Whereas the axons of the peripheral nervous system (PNS) spontaneously regenerate after an injury, the occurring regeneration is rarely successful because axons are usually directed by inappropriate cues. Therefore, finding successful ways to guide neurite outgrowth, in vitro, is essential for neurogenesis. Microfluidic systems reflect more appropriately the in vivo environment of cells in tissues such as the normal fluid flow within the body, consistent nutrient delivery, effective waste removal, and mechanical stimulation due to fluid shear forces. At the same time, it has been well reported that topography affects neuronal outgrowth, orientation, and differentiation.
1. Introduction
Taking into consideration the fact that neurological injuries are hardly self-recoverable, the development of successful methods to guide neurite outgrowth,
in vitro, is of high significance. Indeed, although the peripheral nervous system (PNS) exhibits a higher rate of regeneration than that of the central nervous system (CNS) through spontaneous regeneration after injury, functional recovery is fairly infrequent and misdirected
[1].
Neural tissue engineering has emerged as a promising alternative field for the development of new nerve graft substitutes to overcome the limitations of the current grafts
[2]. Indeed, the ultimate goal of a tissue-engineered construct is to sufficiently mimic the topographic features of the extracellular matrix and the surrounding environment of cells (e.g., mechanical properties, soluble factors, shear stress) so that cells will respond in their artificial environment in the same way they would
in vivo.
Several approaches have been developed to create cellular substrates at the micro- and/or nano-scales that aim to reconstruct the architecture of the extracellular matrix
in vitro [3][4][5][6]. However, apart from topography, it has recently become increasingly evident that neurogenesis may be also driven by mechanical factors
[7][8]. Indeed, shear stress is a significant component of the host environment of regenerating axons
[9]. Flow-induced shear stress can be applied to cells,
in vitro, using specially designed microfluidic systems. In such systems, microfluidic flow replicates the physiological fluid flow inside the body, facilitates mass transport of solutes, and supplies consistent nutrient delivery and effective waste removal resulting in a more
in vivo-like environment
[10]. Previous research and review studies have described the main fabrication techniques, the produced types of microfluidic systems, as well as their respective advantages and disadvantages
[10][11][12][13][14][15][16].
2. Application of Microfluidic Systems for Neural Studies
The positive effect of dynamic cultures on the response of neuronal and/or glial cells (such as adhesion, proliferation, directionality, and differentiation) has been demonstrated by many studies. Studies on the (i) gradient and chemical stimulation of neural cells, (ii) perfusion and shear stress on neural cells, and (iii) co-culture of neural cells via microfluidic systems were reported below.
2.1. Gradient Generation—Chemical Stimulation
Chung et al.
[17] developed a polydimethylsiloxane (PDMS) gradient-generating microfluidic platform that exposed human neural stem cells (hNSCs) to a concentration gradient of known growth factors (GFs), under continuous flow (5 × 10
−5 Pa). As a result, a minimization of autocrine and paracrine signaling was observed. Additionally, they found that the differentiation of hNSCs into astrocytes was inversely proportional to GF concentration while proliferation was directly proportional. Kim et al.
[18] performed a similar study by using a microfluidic chip-generated growth factor gradient system and neural stem cells (NSCs). Results showed that NSCs proliferation and differentiation were directly dependent on the concentration gradient of GF. Additionally, Nakashima and Yasuda
[19] fabricated a microfluidic device to investigate the effect of GF on the differentiation and axon elongation guidance of adrenal pheochromocytoma (PC12) cells. The microfluidic device was composed of a cell culture chamber, a micro-channel, a nano-hole array (containing GF), and a micro-valve allowing the precise release of chemicals from the nano-hole. Nerve growth factor (NGF) was used to stimulate the differentiation of PC12 cells. They showed that the cell growth, differentiation, and axon elongation were dependent on micro-valve switching and release gradient of NGF. Futai et al.
[20] developed a one-layer H-shaped microfluidic channel, for concentration gradient generation, consisting of a thin (~2 μm) but high-aspect-ratio (0.5–1) microchannel. Using a long-lasting concentration gradient of NGF, they examined the axon elongation of primary dorsal root ganglion (DRG) neuronal cells. The results revealed a directional elongation of axons following the NGF concentration gradient during cultivation for 96 h. Furthermore, Park et al.
