Research highlights

Immune response at the microscale

The details of many physiological processes in the body are poorly understood. This is of particular significance in the treatment of diseases like cancer, where understanding the behavior of the immune response at the onset of cancer could potentially lead to improved and more efficient medical treatments.¹ Microfluidic systems provide an opportunity to control the cellular environment where the chemoattractant concentration, topography, cell type and other cellular factors can be tuned for detailed studies of the immune response.

A microfluidic approach was recently taken by Mattei and colleagues in their efforts to understand the interactions between the immune system and cancer cells. Businaro et al.² utilized a simple microfluidic structure consisting of three wide, parallel fluidic channels that were connected via an array of short and narrow capillary migration channels (Fig. 1A, B). Immune cells extracted from the spleen of mice were loaded into one of the outer fluidic channels, and melanoma cells were loaded into the other, such that the melanoma cells could migrate through the whole microfluidic system depending on the chemical cues they detected. The middle fluidic channel served as a buffer channel, and the entire structure was molded in poly(dimethylsiloxane) (PDMS).

Fig. 1 A photograph (A) and partial sketch (B) of the microfluidic device, showing wide cell culture and capillary migration channels. A composite fluorescence image of melanoma (green) and spleen (red) cells interacting with each other inside the device (C). Figure adapted and reprinted with permission from the Royal Society of Chemistry from Businaro et al.²

The key goal of the study was to investigate the effect of interferon regulatory factor 8 (IRF-8) on the immune response. To this end, in one experiment spleen cells from wild-type mice (WT) were allowed to interact with the cancer cells. In a second experiment the spleen cells came from mice whose immune response was compromised by the lack of IRF-8 (IRF-KO cells). As a control, the three cell populations were also cultured individually. In all cases the researchers observed and quantified the cellular migration through the capillary channels and into other parts of the microfluidic device.

It was found, as expected, that when WT spleen cells were allowed to communicate with the cancer cells they migrated swiftly to the melanoma channel, clustered around and interacted with the cancer cells for several hours (Fig. 1C). The clusters could be observed even after 5 days of culture. In contrast, IRF-KO cells approached the cancer cells individually, if at all, and had brief interactions with them (∼few minutes). It was also noted that when co-cultured with WT spleen cells, the proliferation of melanoma cells was severely hindered, but when co-cultured with IRF-KO cells, the melanoma cells invaded all parts of the microfluidic device. This behavior is akin to metastatic processes. Interestingly, when either immune cell populations were cultured separately, there were few, if any, physiological differences between them. This suggests that the WT spleen cells were, in a sense, activated by the cancer cells in co-culture, which triggered the expression of certain immune markers and the onset of the immune response.

The observations reported in this paper would be difficult to uncover in standard well-plate culture experiments, but became available by the use of a microfluidic platform. The same microfluidic system could be utilized to study the migratory behavior of immune cells in space and time, including IRF-KO cells, in the presence of various drugs expected to strengthen the immune response. Similarly, these type of experiments could also shed light on whether certain cancer treatments affect solely the cancer cells, or work by simultaneously activating the immune cells.

A blood–brain-barrier study on-the-chip

Microfluidic devices have been used as inexpensive and efficient miniaturized platforms for the creation of in vitro disease models. These devices could also be utilized to model several neurodegenerative diseases, which are caused by the disruption of the blood–brain-barrier (BBB). The BBB, consisting of a membrane and an endothelial cell layer, controls the selective transport of molecules from the blood to the brain tissue and vice versa. Creating an in vitro model of the BBB may be useful for understanding the progression of the BBB failure.³

Griep et al.⁴ have proposed a miniaturized microfluidic device for studying the erosion of the barrier function in response to both mechanical (shear stress) and biochemical (tumor necrosis factor alpha (TNF-α)) stimuli. Two channels were molded in separate PDMS slabs, placed on top of each other in an orthogonal manner and were separated by a polycarbonate membrane with submicron size pores (Fig. 2). The three layers were bonded using PDMS glue. Immortalized human brain endothelial cells were then statically cultured for 3 days in the top, collagen-coated channel before dynamic culture conditions were introduced. To create a dynamic culture environment, cell medium was flowed through both microchannels for 18 h, generating shear stress values similar to those observed in the capillaries inside the brain tissue (∼0.1 Pa). Then, the transendothelial electrical resistance (TEER) across the membrane was measured as a means to evaluate the barrier stability using Pt electrodes embedded into the PDMS device.

Fig. 2 A sketch of the top (A) and cross-sectional (B) views of the blood–brain-barrier-on-chip. Sizes not to scale.

The impedance spectra of the cell-loaded BBB chip were measured by connecting the chip to an impedance/gain phase analyzer. A cell monolayer, expressing the tight junction marker Zona occludens, was formed in the top microchannel after 24 h. As expected, no cells were observed in the lower device compartment. The TEER increased until day 2 of culture, indicating further tightening of the cell monolayer, and then remained constant.

