were used for performing statistical analyses in comparing data of fluorescent spots between the benchtop and the microchip methods. Moreover, when p-value p 40.05, it was considered as no statistical difference.

In order to test the microfluidic FISH assay, the breast cancer cell lines MDA-MB-468 and MCF-7, which are known to be without HER2 over-expression, were used as negative controls, whereas SK-BR-3 and MDA-MB-361, which are known to overexpress HER2, were used as positive controls. Upon the completion of the microfluidic FISH assay, HER2 genes spots were observed in fluorescence images (Figs. 3(a)–(d)). In these figures, HER2 genes were depicted in orange, while chromosome 17 centromere (CEP17) genes were green. Note that CEP17 was used as a reference here (Lambein et al., 2011). HER2 over-expression was therefore defined as the ratio between the numbers of fluorescence spots associated with the two genes (HER2/CEP17); a ratio greater than 2.2 with 60 cells indicates HER2 over-expression, and a ratio lower than 1.8 indicates a normal status. Furthermore, when a single cell is characterized by more than six CEP17 probes, this is indicative of polysomy, which is one kind of over-expression (Moelans et al., 2010). Fig. 3(a) shows the negative control from MDA-MB-468 which includes around two HER2 spots and two CEP17 in nucleus. Therefore, the ratio (HER2/CEP17) is almost 1, which indicates no HER2 over-expression. Fig. 3(b) is the negative control from MCF-7. This cell line is a border between over-expression and no overexpression. Thus, one nucleus has more than two but less than six CEP17 probes. With these results, MCF-7 is indicated no HER2 over-expression. Fig. 3(c) shows that a single cell is characterized by more than six CEP17 gene. It is indicated to be polysomy. Fig. 3 (d) is a typical status of HER2 over-expression. HER2 has too many copies over CEP17, leading a ratio more than 2.2. This condition is indicative of positive controls. In this study, quantitative fluorescence data were compared between the traditional benchtop protocols and the one obtained from the developed microchip (Table 2(a)). The average ratios by using these two methods for the negative control cell culture MDA-MB-468 were measured to be 1.06 and 0.95, respectively, for the benchtop and microchip; 0.66 and 0.56, respectively, for the negative control cell line MCF-7, and 2.47 and 2.50, respectively, for the positive control cell line MDA-MB-361. However, it should be mentioned that the positive control cell culture SK-BR-3 appeared to be polysomous with respect to CEP17 probe binding. Detection of HER2 gene amplification could also be successfully observed in fluorescent images of tissue samples (Figs. 3(e)–(f)).

3. Results and discussion 3.1. Characterization of the micropump In order to automate the entire FISH protocol, a suction-type pneumatic micropump acting in concert with microvalves played a key role in the microfluidic chip designed herein. The suction-type micropump had several noteworthy merits, such as the capability to control the delivery of a diverse array of samples/reagents in a precise and rapid manner, as well as the ability to prevent contamination by maintaining high pumping rates (Yang et al., 2009). The deflection of the PDMS membrane below the micropump cavity was created by a negative gauge pressure in the air chamber, which provided a driving force to make the fluid flow through the microchannel. Therefore, the performance of the micropump was dependent upon the volume transported and the back pressure. The pumping rate against the applied air pressure is shown in Fig. 2. The maxmum pumping rate of the micropump integrated on this chip was meassured to be 35 μl/s at a gauge pressure of 50 kPa and a duration of 12 s/cycle. As expected, the pumping rate increased with the appliede gauge pressure. The higher the air gauge pressure was applied, the bigger volume of the PDMS membrane was created. However, at high gauge pressures such as 50,  60 and 70 kPa, the pumping rate reacheed a saturated value and had no significant difference since the membrane was deflected maximally to touch down the channel. Furthermore, different operating frequencies provided different durations to the cycle of one pumping motion. Therefore, the more duration was, the more sufficient time was offered in membrane upward deflection. Note that no bonding between the PDMS layers and the glass slide was performed such that one could detach the chips afterwards for optical scannig of FISH chips. A customer-made clamp was adopted in this work to seal the chambers and channels.

