Simple and Inexpensive Paper-Based Astrocyte Co-culture to Improve Survival of Low-Density Neuronal Networks

Abstract

Bottom-up neuroscience aims to engineer well-defined networks of neurons to investigate the functions of the brain. By reducing the complexity of the brain to achievable target questions, such in vitro bioassays better control experimental variables and can serve as a versatile tool for fundamental and pharmacological research. Astrocytes are a cell type critical to neuronal function, and the addition of astrocytes to neuron cultures can improve the quality of in vitro assays. Here, we present cellulose as an astrocyte culture substrate. Astrocytes cultured on the cellulose fiber matrix thrived and formed a dense 3D network. We devised a novel co-culture platform by suspending the easy-to-handle astrocytic paper cultures above neuronal networks of low densities typically needed for bottom-up neuroscience. There was significant improvement in neuronal viability after 5 days in vitro at densities ranging from 50,000 cells/cm² down to isolated cells at 1,000 cells/cm². Cultures exhibited spontaneous spiking even at the very low densities, with a significantly greater spike frequency per cell compared to control mono-cultures. Applying the co-culture platform to an engineered network of neurons on a patterned substrate resulted in significantly improved viability and almost doubled the density of live cells. Lastly, the shape of the cellulose substrate can easily be customized to a wide range of culture vessels, making the platform versatile for different applications that will further enable research in bottom-up neuroscience and drug development.

Keywords: astrocyte; cell viability; co-culture; low-density culture; network activity; neurite length; neuron; paper-based.

Figure 1

Astrocytes isolated from primary rat cortex and cultured on paper-based substrates. (A–D) Influence of extracellular matrix protein coatings on astrocyte proliferation and viability at 7 DIV. Astrocytes were seeded at 100,000 cells/cm² on paper coated with (A) laminin, (B) fibronectin, or (C) Matrigel and on a control of (D) laminin-coated glass. After 7 days in culture with medium containing 10% FBS, the cells were stained with a live/dead viability assay. The cells on Matrigel-coated paper had the highest (E) viability and (F) number of adhered cells likely due to proliferation and/or adhesion properties in comparison to fibronectin and laminin coated paper. (G) Immunostaining with the astrocytic marker GFAP confirmed the astrocyte phenotype of the majority of rat cortex derived cells cultured for 14 days on Matrigel coated paper. Significance levels: ^*p < 0.05; ^**p < 0.005; ^***p < 0.0005; ^****p < 0.00005.

Figure 2

Timeline and schematic of the paper-based co-culture platform supporting low-density neuron cultures. (A) The precursor astrocyte feeders were cultured and proliferated on paper rings in medium supplemented with serum for 1 week before combination with the experimental neuron culture. (B) After a week, freshly dissociated neurons were seeded onto glass coverslips to create the experimental neuron cultures. After 24 h, the 8 DIV astrocyte feeder culture was added to the 1 DIV experimental culture. The paper ring was placed on top of a Parafilm ring submerged in the media to prevent direct contact of the two cultures. (C) Viability and neurite length of the experimental neuron culture was measured at 5 DIV with a live/dead assay. The experimental neuron culture activity was assessed at 14 DIV using calcium imaging. (D) Picture of the co-culture platform in a 24-well plate.

Figure 3

Effect of astrocyte co-culture on the viability of low-density neuron cultures. Primary rat cortical neurons were seeded on PDL-coated coverslips at densities ranging from 1,000 to 50,000 cells/cm² with paper-based astrocyte co-culture and without co-culture as a control. The viability was measured using a live/dead viability assay at 5 DIV. (A) The viability of the experimental neuron culture supported by the co-culture was significantly improved compared to the control without co-culture. (B) The density of live experimental neurons on the coverslip was calculated using the live-dead assay images and compared to the seeding density. The co-culture condition showed minimal loss of cells compared to the control condition. (C) Representative images of the experimental neuron culture with and without co-culture at different seeding densities. Plotted values are mean ± standard error and significance was tested using Mann-Whitney U test with significance levels: ^* if p < 0.05; ^** if p < 0.005; ^*** if p < 0.0005.

Figure 4

Effect of co-culture on neurite length as an indicator for neuronal development. The neurite length of the experimental neuron cultures in co-culture and control condition was measured at 5 DIV. Tracing and quantification was done using the ImageJ Plugin NeuriteTracer. (A) Comparison of neurite lengths of the experimental neuron cultures between co-culture and control conditions. (B) The fold change of the resulting total neurite length indicated a trend toward an increased neurite length in the co-culture condition compared to the mono-culture control. In the plot, the box corresponds to the first and third quartiles with the median. The median is indicated with a horizontal line in the box, and the maximum and minimum are at the ends of the whiskers. Measurements 1.5 interquartile ranges below the first quartile or above the third quartile are classified as outliers and excluded.

Figure 5

Effect of astrocyte co-culture on the spiking activity of low-density neuronal cultures. The cultures were transduced with the gene-encoded calcium indicator GCaMP6f at 7 DIV. At 14 DIV, the activity of the experimental neuronal cultures was measured using the fluorescent calcium indicators. To obtain the spiking activity, the calcium activity videos were segmented into individual cells followed by spike detection. (A) Spontaneous activity was detected in all cultures down to 1,000 cells/cm². The culture-wide spike frequency increases progressively with the seeding density. (B) The fold increase in spike frequency per neuron between the co-cultured and control experimental neuron cultures showed an increase in spike frequency at low densities. In the plot, the box corresponds to the first and third quartiles with the median. The median is indicated with a horizontal line in the box, and the maximum and minimum are at the ends of the whiskers. Measurements 1.5 interquartile ranges below the first quartile or above the third quartile are classified as outliers and excluded. Statistical significance was tested using Mann-Whitney U-test with significance levels: ^* if p < 0.05; ^*** if p < 0.0005.

Figure 6

Applications and versatility of paper-based co-culture platform. (A,B) Effect of astrocytic co-culture on low-density neurons on microcontact-printed patterns. Patterned networks often require low densities to avoid clumping of neurons on the cell-adhesive regions, resulting in low viability. (C,D) Cell viability and density was improved under co-culture condition compared to control mono-cultures. (E) Paper-based culture can be structured to fit any culture vessel, such as well plates, multielectrode arrays and flasks, for versatile applications. Significance levels: ^*p < 0.05.