BBB functions were modulated by shear stress followed by biochemical stimulation. Application of shear stress caused the TEER to increase threefold within 1 day, which could be interpreted as further tightening and stabilization of the cell layer. Under dynamic conditions, the introduction of as little as 1 ng ml⁻¹ TNF-α, however, had the opposite effect resulting in an order of magnitude decrease in TEER. Interestingly, during static culture, TNF-α only led to a 50% reduction of TEER. This observation could mean that endothelial cells have a strong short-term response to shear stress, but over the longer term (a few days) the cell layer can be eroded by the presence of certain chemicals.

The developed microfluidic BBB chip provides a low-cost and versatile platform for drug discovery and drug screening studies for treating neurodegenerative diseases. It may be an alternative to animal models, which are expensive and difficult to study in real-time. Furthermore, this in vitro method could be a useful addition to the standard techniques for evaluating the metastatic potential of brain tumors. Nonetheless, modifications to the BBB-on-chip will be necessary to mimic more closely the in vivo microenvironment. Such modifications may include co-culture of endothelial cells and astrocytes, which are known to contribute to the stability of the BBB, as well as quantification of permeating molecules on both sides of the barrier.

A microfluidic touch sensor

Several industries, including the automotive industry, have seen a wide adoption of robots. However, the use of robots in hospitals or home settings has been hindered by the limited ability of the robots to demonstrate a “fine touch”. A gentle grasping action is often required while handling animate (person or animal) and inanimate objects (e.g. an egg) without crushing them.⁵ To carry out these gentle movements, pressure or tactile sensors need to be incorporated into the handling units for monitoring the force being exerted on the objects. The incorporated pressure sensors should also be soft and flexible to increase the contact area during grasping and to soften impact during handling. Traditional pressure sensors utilizing solid materials or thin films tend to crack and wear out, which suggests that novel solutions might be necessary.

Santos and colleagues reported the use of conductive fluids instead of solid materials in fabricating flexible pressure sensors. Namely, Wong et al.⁶ developed a capacitive tactile unit containing two orthogonal sets of microfluidic channels, a series of air pockets and an array of 5 × 5 pressure-measurement points (taxel plates). All sensor elements were embedded in a flexible PDMS slab that could be attached to a finger, where the microfluidic channels were located directly above and below the air pockets. The microfluidic channels were filled with Galinstan, a liquid metal alloy originally developed for use in thermometers as a nontoxic substitute to mercury. Each intersection of the conductive channels (with the air pockets in between) formed an individual micro-capacitor and created a single taxel (the sensing element of a tactile sensor). The microchannels served both as conduction paths (similar to metal wires), as well as capacitor plates, such that a normal force applied to the taxels changed the electrical and mechanical characteristics of the sensor. Wires were connected to the ports of all fluidic channels and were fed to an operational amplifier. The capacitance of a taxel was proportional to the applied normal force and was calculated from the output voltage of the amplifier.

The sensor could reliably detect a loading force of up to 2.5 N (~250 g). Dynamic loads could be measured at loading/unloading frequencies between 0.4–4 Hz. The spatial resolution of an individual taxel was 0.5 mm. This was the smallest separation at which a force applied to a single taxel was registered by that particular taxel alone. Furthermore, the pressure sensor was reliably used for multiple loading/unloading cycles without showing any evidence of wear and tear.

In the future, the dynamic range of the sensor could be tuned for different applications by filling the air pockets with materials of different dielectric constants. Similarly, the spatial resolution of the sensor could be increased using softer PDMS layers with an increased mass ratio of PDMS base to curing agents. These modifications would benefit the development of robotic arms capable of handling fragile or tender objects. Other applications may include hospital settings where robots wearing these normal force sensors can be of better assistance to patients.

References

1. K. E. de Visser, A. Eichten and L. M. Coussens, Nat. Rev. Cancer, 2006, 6, 24–37.
2. L. Businaro, A. D. Ninno, G. Schiavoni, V. Lucarini, G. Ciasca, A. Gerardino, F. Belardelli, L. Gabriele and F. Mattei, Lab Chip, 2012, 12 10.1039/C2LC40887B.
3. (a) A. M. Palmer, J. Alzheimers Dis., 2011, 24, 643–656; (b) P. M. Carvey, B. Hendey and A. J. Monahan, J. Neurochem., 2009, 111, 291–314.
4. L. Griep, F. Wolbers, B. de Wagenaar, P. ter Braak, B. Weksler, I. Romero, P. Couraud, I. Vermes, A. van der Meer and A. van den Berg, Biomed. Microdevices, 2012, 1–6,  DOI:10.1007/s10544-012-9699-7.
5. H. Yousef, M. Boukallel and K. Althoefer, Sens. Actuators, A, 2011, 167, 171–187.
6. R. D. Ponce Wong, J. D. Posner and V. J. Santos, Sens. Actuators, A, 2012, 179, 62–69.

---