Biosensor and microfluidic chip based point-of-care (POC) testing offers many advantages over centralized laboratory testing, such as small sample and reagent volumes required, rapid detection and on-site analysis. In Temasek Polytechnic Microelectronics Center, we are focusing on developing biosensor and microfluidic chip based POC systems for immunological detection of diseases, such as Sepsis and Urinary Tract Infection (UTI). The systems integrate several elements: (1) biosensor, (2) microfluidic chip, (3) pumps and valves for fluid delivery, (4) signal detection components, and (5) a computer for controlling the fluidics, logging and data analysis.

A POC system for quantitative analysis of procalcitonin (PCT) was developed based on a plasmonic biosensing chip. The plasmonic biosensing chip with gold nanopillar array structure was fabricated by high fidelity nanoimprinting technology. The chip was integrated into a microfluidic device to channel reagents over the nanopillar array. A robust sandwich bioassay of capture antibody / PCT / detection antibody labelled with quantum dot (QD) was established on the chip. To detect the QD emission from the chip, a CCD-based fluorescence detection system was built, which uses a laser beam as light source and an air-cooled CCD camera as detector. A compact circular multi-purpose holder for receiving the microfluidic device was designed with reduced tubing length and optimized valves and reservoirs configuration. With incorporation of the control cum switching software program, this POC system is user friendly and provides a desirable on-site yet sensitive detection of PCT at 1ng/ml for clinical practice. Another POC system integrated with dielectrophoresis on-chip concentration module was developed for UTI detection. A sandwiched immunoassay was developed on microfabricated interdigitated electrode (IDE) biosensor chip for UTI E.Coli detection. Dielectrophoresis (DEP) technology was successfully used to concentration E.Coli and increased the local concentration of bacteria for more than 1000 folds. The DEP-enhanced microfluidic immunoassay can detect E.Coli at the concentration level of 10⁵ CFU/ml.

Myelin sheaths, formed by oligodendrocytes (OLs), enable rapid signal transfer in central nervous system (CNS) and loss or damage of myelin sheaths results in many neurological disorders such as Alzheimer’s or multiple sclerosis.  However, detailed mechanism of myelination process remains largely unknown, not only because of the complicated reciprocal signaling process but also the lack of in vitro model systems that allow easy experimental manipulations.  We present a 3D microfluidic neural stem/progenitor cell (NSPC) culture platform (i.e., Brain-on-a-Chip) capable of culturing NSPCs in an aggregate form under spatially controlled microenvironment and expect it to be used as an in vitro model system for studying CNS myelinogenesis.  The platform is composed of four culture chambers each containing 10 horseshoe shaped trapping sites.  Finite element method (FEM) computer simulation was used to confirm a uniform media flow within the culture chamber and to optimize the design/dimension of the trapping structures for increased aggregate trapping efficiency.  NSPC aggregates with uniform pre-determined size (150 µm) were prepared by culturing dissociated NSPCs from E16 rats in a micro-well array (depth: 150 µm, diameter: 150 µm) for 3 days.  NSPC aggregates were collected and loaded into the culture platform where they were captured at each trapping sites with trapping efficiency of approximately 85%.  Robust neurite outgrowth and glial migration were observed during the first week of culture.  At DIV 14, the NSPC aggregates were treated with retinoic acid (500 nM) to investigate its effect on myelin formation in vitro.  After two weeks of treatment, increased number of myelinating OLs and myelin segments were found in retinoic acid-treated group as compared to controls.  In summary, we have developed a microfluidic NSPC culture platform that can be exploited as a powerful tool for investigating neural development and myelination in vitro and to test potential drug candidates capable of promoting CNS myelination.

Manufactured nanoparticles (MNPs) are microscopic materials (sub-100 nm) having physical and chemical properties uniquely different from their respective bulk materials. MNPs have found applications in optical, electronic, and biomedical fields and numerous novel products containing MNPs are entering the market each year [1, 2]. However, there are many reports of cyto- or genotoxicity of MNPs [3]. A rapid and reproducible screening tool for analyzing and determining MNP toxicity in vitro in a high throughput manner would be desirable to safeguard against toxic effects to humans.

This paper reports a novel approach of using a simple design of microfluidic device to assess the cyto- and genotoxicity of MNPs in a high throughput manner. The microfluidic toxicological testing platform is based on mammalian cells that are immobilized inside microchannels and subsequently used for toxicological analysis of MNPs. Our group has previously developed a lab-on-a-chip toxicological testing platform that contains different cell types immobilized to the surface using antibody-cell surface antigen recognition [4]. While this lab-on-a-chip platform and other commonly used well-plate culture techniques [5] used for MNP toxicological testing are performed under static conditions, there are reports showing that flow through systems may be significantly more relevant highlighting the advantages of using a microfluidic platform approach [6].

