Biomedical Engineering Reference
In-Depth Information
approaches and for specific applications in the domains of pharmacology, toxicology, and
drug development.
The aim of this chapter is to describe and summarize the experimental basis for the sug-
gested applications with emphasis on biosensors. Once the utility of these systems is
validated, technical improvements can quickly adapt NNBS procedures for a great variety
of applications. Prototype systems of compact portable recording have already been real-
ized [7, 8] and are briefly described herein. Dual network recording on the same MEA,
providing a “twin” network as a control, is already a routine procedure; and eight-
network recording modules, served by a liquid handling robot, are under development at
the Center for Network Neuroscience (CNNS).
6.2
Methods
6.2.1
Cell Culture
The cell culture procedures are well established and generate robust, spontaneously active
networks that can withstand 50 g of tangential impact and allow shipping to other labo-
ratories. Many experiments performed at the U.S. Naval Research Laboratory have relied
on shipment of intact living networks grown on MEAs at the CNNS [9, 8]. In addition to
the spinal cord and frontal cortex networks discussed in this chapter, other neuronal cell
systems such as auditory cortex [10, 2] and brain stem [11] have been used and grown on
MEAs with nearly identical methods. The only major difference between spinal and corti-
cal cell culture development is the use of a glycine-free medium for spinal tissue to avert
glycine-induced inhibition of spinal circuit activity. Cultures are prepared according to the
basic pioneering methods established in the laboratory of Dr. Phillip Nelson [12]. Central
nervous system (CNS) tissues are obtained from embryonic HSD:IRC mice (Harlan,
Sprague, Dawley, Indianapolis, IN). Spinal cord tissue is dissected at 14-15 days' gestation
and cortical tissue at 16-17 days. The tissue is dissociated enzymatically (papain) and
mechanically (via trituration), and seeded at a concentration of 0.2-0.5
10
6
cells per mil-
liliter onto electrode arrays coated with polylysine and laminin to promote cell adhesion.
The cell pool is seeded as 0.1 ml droplets (approximately 20,000-50,000 cells or 2,800-7,100
cells per square millimeter). Assuming a neuron to nonneuron cell ratio of 1:10, this pro-
cedure provides a neuronal density of approximately 300-700 neurons per square mil-
limeter. After cell adhesion (~2 h), 1 ml of medium is added, confined by a silicon gasket.
Spinal cord and cortical cultures are incubated in minimal essential medium (MEM) and
Dulbecco's modified minimal essential medium (DMEM), respectively, each supple-
mented with 5% horse serum and 5% fetal bovine serum for a week. Thereafter, the cells
are fed twice a week using half-media changes with the respective media containing only
5% horse serum. Typically, 80% of the cultured networks exhibit robust spike activity on
at least 50% of the 64 recording sites. Mean SNRs are generally in a range from 3-6 with
several channels exceeding SNRs of 10. This level of activity has proven to be suitable for
reproducible pharmacological studies. Thereafter, the cells are fed twice a week using half-
media changes with the respective media containing only 5% horse serum. Moreover,
pooled tissues from 10 embryonic mouse spinal cords, for example, can seed more than
200 MEAs. Similarly, specific brain regions, such as frontal cortex, can easily seed more
than 50 MEAs. This provides a heretofore unprecedented efficiency in tissue utilization
and could clearly serve to reduce the number of animals presently necessary for neuro-
toxicological investigations.