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

The average ratios (n ¼3) of the negative clinical tissues by using the benchtop method and the microchip were measured to be 0.97 and 1.00, respectively, with the negative control tissue S2013, indicating no HER2 over-expression. The ratio of the positive clinical tissues with the benchtop method and the microchip were measured to be 2.28 and 2.35, respectively, with the positive control tissue S2005, which indicated HER2 over-expression. Notably, the entire process could be completed automatically within 20 h, and the reagent consumption was reduced by ∼70% with the microchip. FISH diagnosis could be performed in cell culture as reported in the previous work (Vedarethinam et al., 2010). Nevertheless, their approach needed more reagent consumption, was performed without automation, and did not use clinical samples. Our results reveal that the developed FISH chip is capable of performing the entire protocols by using less reagents (∼70% reduction as shown in Table 1, especially 2-μl usage for DNA probe) in an automatic format. According to the reagent consumption, the assay costs could be significantly reduced. Moreover, three modules, including a fluid control module, a temperature control module and a hybridization module were integrated on a single chip. Notably, both clinical tissues and cell lines could be applied in this chip. Our previous study also presented a microfluidic FISH chip (Tai et al., 2013). It required the permanent bonding of the PDMS layer and the glass slide. Furthermore, only f suspended cells, not tissue samples were demonstrated. The chip presented in this study required no bonding between the PDMS and the glass slide such that it could be easily adopted for subsequent microscopic observations. Furthermore, a dual temperature control module was adopted such that the reagents may not be damaged during the operation. Additionally, clinical tissue samples were performed successfully on this chip. Data acquired in Table 2(a) and (b) show that HER2 genes detection on the benchtop and the chip-based exhibited comparable performance of targeting HER2 spots (orange) and CEP17 spots (green). Simple student's t-test values shown in these tables are 0.15 for MDA-MB-468, 0.13 for MCF-7, 0.85 for MDA-MB-361 and 0.64 for tissue S2013, respectively. When p-value is more than 0.05, it means the microchip efficacy has no significant difference against the benchtop. We also calculated RSD for each sample using the benchtop protocol and the chip-based protocol. RSD for the benchtop and the chip-based methods are 8.50% and 6.32%, respectively, for MDA-MB-468; 8.46% and 4.09%, respectively, for MCF-7; 9.48% and 5.13%, respectively, for MDA-MB-361 and 4.81% and 11.56%, respectively, for tissue S2013. Experimental data showed that with the aid provided by the dual temperature control module and the microfluidic control module, the microchip exhibited a stable heating process. Furthermore, no leakage and no bubbles were observed to provide a benchtop-like environment within this chip. Therefore, the performance of these two Fig. 3. (a) The negative control sample (breast cancer cell line MDA-MB-468) was characterized by measuring HER2(orange)/CEP17(green) ratios of 1.06 and 0.95 for the (a-1) benchtop and (a-2) chip-based protocols, respectively. (b) The negative control sample (breast cancer cell line MCF-7) was characterized by measuring HER2(orange)/CEP17(green) ratios of 0.66 and 0.56 for the (b-1) benchtop and (b-2) chip-based protocols, respectively. (c) The positive control sample (cell line SK-BR3), which appeared to be polysomous. HER2 and CEP17mRNAsþprobes were orange and green, respectively. (d) The positive control sample (breast cancer cell line MDA-MB-361) was characterized by measuring HER2(orange)/CEP17(green) ratios of 2.47 and 2.50 for the (d-1) benchtop and (d-2) chip-based protocols, respectively. (e) The negative control sample (gastric cancer tissue) was characterized by measuring HER2(orange)/CEP17(green) ratios of 0.97 and 1.00 for the (e-1) benchtop and (e-2) chip-based protocols, respectively. (f) The positive sample (gastric cancer tissue) was diagnosed by measuring HER2(orange)/CEP17(green) ratios of 2.28 and 2.35 for the (f-1) benchtop and (f-2) chip-based protocols, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

277

diagnostic approaches is comparable. This study has therefore presented a new microfluidic FISH system capable of rapidly and automatically detect over-amplification of the breast cancer and

278

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

Table 2 (a) Comparisons of cultured cell FISH data between the benchtop FISH protocol and the microchip-based protocol (n¼ 3). Cell line

Assay

HER2

CEP17

Ratio

p-value

MDA-MB-468

Benchtop Microchip Benchtop Microchip Benchtop Microchip Benchtop Microchip

1147 9 1127 8 1207 12 1107 9 Pa Pa 4777 30 503 7 18

108 79 1187 3 1817 8 198 712 Pa Pa 194 77 202 7 17

1.06 0.95 0.66 0.55 Pa Pa 2.47 2.50

0.15

MCF-7 SK-BR-3 MDA-MB-361

(b) Comparisons of tissue FISH data between a benchtop protocol and microchip-based protocol. Tissue Assay HER2 CEP17 S2013 (n ¼3) S2005 (n¼ 1)

a

Benchtop Microchip Benchtop Microchip

1097 3 1077 7 239 224

1137 3 1097 7 105 95

0.13 Pa 0.85

Ratio

p-value

0.97 0.98 2.28 2.35

0.64

Polysomy.

gastric cancer biomarker HER2 in cell cultures and tissue samples.