A schematic diagram of the crossed flow design of the microfluidic system is shown in Figure 1 where a glass substrate is functionalized with cell adhesion molecules to bind to various types of cells, and parallel streams of the cell suspensions and nanoparticle suspensions are perfused into the chip sequentially.  As a proof of concept, a polydimethylsiloxane (PDMS)-based microfluidic chip containing 5 inlets for cell suspension and 5 inlets for nanoparticle solutions is designed, fabricated, and characterized (Fig. 2).

Collagen 1 or antibodies against cell surface antigens were first immobilized on epoxy-silane functionalized glass substrates using a microarrayer system prior to the perfusion of cells (Fig. 3). An array of 25 spots (diameters of 350 µm) of cell clusters were thus generated. Using two cell dyes as model treatments, we have demonstrated to be able to deliver up to 5 conditions per treatment simultaneously to the immobilized cells array (Fig. 4). Furthermore, we have used this platform to assess cell viability when subjected to simultaneous treatments with solutions of different tonicity (hence toxicity). Finally, we have shown to perfuse solutions of microparticles with various surface modifications (decorated with different antibody concentrations) to the cell microarray (Fig. 5). We have chosen microparticles as proof of concept working towards using nanoparticles in the future.

We are now scaling up the number of microchannels for immobilization of larger number of cell populations and for screening of large number of nanoparticles. This cross flow microfluidic configuration platform can provide a faster readout, is cost effective, offers time-resolved monitoring, and retains high magnification imaging capability. It could potentially lead to the reduction of animal model experiments and achieve a new testing standard for nanoparticle toxicity.

This study integrated the microfluidic system and ODEP technology for the isolation of CTC clusters from the background leukocytes. The working principle is based on the size difference between the CTC clusters and leukocytes, and thus different magnitude of ODEP force acting on them. ODEP mainly use a controllable light pattern, acting as a virtual electrode, to generate a non-uniform electric field that is in turn utilized to manipulate the electrically-polarized cells. The utilization of ODEP-based mechanism for CTCs isolation has been successfully demonstrated in our previous study [1].

Since 1970, a series of clinical studies have shown that single CTCs may not be the main cause of cancer metastasis, but two or more aggregated CTCs [2, 3]. In order to isolate CTC clusters for back-end analysis. For the biological-based methods (e.g. ^HBCTC-Chip [4], or CTC-iChip [5]), although CTC clusters can be specifically separated by antibody-based schemes, but the surface-area-to-volume ratio of CTC clusters is relatively low which might affect the binding efficiency of CTC clusters and antibody. Alternatively, some studies proposed physical-based methods to separate CTC clusters (e.g. ISET [6], FMSA [7], Cluster-chip [8]). Although these methods have been demonstrated to effectively isolate the CTC clusters from the background cells mainly based on their size difference, the influence of shear stress on physical size of CTC cluster, or the viability of the cells isolated is still a problem.

To address this issue, The key advantages of ODEP mechanism for cell isolation including: (1) no need of complex microfabrication process for constructing microfilter structures, and (2) the reduction of shear stress acting on the cells manipulated.

However, the feasibility of using ODEP-based mechanism for the isolation of CTC cluster (i.e. CTC cell aggregates) has not yet been explored. To test its feasibility, a T-shaped microfluidic chip was designed (Fig.1). A virtual microfilter consisting of multiple light patterns was designed at the CTC clusters isolation zone (Fig.1). By continuously moving and rotating the light patterns in the microfilter, the larger CTC clusters can be separated from the background leukocytes, and also transported to the side microchannel (Fig.2). In this work, the optimum ODEP operating conditions (e.g. moving velocity of light pattern) was explored. Results revealed that moving velocity of light pattern that can manipulate the CTC clusters (containing 2-13 cells) was significantly higher than the background of leukocytes (Fig.3). Based on this, the moving velocity of light pattern was set at 100 μm/sec (Fig.3). At a given sample flow rate of 0.5 μl/min, moreover, we found that the rotation speed of light patterns at 14 RPM could significantly increase the purity of CTC clusters isolation (Fig.4). Based on the set operating conditions, the recovery and purity of the isolated CTC clusters were experimentally evaluated to be 70.1 ± 7.1% and 60.8 ± 2.7%, respectively (Fig.4). As a whole, we have established a high purity CTC clusters isolation method that is easy to operate, and is possible to avoid the problem caused by the shear stress acting on the cells or particles.

