**4.1.1 Cells adhesion**

24 Advances in Unconventional Lithography

PLL could not be guarantee to coat the microchannels effectively after days. There is still a lot of work to do to construct the neuronal networks on MEA to study the cells function as

(a) (b)

neurons in the striatum cultured at 3 days (b) PC 12 cells cultured at 6 days

**4. Construction of neural network by applying soft lithography** 

Fig. 8. Microscopy of the cells cultured on MEA with PI microchannels (a) GABAergic

A man-made neural network on electrode can be applied to do basic research of neuroscience, be a kind of biosensor for drug discovery [9] and even be implanted into brain to establish artificial connections that could form the basis of a neural prosthesis[10]. These fields have caused much attention in the world [11]. In the past, there are three strategies to realize a simplified neural network in vitro, such as mechanical fixation-applying spatial restrictions[12,13], physical modification of surface roughness and surface topography[14,15], chemical polymer microfabrication on surface—using soft lithography. Soft lithography is created by Whitesides in 1993[16]. It is used to create chemical structures on surfaces, including µCP, µFN and other downstream techniques. These microfabrication techniques that control both the size and shape of the cell anchored to a particular surface are extremely useful for understanding the inuence of the cell–material interface on the behavior of cells [17, 18].The adhesion and survival of neural cells should be considered firstly for the patterned neural cell culture in vitro. So the selection of appropriate cellattracting substances is an important step for pattern design in micro-contact printing. Ideal substances encourage good cell–substrate interactions, constantly stimulate the cells by substrate-bound chemical, biological, electrical or mechanical signals [19, 20] and even regulate neuritis growth on designed patterns. The most commonly used coating reagents to promote cell adhesion are extracellular matrix (ECM) proteins like laminin (LN), positively charged polymers such as poly-l-lysine (PLL) and synthetic amide-linkage-free compounds such as polyethylenimine (PEI). Therefore, the characteristics of three different substrates, PEI, PLL, LN were compared by the primary neuron culture in our previous work. The PEI characterized with strong positive surface charges was validated to fabricate more continuous and integrated micro-contact printing neural patterns under serum-free culture

in vitro.

conditions than PLL, LN[21].

The number of cells that adhere to the different substrates after 24 h in culture is shown in Fig. 9. The adhesive effects were evaluated by calculating the ratio of cell numbers that adhere to the grid-like patterns divided by the total area of printed polymer. We found that the positively charged polymers (PLL and PEI) had a signicantly higher level of cell attachment than LN (p < 0.05)[13].

Fig. 9. The adhesive effects were test by analysis of the number of neurons on the area (mm2) of LN, PEI and PLL grid patterns after 24 h in culture. The asterisks indicated neurons on PEI and PLL patterns had signicantly higher lever than on LN patterns, n = 12, p < 0.05.

Application of Soft Lithography and Micro-Fabrication on Neurobiology 27

Parkinson's disease is due to the loss and injuries of dopaminergic neuron in substantia nigra which cause a decrease in nerve fibers projected to the new striatum. Finally, reasonable synaptic connections and neural network can not be established. Therefore, it is expected to establish cell models to investigate the relationship between these two kinds of coherent neurons and construct an artificial neural network in vitro by the application of soft lithography. In present work, specific neural network with dopaminergic neurons and GABAergic neurons co-culture was established by μCP PEI grid patterns. Meanwhile, PEI was validated again to fabricate more continuous and integrated neural patterns by using

Different kinds of neural network by μCP were established with neuron from the striatum, dopaminergic neurons from the substantial nigra and both of them co-culture. The conditions of neuronal adhesion on different pattern figures were observed using several techniques, including immunocytochemical staining, transmission electron microscope and scanning electron microscope. Using immunocytochemical staining, transmission electron microscope, we identified the types of neural cells and observed some neurosynapse-like structures near the neuronal soma on PEI-coated coverslips. These ndings indicate that PEI is a suitable surface for establishing a functional neuronal

**4.2.1.1 Investigation of neural cell types and neurite elongation along the grid-like** 

PEI, PLL and LN were used to produce grid-shape patterns on glass coverslips by microcontact printing. GABAergic neurons and medium spiny neuron from the rat striatum, dopaminergic neurons from the rat substantial nigra and both of them co-culture were researched separately on the different polymers coated surface. The viability and morphology of these neurons under serum-free culture conditions were observed using uorescent microscopy in Fig. 11, Fig. 13, Fig. 14. After 7 days in culture, we found that the neural cell bodies on the PEI patterns were located mostly at the cross-points of the grid, whereas neurites extended along the line of the grid-like patterns. More continuous and integrated neural network was achieved finally. On the PLL-coated coverslips, the neural patterns appeared to be integrated. But several cells clustered at the cross-points of the grid disappeared gradually after the media was replaced. In contrast, cells cultured on the LN-coated grids were generally clustered into clumps and cannot form satisfied patterns. Different sizes of PEI pattern were produced by microcontact printing. In Fig. 12, compared with 50μm, 100μm, 200μm pattern sizes, we found few difference early. After 7 days or 14 days culture, most neural cells on 200μm size grew well and were seldom found to overlap each other, unlike those on 50μm size clustered into clumps at the cross-points of grid and disappeared gradually. Identified with immunocytochemical staining, we found that neural cells from the rat substantial nigra were TH positive, synaptic vesicle protein were synaptophysin positive in Fig. 13 and cells from the rat striatum were GABA positive or DARPP-32 positive in Fig. 11, Fig. 12, Fig. 14. SEM show neurons outgrowth on PEI-coated patterns and validate the findings by

μCP and microfluidic technique both.

**patterns by immunocytochemical staining and SEM** 

immunocytochemical staining in Fig. 15, Fig. 16.

**4.2.1 Neural network by μCP** 

network in vitro.
