An extremely rich repertoire of bursting patterns during the development of cortical cultures

Abstract

Background: We have collected a comprehensive set of multi-unit data on dissociated cortical cultures. Previous studies of the development of the electrical activity of dissociated cultures of cortical neurons each focused on limited aspects of its dynamics, and were often based on small numbers of observed cultures. We followed 58 cultures of different densities--3000 to 50,000 neurons on areas of 30 to 75 mm2--growing on multi-electrode arrays (MEAs) during the first five weeks of their development.

Results: Plating density had a profound effect on development. While the aggregate spike detection rate scaled linearly with density, as expected from the number of cells in proximity to electrodes, dense cultures started to exhibit bursting behavior earlier in development than sparser cultures. Analysis of responses to electrical stimulation suggests that axonal outgrowth likewise occurred faster in dense cultures. After two weeks, the network activity was dominated by population bursts in most cultures. In contrast to previous reports, development continued with changing burst patterns throughout the observation period. Burst patterns were extremely varied, with inter-burst intervals between 1 and 300 s, different amounts of temporal clustering of bursts, and different firing rate profiles during bursts. During certain stages of development bursts were organized into tight clusters with highly conserved internal structure.

Conclusion: Dissociated cultures of cortical cells exhibited a much richer repertoire of activity patterns than previously reported. Except for the very sparsest cultures, all cultures exhibited globally synchronized bursts, but bursting patterns changed over the course of development, and varied considerably between preparations. This emphasizes the importance of using multiple preparations--not just multiple cultures from one preparation--in any study involving neuronal cultures. These results are based on 963 half-hour-long recordings. To encourage further investigation of the rich range of behaviors exhibited by cortical cells in vitro, we are making the data available to other researchers, together with Matlab code to facilitate access.

Figure 1

Development of burst patterns in a dense culture. Graphs show array-wide spike detection rates (ASDR) per second in recordings from one culture at various ages (numbers in top right corners). Note the different vertical scales used for each day. The spike detection rate of individual electrodes is represented by gray-scale rasters for all 59 electrodes, stacked vertically below each graph. (Each horizontal line pertains to one electrode; gray levels indicate firing rates. The same scale (see color bar next to '4 div') is used for all days. Note that the extreme increase in ASDR during bursts commonly saturates the gray scale. This representation is also used in subsequent figures.) Analogous figures from all 58 cultures used are available online.

Figure 2

Phase contrast micrographs of central area of cortical cultures. Photographs are of three typical cultures of different densities, at 1, 15, and 32 days in vitro. Note that even our 'sparse' cultures are considerably denser than those commonly used for investigating synaptic plasticity with intracellular electrodes (see, e.g., [32]). Scale: electrode spacing is 200 μm center to center.

Figure 3

Classification of observed bursting behaviors. A Overview of the different classes of bursting behavior observed in our cultures. Numbers in parentheses indicate plating batch. Vertical bars indicate partial medium replacement times. Hash patterns indicate burst frequency for all types of burst patterns except superbursts. In batch 3, three cultures received full medium replacements (indicated by thicker bars in the lower three cultures of batch 3). One culture in batch 6 got infected after 20 div, and had to be discarded. B Examples of burst pattern classes, with array-wide spike detection rates and gray-scale rasters for all electrodes, all taken from dense cultures. B1 No bursting. B2 Tiny bursts. B3 Fixed size bursts. B4 Variably sized bursts. B5 Long-tailed bursts. B6 Regular superbursts. B7 Inverted superbursts. B8 Dramatic burst rate variation. Gray scales are as in Figure 1.

Figure 4

Development of firing and bursting activity in dense cultures. A Median ASDR as a function of developmental age. Medians were taken across all 1800 one-second-wide time bins in individual half-hour long recordings. (Since bursts occupy a small fraction of time bins, a culture's median ASDR over time is a good indication of the baseline ASDR outside of bursts.) B Burstiness index as a function of developmental age. Dots in A and B are measurements from individual cultures, colored by plating batch. Black lines are interpolated averages across all cultures, using a Gaussian window with a half-width of 1 day. Dots were horizontally jittered by ± 0.25 days for visual clarity.