This paper reports a novel microfluidic viscometer with an embedded pressure sensor constructed using electrofluidic circuits, which are built by filling ionic liquid into microfluidic channels. The developed sensor-integrated microfluidic viscometer is made of polydimethylsiloxane (PDMS) with transparent electrofluidic circuits, which makes it feasible to real-time monitor samples under tests.  In addition, the entire device is fully disposable to prevent cross contamination between samples, which is desired for biomedical applications.  The electrofluidic circuit pressure sensor also provides great sensing linearity, long-term and thermal stability for reliable measurements. Various microfluidic viscometer designs have been developed in previous studies. However, several drawbacks, including: limitations for Newtonian fluids, limited shear rate ranges, temperature sensitivity, incapability for real-time monitoring, and complicated setup [1-5], make the devices impractical for routine measurements.  In this paper, we develop a fully disposable and optically transparent PDMS microfluidic viscometer for biomedical applications. The device can be used for Newtonian and non-Newtonian fluid measurement under different temperatures and shear rates.  The microfluidic viscometer consists of a glass substrate, a PDMS-made bottom microfluidic layer and top electrofluidic channel layer. A PDMS membrane is sandwiched between the two layers (Figure 1). In the bottom microfluidic layer, the hydrostatic pressure at the upstream is built up during the sample injection due to viscous force. By measuring the hydrostatic pressure at the upstream with known geometry of the channel, sample viscosity can be estimated. To measure the hydrostatic pressure, an integrated electrofluidic pressure sensor is designed in the top layer that provides great long-term and temperature stability [6]. A pressure transduction hole is fabricated near the inlet of the microfluidic channel, which is exploited to transfer hydrostatic pressure from bottom layer to top layer. Once the hydrostatic pressure deforms the membrane through the pressure transduction hole, electrical resistance of an electrofluidic resistor aligned to the hole will be changed due to its cross-sectional area variation.  To precisely measure the resistance change, Wheatstone bridge circuit architecture is exploited. The Wheatstone bridge circuit also provides the sensor great sensing linearity and temperature stability. Consequently, the hydrostatic pressure can be estimated by measuring the electrical voltage across the circuit, and the viscosity can be calculated based on fluid mechanics theories.  Four types of fluids are analyzed in the experiments. Water and glycerol solutions of different concentrations are exploited to demonstrate the viscosity measurement of Newtonian fluids, and the results are shown in Figure 3. To demonstrate shear thinning property of non-Newtonian fluids, 1000 ppm xanthan gum solution is tested (Figure 4). In order to further demonstrate usage of the device for real biological samples, human whole blood with EDTA is tested under different temperatures (Figure 5).  The results demonstrate the advantages of small volume (< 1 ml), disposability, and different temperature operation capability of the developed microfluidic viscometer for practical biomedical applications. We have successfully developed a fully disposable and optically transparent microfluidic viscometer based on ionic liquid electrofluidic circuit. The device can be applied widely in applications where long-term observations and various temperature operations are required.

We developed here a fluidic chip for co-culture of adipocytes and immune cells to mimic in vivo environment of diabetes. Type II diabetes is commonly developed as a result of obesity. Evidence suggest that there is excess infiltration and accumulation of immune cells into obese adipose tissue [1-3]. These immune cells release different types of inflammatory cytokine that causes adipose tissue to become chronically inflame which subsequently reduce adipose tissue sensitivity toward insulin as observed in diabetes. Understanding interaction between immune cells and adipocyte as well as pattern of cytokine profile of the adipo-immuno system could lead to insight mechanisms of early diabetes development and possibly early markers of diabetes.

Study of adipo-immuno system in human obesity and diabetes is not well established due to small amount of sample obtained from patients. Current in vitro cell culture often requires large amount of sample for repetitive cell culture and activation making the in vitro method not suitable for human sample study. Alternative system which is able to handle and efficiently utilize small sample is demanded. We develop here a miniaturized adipo-immuno co-culture system to mimic diabetes environment. Adipocyte and immune cells are cultured in separate but fluidically connected compartments with constant supply of nutrient via perfusion (Figure 1). The co-culture cells also fluidically linked to an immunoassay chamber that allows cytokine to be monitored without interfered the co-culture environment.

