Synchronous, focally modulated beta-band oscillations characterize local field potential activity in the striatum of awake behaving monkeys

Abstract

Synchronous oscillatory activity has been observed in a range of neural networks from invertebrate nervous systems to the human frontal cortex. In humans and other primates, sensorimotor regions of the neocortex exhibit synchronous oscillations in the beta-frequency band (approximately 15-30 Hz), and these are also prominent in the cerebellum, a brainstem sensorimotor region. However, recordings in the basal ganglia have suggested that such beta-band oscillations are not normally a primary feature of these structures. Instead, they become a dominant feature of neural activity in the basal ganglia in Parkinson's disease and in parkinsonian states induced by dopamine depletion in experimental animals. Here we demonstrate that when multiple electrodes are used to record local field potentials, 10-25 Hz oscillations can be readily detected in the striatum of normal macaque monkeys. These normally occurring oscillations are highly synchronous across large regions of the striatum. Furthermore, they are subject to dynamic modulation when monkeys perform a simple motor task to earn rewards. In the striatal region representing oculomotor activity, we found that small focal zones could pop in and out of synchrony as the monkeys made saccadic eye movements, suggesting that the broadly synchronous oscillatory activity interfaces with modular spatiotemporal patterns of task-related activity. We suggest that the background beta-band oscillations in the striatum could help to focus action-selection network functions of cortico-basal ganglia circuits.

Figure 2.

Properties of striatal LFP oscillations in a macaque monkey (M8) during quiet rest. A, The top trace shows an example of a raw LFP voltage plot with segments identified as oscillatory highlighted by gray shading. The bottom trace shows the same LFP signal, bandpass filtered (10-25 Hz) and squared, illustrating the waxing and waning of the 10-25 Hz oscillations. The horizontal dashed line indicates the threshold used for detection of oscillatory episodes (at or above the 70th to 75th percentile of the filtered and squared voltage). Successive peaks of the filtered and squared LFP that remained higher than this threshold level were considered part of an oscillatory episode. The minimal episode length was set at 200 msec (4 cycles). B, Histogram depicting the percentage of the LFP signal between 10 and 25 Hz for the successive 640 msec time windows (50% overlap) measured during one recording session. The vertical dashed line indicates the 75th percentile level of 10-25 Hz oscillation content. Windows in the upper quartile were used for analysis of frequency and synchrony. C, Number of windows with a particular frequency of LFP oscillations in the sample of 33 rest-condition experiments. D, Spatial analysis of the prevalence of 10-25 Hz oscillatory activity across recording sites in the striatum. The recording chamber on M8 was installed at a 20° off-horizontal and was centered on stereotaxic lateral coordinate L14. Representative recording sites are shown for sagittal 20° oblique planes within the caudate nucleus (CN) (front panel, L10) and putamen (Put) (back panel, L19). For each location within the coordinate system, a graph shows the average FFT compiled from 206 to 633 windows recorded at that site. The shaded zone in each graph denotes the 10-25 Hz frequency band. Axes for the FFT analyses are indicated for exemplars in the caudate nucleus and in the putamen. The x-axis indicates the standard anteroposterior (AP) coordinates; the y-axis indicates the depth relative to the top of the grid in the recording chamber. For simplification, not all recording sites mapped in the two planes shown are represented.

Figure 1.

LFPs recorded from the same microelectrode from a macaque monkey (M8) during rest (A), performance of a visually guided saccade task (B), and drowsiness (C). For each condition, graphs are shown for, from top to bottom, raw voltage plots of the LFPs, bandpass-filtered (10-25 Hz) LFPs, horizontal and vertical eye position, and FFT analysis. FFT analyses were performed during the period within the FFT-labeled zones in A-C. The 10-25 Hz oscillations were spontaneously present during rest (A) and occurred during task trials (B) but were minimal during drowsiness (C).

Figure 3.

Synchrony of 10-25 Hz oscillatory LFP activity across the striatum during rest in monkey M8. A, Examples of simultaneously recorded LFP traces at five recording sites in the striatum (CN1-CN4, caudate nucleus; Put, putamen), and one in the FEF of the same hemisphere. B, FFT analysis for the six traces shown in A in corresponding colors. The graphs in C and E illustrate the degree of correlation of the 10-25 Hz content of LFP oscillations on pairs of electrodes at C sites in the caudate nucleus separated by 2 mm (CN3 × CN4) or 6 mm (CN1 × CN3) and E sites in the caudate nucleus separated by 4 mm (CN2 × CN4) and sites in the caudate nucleus and putamen separated by 10 mm (CN2 × Put). The graphs in D and F illustrate the cross-covariance correlograms for the same pairs of sites. The closer sites showed a stronger linear relationship between the oscillatory content at each site and also a greater cross-covariance between sites. G, Correlation of 10-25 Hz content of the LFP oscillations between a site in the caudate nucleus (CN3) and a site in the FEF. H, Cross-covariance correlogram for the same combination of sites as in G. There was no relationship between the oscillatory content at the two sites, and there was low cross-covariance amplitude between the oscillatory FEF and caudate nucleus LFPs.

Figure 4.