[21] designed a microfluidic platform to expose human embryonic stem cells (ESCs)-derived neural progenitor cells to stable concentration gradients of extracellular signaling molecules. Human ESC-derived neural progenitor cells were cultured in the microfluidic system under continuous cytokine gradients (0.15 μL/h) for 8 days. Neural progenitor cells proliferated and differentiated, into neurons, in a controlled manner, and the cell properties reflected the different concentrations of extracellular signaling molecules. Bhattacharjee and Folch
[22] fabricated a microfluidic chip containing 1024 biochemical gradient generators, with each generator entrapping a single neuron, to investigate the axon guidance and growth dynamics of primary hippocampal neurons in response to biochemical cues. The microfluidic chip produced a reproducible, stable gradient with negligible shear stress on the culture surface (100 μL/h). Using this platform, it was demonstrated that the hippocampal axon guidance was dependent on the concentration and incidence angle of the netrin-1 gradient. In particular, regarding the concentration, it was found that hippocampal growth cones close to the source of netrin-1 (high concentration) were strongly attracted, while those far from the source of netrin-1 (low concentration) were repelled. With regards to the angle of incidence, it was shown that hippocampal growth cones oriented away from the gradient axis (90–135°) turned toward the netrin-1 source, while those oriented toward the gradient (less than 45°) were strongly repelled. Finally, Cheng et al.
[23] developed a microfluidic system that could analyze the effect of chemical and mechanical stimulation on the neuronal differentiation of placenta-derived multipotent stem cells (PDMCs). They investigated the effect of shear stress using different flow rates (1.4 × 10
−4, 3.31 × 10
−3, 4.97 × 10
−3 Pa) on PDMCs beside 3-isobutyl-1-methylxanthine (IBMX), as the chemical stimulant, for 3 days. They showed that shear stress could not differentiate PDMCs into other cell types. Although chemical stimulation played a crucial role in the differentiation of PDMCs, shear stress enhanced PDMCs for earlier neuronal differentiation. In the 48-h condition, the maximum flow rate and IBMX showed the highest cell differentiation ratio of 42.4%.
The neuronal response to gradient and chemical stimulation through microfluidic systems is summarized in Table 1. The different approaches and the respective nerve responses are presented.
2.2. Perfusion—Shear Stress
Chafik et al.
[9] developed a custom-designed flow chamber that applied shear stress (1.33 Pa for 2 h), through laminar fluid flow, to Schwann cells. They showed that mechanical stimuli enhanced the proliferation of Schwann cells and caused a slight movement from their original positions. Gupta et al.
[24][25] used an
in vitro model to apply shear stress on primary Schwann cells in the form of laminar fluid flow (3.1 Pa for 2 h). They observed increased proliferation and downregulation of two pro-myelinating proteins, myelin-associated glycoprotein (MAG) and myelin basic protein (MBP). These results implied that a low level of mechanical stimulus may directly trigger Schwann cell proliferation. Moreover, Millet et al.
[26], considering that cell-to-cell signaling is local, developed a specific culture system that sustained small numbers of primary hippocampal neurons and enabled analysis of the microenvironment. They observed that cultured neurons inside perfused channels (by gravity flow) composed of native, autoclave, or extracted PDMS showed increased viability and channel-length capacity (increasing 2-fold for all but native PDMS). Park et al.
[27] designed a two-dimensional microfluidic system to study the effect of continuous flow shear stress (10
−4 and 10
−3 Pa) on radial glial cells (RGCs). They found that flow shear stress possibly activated mechanosensitive Ca
+2 channels that significantly enhanced the proliferative capacity of RGCs in response to increased shear stress. Furthermore, Park et al.
[28] developed a microfluidic chip to apply a continuous flow of fluid that is readily observed in the brain’s interstitial space, on three-dimensional (3D) micro-spheroidal neural tissue (neurospheroids). The slow interstitial level of flow (0.15 μL/min) was maintained using an osmotic micropump system and the uniform neurospheroids were formed in concave microwell arrays. Using this system, the effect of flow on the size of neurospheroids, neural networks, and neural differentiation was investigated. Under flow conditions, the results indicated that the size of neurospheroids was larger and formed more complex and robust neural networks (increased levels of synapsin IIa and β-ΙΙΙ tubulin, decreased levels of nestin) compared to static conditions. This phenomenon was attributed to the continuous nutrient, cytokine, and oxygen transport, as well as the removal of metabolic wastes provided by the presence of slow interstitial flow. Additionally, this system was utilized to investigate the toxic effects of amyloid-β, which is thought to be the main cause of Alzheimer’s disease. Under flow conditions, the results revealed a decreased viability of neurospheroids, as well as increased neural destruction and synaptic dysfunction compared to static conditions. Conclusively, this system had significant potential as an
in vitro brain model since it offered an environment that was similar to that found
in vivo.
The neuronal response to perfusion and shear stress through microfluidic systems is summarized in Table 2.
2.3. Microfluidic Cell Co-Culture Platforms
Majumdar et al.