We first tested fluidic diffusion on chip using colored dye to ensure crosstalk between cells in different compartments (Figure 2) [4]. Once fluidic crosstalk is confirmed, cells were cultured on chip. For co-culture cell model, differentiated human pre-adipocytes were used as adipocytic model whereas peripheral blood mononuclear cell (PBMC) were used as immune cell model. Data indicate that adipocytes can be grown, differentiated and maintained on chip for up to 20 days whereas PBMC could maintain >80% viability on-chip for up to 2 days (Figure 3). PBMC viability on-chip is consistent with PBMC cultured on standard cell cultured plate indicating that the short PBMC lifetime is not due to the on-chip cultured condition but it is the nature of PBMC that could not survive for too long once isolated. Next, glucose uptake experiments were conducted to observe the co-culture system biological activity. LPA was used to induce system inflammation. Result show that adiopo-immuno co-culture with LPA treatment has a tendency to take up more glucose as compared to other groups. Nevertheless no conclusive data can be yet interpreted. More data on insulin treatment and cytokine profile of different treatment conditions are needed.

In conclusion, we have demonstrated a fluidic chip that could be used for adipo-immuno co-culture. Nevertheless, functional study of the chip is still at the early stage and more biological data is needed for study of adipo-immuno interaction.

Acknowledgment

The authors would like the Blood Bank of Health Sciences Authorities, Singapore for kindly providing blood from healthy volunteers for this study. We also thank A*STAR, Singapore for providing support for this project (JCO Grant#1431AFG123).

Single-cell analysis of cancer cells is important to improve our understanding of tumor heterogeneity, which could contribute to various clinical applications, such as diagnosis, prognosis, and choice of therapy. In order to improve single-cell analysis throughput, cell-sized microwells in an array are often fabricated and single cells are trapped via various methods such as dielectrophoresis, magnetic and optical techniques. After trapped individually, the cells are either lysed to get intracellular materials or directly analyzed to investigate various intracellular responses to external stimuli. However, besides confinement of each cell in an individual microwell, a real single-cell analysis needs to give individual stimuli to each cell, which prevents the cross-contamination of the intracellular materials among each cell. In this study, we firstly trapped individual cells in cell-sized microwells and obtained a high trapping yield via a dielectrophoresis electrode array. Moreover, an analysis microwell array with photolithographically patterned microgaskets was used for high throughput single cell analysis. The droplets containing stimuli were accurately injected into individual analysis microwells by using an inkjet spotter and the analysis microwells with stimuli were reversibly sealled for liquid tight and were aligned to the trapped cells for single cell analysis. For example, we demonstrated the utility of the developed platform by monitoring the kinetics of calcium fluxes and oscillations in individual Hela cells under the stimulation of drugs. As a universal second messenger in virtually all eukaryotic cells, calcium ions (Ca²⁺) plays a vital role in cellular responses to stimuli and mediates various physiological processes, including neurosecretion, skeletal muscle contraction, cell growth and differentiation. The resulted calcium imaging data revealed significant difference between different cells under the stimulation of drugs. We believe that such a technology will be an essential tool for obtaining high throughput single cell analysis data, in order to indentify statistically significant trends and achieve significant improvements in reagent consumption and analysis time.

Herein, we introduce a novel Dean migration phenomenon in spiral microchannel, termed High-resolution Dean Flow Fractionation (HiDFF), for separation of small microparticles based on differential Dean migration profiles. Inertial microfluidics is an emerging class of passive size-based sorting technique for cell separation [1,2]. However, inertial focusing of small microparticles (particle size, a_p < 1-2 µm) remains a huge technical challenge as channel dimensions have to be scaled down (hydraulic diameter, D_h ~10 µm) for them to experience significant inertial forces (F_L) and undergo lateral migration. Our group previously developed a spiral microfluidics sorter (Dean Flow Fractionation, DFF) which enables well-controlled, Dean-induced lateral migration of small particles including bacteria [3], nanoparticles [4] and biomolecules [5] to channel outer wall while target cells (>10 µm) focus near inner wall for separation. However, it cannot further size-fractionate small particles (1 µm vs. 2 µm) as they would recirculate continuously due to Dean vortices. Using HiDFF, we were able to demonstrate efficient separation of 1–3 µm microbeads mixture, and then apply the technology to size-fractionate different small biological components including bacteria and nanoparticles.

The 2-inlet, 2-outlet spiral microdevice (60×300 µm (H×W)) is fabricated using polydimethylsiloxane (PDMS) (Fig. 1A). Small particles (a_p/ D_h <0.07) introduced at channel outer wall experience Dean drag forces (F_D) due to Dean vortices and migrate laterally towards inner wall (Fig. 1B). Near the inner wall, particles occupy different innermost transient positions under influence of size-dependent wall-induced inertial lift force (F_WL) (Fig. 1C). Hence, smaller particles are positioned closer to inner wall and separated into the inner outlet (outlet 1) while larger particles are sorted into outer outlet (outlet 2).