4. Conclusions We have presented for the first time that a microfluidic chip capable of conducting the entire FISH process in an automatic format for the detection of HER2 amplification in breast cancer cell lines and gastric clinical tissues within 20 h. Moreover, in addition to this convenient automation, the FISH microfluidic chip completed the whole assay with 70% reduction of reagents (especially 2-μl usage for DNA probes) when compared with the traditional benchtop protocols (10-μl usage for DNA probes). Besides, the assay time was considerably lower than that of the benchtop protocols. Four cell lines and two clinical tissue samples have been used to verify the performance of this developed microchip. Pvalue showed no significant difference between the benchtop protocol and the chip-based protocol. It has a great potential for future pharmacogenetic diagnostics and therapy.

Acknowledgements The author would like to thank the Chang Gung Medical Hospital at Linkou in Taiwan and National Tsing Hua University for their financial support of this project (102N2768E1). The authors also gratefully acknowledge the partial financial support provided to this study by the Ministry of Science and Technology, Taiwan (NSC102-2218-E-007-001) and Hsinchu Science Park, Taiwan (103A03).

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.03.003.

References Bruneau, G., Gross, M.S., Krieger, M., Bernheim, A., Thibault, J., Nguyen, V.C., 1990. Preparation of a human dopa decarboxylase CDNA probe by PCR and its assignment to Chromosome-7. Ann. Génét. 33, 208–213. Byun, I., Yang, J., Park, S., 2008. Fabrication of a new micro bio chip and flow cell cytometry system using Bio-MEMS technology. Microelectron. J. 39, 717–722. Courtois, B., Blanton, R.D., 1999. MEMs-Introduction. IEEE Des. Test Comput. 16, 16–17.