Cross-covariance analysis of 10-25 Hz LFP oscillations recorded during rest in macaque monkey M8. A, Cross-covariance for the CN3 × CN4 pair illustrated in Figure 3C. The solid line shows the average, and the shaded zone shows ±1 SD for 109 of 438 windows (upper quartile of 10-25 Hz content). B, Cross-covariance plots for all 64 pairs. Each correlogram line corresponds to one pair-specific average (51-233 windows analyzed per pair). C, Distribution of the peak average cross-covariance values across the 64 pairs of sites. D, Lag measurements corresponding to the cross-covariance pairs analyzed. E-H, Scatterplots of the cross-covariance values for pairs of striatal sites separated by different distances in the anteroposterior (E), mediolateral (F), and depth (G) planes and by absolute distance separation (H). Depth was measured relative to the top of the grid installed in the recording chamber. The straight lines represent the regression plots for the pairs of variables.

Figure 5.

Phase relation ship of neuronal spike activity of striatal neurons to the 10-25 Hz LFP oscillations recorded at the same sites at which spike activity was recorded. A, LFP-triggered histogram of spike activity of a neuron classified as a phasically active neuron. B, LFP-triggered histogram for a neuron classified as a tonically active neuron. For the analysis in A and B, cycles of the 10-25 Hz filtered LFP activity of each neuron were detected with an amplitude threshold to identify strongest oscillations (threshold level set at 85th percentile of all LFP voltage values). Each of these was marked as a digital event at the peak of the oscillation. Single-unit spike activity occurring within ±250 msec of each peak of the LFP was then plotted (bin width, 10msec, 200 msec period shown). The dashed line shows the average histogram for 50 artificial spike trains (described in Materials and Methods), and the shaded zone shows these values ±1 SD. Peaks in the LFP-triggered histogram deviating from the artificial spike train average by >2 SD were considered significantly related to the LFP cycle. The neurons shown in A and B were both significantly related to the LFP oscillations: A, anti-phase; B, in-phase. Below these LFP-triggered histograms are autocorrelograms of unit activity and interspike interval histograms for the corresponding neurons (bin width, 5 msec). C illustrates the phase-locking relationship for the same neuron illustrated in B with a polar plot of spike times relative to the simultaneously recorded LFP peak converted to angular coordinates (see Materials and Methods) with 15° bins. This neuron had a tendency to fire more in-phase than out-of-phase with the LFP peak and had its largest spike bin at zero-phase. D, Phase relationship for the 27 phase-locked neurons (Table 2) on the basis of the highest histogram peak for each cell calculated in C. This graph shows a tendency for the cell firing to occur just after the peak and just after the valley of the simultaneously recorded LFP cycles.

Figure 6.

Dynamic modulation of 10-25 Hz rhythmic LFP synchrony during performance of the single-saccade oculomotor task by monkey M7. A, Raster plots and peristimulus time histograms of multiunit activity (bin width, 10 msec) at two recording sites in the caudate nucleus recorded during the visually guided single-saccade tasks shown from 500 msec before fixation (0 on time line) to 2500 msec after fixation. Postsaccadic activity was pronounced on E3, but saccade-related activity was slight on E2. A third site (electrode 1, data not shown) in A did not have saccade-related activity. B, Horizontal eye position plots corresponding to time lines shown in C and D. C, Modulation of 10-25 Hz LFP oscillatory activity during the oculomotor task (percentage of the LFP power spectrum between 10 and 25 Hz for successive 128 msec windows) averaged across the entire set of 104 trials analyzed. D, Cross-covariance (synchrony) of the LFP signals between each pairing of the three recording sites. Note that compared with values just before saccade onset, there is a marked drop in synchrony between the site with saccade-related activity (E3) and the site with no saccade-related activity (E1), as shown by E1 × E3.

Figure 7.

Focal task-related desynchronization of striatal LFP activity in the caudate nucleus. A, B, Activity at the three recording sites shown for two consecutive days of recording: the same day as shown in Figure 6 (dotted line) and for the immediately preceding day (solid line). Firing rates normalized to the prestimulus firing rate are shown for the three sites and illustrate consistent saccade-related activity recorded on E3. B, Cross-covariance (synchrony) levels for the pairwise comparisons among the sites for the first day (solid line) and the second day (dotted line). A similar local desynchronization pattern, most prominent for comparison of the task-related E3 site and the non-task-related E1 site, holds across both days (E1 × E3). Vertical lines for saccade and reward indicate the average behavioral event timing for the sessions (dashed for day 1, dotted for day 2). C, D, Focal desynchronization of striatal LFPs during another session of visually guided saccades (29 trials) for three simultaneously recorded LFP sites in the caudate nucleus of monkey M7 recorded 1 year after the recordings shown in Figure 6 and in A and B of this figure. C, Raster plot and histogram of firing rate of multiunit activity located at electrode site Ec. This site showed activity related to fixation and also a postsaccade peak. Site Eb also showed a milder yet similar modulation pattern, unlike site Ea. D, Top, Modulation of 10-25 Hz LFP oscillatory activity during the oculomotor task (percentage of the LFP power spectrum between 10 and 25 Hz for successive 128 msec windows) averaged for the 29 trials. All three sites show a drop in oscillation content after an eye movement (to capture the fixation point or the target). D, Bottom, Cross-covariance (synchrony) of the LFP signals between each pairing of the three recording sites. Note that as shown in Figure 6, here again, LFP activity can stay quite correlated (Eb × Ec) at similar sites, but synchrony can also drop around task execution for sites with differing multiunit activity (e.g., Ea × Ec).