Abstract
Laser-deinsulated, printed-ircuit electrodes integrated into the floor of culture chambers have been used to monitor the spontaneous activity of mouse spinal monolayer cell cultures. This technique has allowed a multisite analysis of activity over long periods of time in closed chambers. In 63 cultures investigated 3–5 weeks after seeding, 89% included single- or multiunit bursting. Based on a subset of 40 cultures in which all electrodes were sequentially scanned, bursting was found on 41% of the active electrodes (approximately 38% of all units monitored). A total of 35% of the electrodes monitoring spontaneous bursting activity revealed rhythmic sequences that were usually coupled among multiple electrodes. Although most of this coupling was in-phase, three out of 40 cultures exhibited antiphasic bursting. In all cases where coupling was observed, each electrode monitored different burst compositions, demonstrating that the activity was generated by different units. Some rhythmic patterns persisted for over 12 hr and were observed in 400 mm2 monolayer cultures, as well as in much smaller 3 mm2 adhesion islands. The addition of 10 mM MgCl2 consistently blocked both random and patterned (i.e., bursting) spontaneous activity at all recording sites. Strychnine (10(-6) M) typically increased firing frequencies and either disrupted pretest bursting or generated rhythmic activity from random phasic patterns. In certain cases, strychnine also blocked activity on specific electrodes, indicating that glycine is not the only inhibitory transmitter involved. The spontaneous appearance of rhythmic activity in low- density, monolayer cell cultures established from dissociated and randomly seeded spinal tissue can be explained by one or a combination of two hypotheses: an inherent specificity of some interconnections in developing mammalian cultures and the generation of organized activity by random circuits at certain stages of complexity.