[29] designed a PDMS microfluidic cell co-culture platform that allowed individual manipulation of various cell types with the placement of a microfabricated valve that served as a reversible barrier between the chambers. As a result, healthy co-cultures of hippocampal neurons and glia were maintained for several weeks under optimal conditions. In particular, co-culture with glia provided nutrient media for maintaining healthy neural cultures, eliminating the need to supply neurons with pre-conditioned glia media, thereby enhancing the transfection efficiency of neurons in the platform. Similarly, Shi et al.
[30] fabricated two PDMS microfluidic cell culture systems, a vertically-layered set-up, and a four-chamber set-up for studying communication between neurons and glia in close proximity. The chambers were separated by pressure-enabled valve barriers that allowed them to control communication between the two cell types. In this study, the number and stability of synaptic contacts, as well as the secreted levels of soluble factors, were increased in the co-culture system, thus confirming the importance of communication between neurons and glia for the development of stable synapses in microfluidic platforms. Robertson et al.
[31] developed an
in vitro system to examine the synaptic interaction between two interconnected populations of mixed primary hippocampal co-cultures by integrating microfluidics with calcium imaging techniques. Moreover, a computational model was verified to characterize the fluidic characteristics of the system and improve the experimental protocols (prevent substance cross-contamination between co-cultures). The results revealed that neurons and glia, in each of the separated chambers, grew within the microchannels where they physically interacted and formed synapses. In addition to this, the function of the neuron-glia synapse was confirmed by calcium imaging. Yang et al.
[32] fabricated a microfluidic array platform to modulate hNSC differentiation in a 3D ECM microenvironment using recapitulation of paracrine action of genetically engineered human mesenchymal stem cells (hMSCs). hMSCs were genetically engineered to increase the expression of glial cell-derived neurotrophic factor (GDNF) utilizing cationic polymer nanoparticles. A simulation study produced by mathematical modeling of accumulated GDNF secreted from engineered hMSCs confirmed that
in vivo-like signaling of secreted factors could be created in a 3D ECM hydrogel in the microfluidic system. Specifically, in the central channels of the microfluidic system, which were filled with a 3D ECM hydrogel, hNSCs were cultivated, and GDNF-overexpressing hMSCs (GDNF-hMSCs) were cultured in the channels on each side of the central channel. Reduced differentiation of hNSCs into glial cells and increased differentiation of hNSCs into neuronal cells, including dopaminergic neurons, were observed in the co-culture of hNSCs with GDNF-hMSCs in the 3D microfluidic system. Moreover, neuronal cells demonstrated functional neuron-like electrophysiological characteristics. Finally, an animal model of hypoxic-ischemic brain injury was used to confirm the improved paracrine capacity of GDNF-hMSCs. Park et al.
[33] fabricated a circular microfluidic co-culture platform where embryonic CNS neurons and postnatal oligodendrocytes (OLs) were co-cultured in two separated compartments which were connected by arrays of axon-guiding microchannels. These microfluidic channels allowed physical isolation for cell bodies, but not for axons, and maintained fluidic isolation. The embryonic CNS neurons and postnatal OL progenitors were co-cultured in the platform for up to four weeks to study the axon-glia interaction and myelination. The circular design demonstrated excellent cell loading characteristics where a significant number of cells were positioned near the axon-guiding microchannels. Moreover, it showed enhanced axonal growth characterized by the significantly increased axon coverage ratio in the axon-glia compartment. The co-culture capability of the platform was confirmed by successfully co-culturing OL progenitors with axons in the axon/glia compartment resulting in the maturation of OLs. In a related study, Park et al.
[34] developed a multi-compartment microfluidic co-culture platform for studying axon-glia interaction at a higher throughput. The platform allowed for conducting parallel localized biomolecule and drug treatments while carrying out various co-culture conditions in a single device. Using this platform, they were able to simultaneously study the axon-glia communication, the development and differentiation of oligodendrocytes, as well as the axonal-specific response to different stimuli. The results revealed that mature oligodendrocytes were needed in order to obtain a robust myelin sheath instead of oligodendrocyte progenitor cells. Additionally, it was shown that astrocytes stimulated the development of oligodendrocytes and were detrimental when added to an already existing axonal layer. Finally, Ristola et al.
[35] developed a novel PDMS-based microfluidic cell culture device, with open compartments, for neuron-oligodendrocyte
in vitro myelin studies. It was shown that the primary rat DRG neurons were successfully co-cultured with the oligodendrocytes in the device, which could also be used for time-lapse imaging. The results also demonstrated successful interactions and contacts between neurites and oligodendrocytes, as well as the deposition of myelin segments in an aligned distribution in the device.
Table 3 summarizes the co-culture studies of neural cells using microfluidic systems.