Figure 5

Comparison of the development of cultures of different sizes. A Actual density of cultures at 1 div. B Maximum (across days) of ASDR (averaged over 30 minutes of recording) observed in the first 35 div. C First day on which ASDR reached half of its maximum. D First day with burstiness index greater than 0.25. Error bars indicate the mean and the sample standard deviation. Horizontal jittering of dots is for visual clarity only. Note that the vertical scale is logarithmic in A and B, which explains why the error bars appear asymmetric. In small-and-sparse and ultra-sparse cultures, the ASDR remained so low that the age at which half of the maximum was reached could not be measured accurately, and the BI never reached 0.25. Therefore, no data are shown.

Figure 6

Characterizing burst shapes in dense cultures. A Parameters that define the shape of population bursts. We measured the ASDR as a function of time during the burst, smoothed with a 10 ms Gaussian filter. The 20% and 80% points between baseline and peak ASDR were determined and used to define the various phases. B Total duration of individual bursts. (Note log scale on y-axis.) C Duration of onset phases. D Duration of offset phases. In all panels, dots represent individual bursts, horizontally jittered for clarity, and colored by plating batch. Lines are interpolated averages, computed in log-space, using a Gaussian window with a half-width of 1 day.

Figure 7

Comparison of burst sizes during culture development. Scatter plot of total number of spikes in burst and number of participating electrodes. Colors represent bursts from different (dense) cultures. Black traces are the frequencies (in bursts per minute; bpm) of bursts with a given number of participating electrodes, averaged across all cultures represented. Note log scale on y-axis.

Figure 8

Strength of responses to stimulation during the development of dense cultures. A Increase in ASDR during the first 50 ms post-stimulus, averaged over all 50 presentations and 59 electrodes. Each dot represents a set of stimuli delivered to one culture on a given day, colored by plating batch and horizontally jittered for visual clarity. The line is the interpolated average across all cultures, using a Gaussian window with a half-width of 1 day. B Fraction of stimuli that evoked a burst.

Figure 9

Development of functional projections in cultures of different densities. Median distance of sites of non-synaptic responses to stimulated electrode (solid, dots) and 90th percentile of distance distribution (dashed, circle) in individual recordings from dense culture, small cultures, and sparse cultures (left to right). The diameter of the MEA (maximum electrode distance) is 1.72 mm. Projections likely continued to grow beyond this length, especially in the dense cultures, but our method is incapable of following that development. (Interpolation was performed with a window half-width of 1 day for the leftmost panel, and 2 days for the other two, to obtain a smooth curve for the smaller data sets.)

Figure 10

Sensitivity of spiking activity to mild mechanical perturbation (movement of the culture dish). A Number of bursts in first minute after culture was moved into the recording device, normalized to burst rate 10–30 minutes later. B Mean ASDR in first minute after culture was moved into the recording device, normalized to ASDR 10–30 minutes later. C Example of mechanically-induced bursting, recorded at 8 div from a dense culture. Gray scale is as in Figure 1. Lines are interpolations of the data, using a Gaussian window with a half-width of 1 day.

Figure 11

Comparison of activity after moving a culture into the recording device with activity around 12 hours later. Comparison of activity in first 15 minutes after moving a culture into the recording device (left-hand graphs) with activity in the same culture around 12 hours later (right-hand graphs), in cultures of various ages in vitro. Examples are from different cultures. Gray scales are as in Figure 1.

Figure 12

Quantifying sources of variability in activity levels. A Variability of median ASDR. B Variability of burstiness index. We compared the day-to-day variability for individual dense cultures with variability between sister cultures, and with variability between different platings. Interpolated using a Gaussian window with a half-width of 2 days.

Figure 13

Distribution of inter-burst intervals and burst sizes. A–B Distribution of burst sizes. C–D Distribution of interburst intervals. Data are shown from 8 recordings from arbitrarily selected cultures (different colors), at around 7 div (A and C), and around 35 div (B and D). Bimodality in the IBI distribution results from temporal clustering of bursts. Data were binned using a Gaussian window in logarithmic space, bin size was 5% of a decade.