We first used 50 nm fluorescent polystyrene microbeads to study the lateral migration of small particles at different channel positions. Particles migrated along the channel top and bottom towards inner wall, and recirculated outwards as a tight band along the channel midline (Fig. 2A). Interestingly, the innermost transient particle position moved further away from inner wall with increasing particle size at same flow conditions, which led to distinct transient positions for each particle size (50 nm–3 µm) (Fig. 2B). We next characterized the separation efficiencies of different binary bead mixtures including 2 and 3 µm, and 1 and 2 µm beads (Fig. 3). Approximately 20% of the smaller particles were sorted into outlet 1 which resulted in a ~10 to 30–fold enrichment. Lastly, we used the developed HiDFF technology to size-fractionate Escherichia coli (E. coli) and Poly(lactic-co-glycolic acid) nanoparticles (PLGA NPs) into 2 different size groups (Fig. 4), and verified the particle size distribution using optical imaging and scanning electron microscope (SEM).

Currently, there is a critical need for small particle sorting which is important in applications ranging from material sciences, environmental bio-sampling to bacterial/exosomes clinical diagnostics. The developed HiDFF technology represents an important progress towards this goal, as it enables high throughput (~100 µLmin^-1) membrane-free separation of small particles (<1 µm), and can be easily integrated to existing point-of-care platforms for small biological targets purification and detection.

Technologies for mapping dynamic patterns of neural activity over time and space have advanced our understanding of brain function in both health and disease. An important application of these technologies is the discovery of next-generation neurotherapeutics for neurological and psychiatric disorders lacking effective treatments. Here, we describe an in vivo drug screening strategy that combines high-throughput brain activity mapping (BAM) technology with bioinformatic analysis. This platform enables evaluation of a compound’s therapeutic potential based on information rich BAMs derived from drug-treated zebrafish larvae. From a screen of clinically used drugs, we found intrinsically coherent drug clusters that are significantly associated with known therapeutic categories. Using the BAM-based clusters as a functional classifier, we successfully predicted anti-epileptic candidate drugs from non-clinical compounds and implicate epigenetic mechanisms for the development of novel anti-epileptic drugs. Collectively, these BAMs linked to specific compounds provide an experimental framework to advance the field of systems neuropharmacology.

Organotypic culture of human tissue has a great potential to bridge the existing gap between in vivo and in vitro studies and to enable in-depth study of disease pathogenesis and therapy with low cost. Recent advances in microfluidic-based co-cell culture have enabled the realization of physiologically relevant in vitro models of specific human tissues. These systems have been used to study the contribution of micro-environmental factors to cellular morphogenesis in health and disease states,   therefore, can be used to answer fundamental biological questions and develop a reliable alternative to animal models, and enable screening of potential new drug compounds. In this paper, we will discuss three immune competent micro-physiological models, namely the human gastrointestinal tract¹, the human epidermis² and the adipose tissue³, the interaction of immune cells with these tissues and their role in inflammation. We will emphasis on the interaction between the adipocytes and immune cells to highlight the role of immune cell infiltration in the adipose tissue in the pathogenesis of diabetes type 2. A three-dimensional perfusion-based microfluidic system has been realized and utilized to host the three biological models. It enables co-culturing of two different types of cells that are physically separated but fluidically and chemically connected which enables the exchange of paracrine signals between the two cell types. The co-culture was maintained viable for more than three weeks during which cells were differentiated such that their phenotypes, morphologically and functionally, resemble the corresponding cells in the human tissue. Inflammation was induced in the biological system, and the response of the co-culture was detected. The biological models have been characterized in both health and disease states. Various commercial drugs are being tested to evaluate the in vitro system response.

References

1. Qasem Ramadan and Lin Jing. Characterization of tight junction disruption and immune response modulation in a miniaturized Caco-2/U937 coculture-based in vitro model of the human intestinal barrier, Biomedical Microdevices, 2016,18(1):11.
2. Qasem Ramadan and Fiona Chia Wan Ting. In vitro micro-physiological immune-competent model of human skin. Lab Chip, 2016,16, 1899-1908.
3. Yunxiao Liu, Patthara Kongsuphol, Sajay Bhuvanendran Nair Gourikutty, Qasem Ramadan. Human adipocyte differentiation and characterization in a perfusion-based cell culture device. Biomedical microdevices, Accepted.

Acknowledgment

This research was supported by the Agency for Science, Technology and Research (A*STAR) (Grant number: 1431AFG123).