Fox, J.L., Hsu, P.H., Legator, M.S, Morrison, L.E., Seeling, S.A., 1995. Fluorescence in situ hybridization: powerful molecular tool for cancer prognosis. Clin. Chem. 41, 1554–1559. Hsu, J.T., Chen, T.C., Tseng, J.H, Chiu, C.T., Liu, K.H., Yeh, C.N., Hwang, T.L., Jan, Y.Y, Yeh, T.S., 2011. Impact of HER-2 overexpression/amplification on the prognosis of gastric cancer patients undergoing resection: a single-center study of 1036 patients. Oncologist 16, 1706–1713. Huang, C.W., Huang, S.B., Lee, G.B., 2006. Pneumatic micropumps with serially connected actuation chambers. J. Micromech. Microeng. 16, 2265–2272. Hung, L.Y., Wang, C.H., Hsu, K.F., Chou, C.Y., Lee, G.B., 2014. Automatic Selection of high-affinity aptamers specific to different histologically classified ovarian cancer cells by using on-chip Cell-SELEX. Lab Chip 14, 4017–4028. Lambein, K., Praet, M., Forsyth, R., Van den Broecke, R., Braems, G., Matthys, B., Cocquyt, V., Denys, H., Pauwels, P., Libbrecht, L., 2011. Relationship between pathological features, HER2 protein expression and HER2 and CEP17 copy number in breast cancer: biological and methodological considerations. J. Clin. Pathol. 64, 200–207. Langer-Safer, P.R., Levine, M., Ward, D.C., 1982. Immunological method for mapping genes on Drosophila polytene chromosomes. Proc. Natl. Acad. Sci. USA 79, 4381–4385. Leversha, M.A, Han, J., Asgari, Z., Danila, D.C., Lin, O., Espinoza, R.G., Anand, A., Lilja, H., Heller, G., Fleisher, M., Scher, H.L., 2009. Fluorescence in situ hybridization analysis of circulating tumor cells in metastatic prostate cancer. Clin. Cancer Res. 15, 2091–2097. Li, J.F., 2002. Microfabrication technology of three-dimensional microdevices and their MEMS applications. J. Inorg. Mater. 17, 657–664. Lichter, P., Ledbetter, S.A., Ledbetter, D.H., Ward, D.C., 1990. Fluorescence in situ hybridization with Alu and L1 polymerase chain reaction probes for rapid characterization of human chromosomes in hybrid cell lines. Proc. Natl. Acad. Sci. USA 87, 6634–6638. Mayer, J.A., Pham, T., Wong, K.L., Scoggin, J., Sales, D.V., Clarin, T., Pircher, T.J., Mikolajczyk, S.D., Cotter, P.D., Bischoff, F.Z., 2011. FISH-based determination of HER 2 status in circulating tumor cells isolated with the microfluidic CEETM platform. Cancer Genet. 204, 589–595. Mitri, Z., Constantine, T., O'Regan, R., 2012. The HER2 receptor in breast cancer: pathophysiology, clinical use, and new advances in therapy. Chemotherapy Research and Practice, vol. 2012, p. 743193. Moelans, C.B., de Weger, R.A., van Diest, P.J., 2010. Absence of chromosome 17 polysomy in breast cancer: analysis by CEP17 chromogenic in situ hybridization and multiplex ligation-dependent probe amplification. Breast Cancer Res. Treat. 120, 1–7. Nath, J., Johnson, K.L., 2000. A review of fluorescence in situ hybridization (FISH): current status and future prospects. Biotech. Histochem. 75, 54–78. Pinkel, D., Landegent, J., Collins, C., Fuscoe, J., Segraves, R., Lucas, J., Gray, J., 1988. Fluorescence in situ hybridization with human chromosome-specific libraries: detection of trisomy 21 and translocations of chromosome 4. Proc. Natl. Acad. Sci. USA 85, 9138–9142. Tai, C.H., Ho, C.L., Chen, Y.L., Chen, W.L., Lee, G.B., 2013. A novel integrated microfluidic platform to perform fluorescence in situ hybridization for chromosomal analysis. Microfluid. Nanofluid. 15, 745–752. Vanneste, E., Melotte, C., Debrock, S., D'Hooghe, T., Brems, H., Fryns, J.P., Legius, E., Vermeesch, J.R., 2009. Preimplantation genetic diagnosis using fluorescent in situ hybridization for cancer predisposition syndromes caused by microdeletions. Hum. Reprod. 24, 1522–1528. Vedarethinam, I., Shah, P., Dimaki, M., Tumer, Z., Tommerup, N., Svendsen, W.E., 2010. Metaphase FISH on a Chip: miniaturized microfluidic device for fluorescence in situ hybridization. Sensors 10, 9831–9846. Velten, T., Ruf, H.H., Barrow, D., Aspragathos, N., Lazarou, P., Jung, E., Malek, C.K., Richter, M., Kruckow, J., Wackerle, M., 2005. Packaging of Bio-MEMS: strategies, technologies, and applications. IEEE Trans. Adv.d Packag. 28, 533–546.

K.-J. Kao et al. / Biosensors and Bioelectronics 69 (2015) 272–279

Weng, C.H., Lien, K.Y., Yang, S.Y., Lee, G.B., 2011. A suction-type, pneumatic microfluidic device for liquid transport and mixing. Microfluid. Nanofluid. 10, 301–310. Yang, S.Y., Cheng, F.Y., Yeh, C.S., Lee, G.B., 2010. Size-controlled synthesis of gold nanoparticles using a micro-mixing system. Microfluid. Nanofluid. 8, 303–311. Yang, Y.N., Hsiung, S.K., Lee, G.B., 2009. A pneumatic micropump incorporated with a normally closed valve capable of generating a high pumping rate and a high

279

back pressure. Microfluid. Nanofluid. 6, 823–833. Young, S.R., Liu, W.H., Brock, J.K., Tutera, A.M., Smith, S.T., 1995. Studies of Her2/Neu and C-MYC amplification in ovarian-cancer using dual-color fluorescence in situ hybridization (FISH). Am. J. Hum. Genet. 57, 438–439.

A fluorescence in situ hybridization (FISH) microfluidic platform for detection of HER2 amplification in cancer cells.

Over-expression/amplification of human epidermal growth factor receptors 2 (HER2) is a verified therapeutic biomarker for breast and gastric cancers. ...
2MB Sizes 0 Downloads 12 Views