A causal role for the pulvinar in coordinating task-independent cortico-cortical interactions

Abstract

The pulvinar is the largest nucleus in the primate thalamus and has topographically organized connections with multiple cortical areas, thereby forming extensive cortico-pulvino-cortical input-output loops. Neurophysiological studies have suggested a role for these transthalamic pathways in regulating information transmission between cortical areas. However, evidence for a causal role of the pulvinar in regulating cortico-cortical interactions is sparse and it is not known whether pulvinar's influences on cortical networks are task-dependent or, alternatively, reflect more basic large-scale network properties that maintain functional connectivity across networks regardless of active task demands. In the current study, under passive viewing conditions, we conducted simultaneous electrophysiological recordings from ventral (area V4) and dorsal (lateral intraparietal area [LIP]) nodes of macaque visual system, while reversibly inactivating the dorsal part of the lateral pulvinar (dPL), which shares common anatomical connectivity with V4 and LIP, to probe a causal role of the pulvinar. Our results show a significant reduction in local field potential phase coherence between LIP and V4 in low frequencies (4-15 Hz) following muscimol injection into dPL. At the local level, no significant changes in firing rates or LFP power were observed in LIP or in V4 following dPL inactivation. Synchronization between pulvinar spikes and cortical LFP phase decreased in low frequencies (4-15 Hz) both in LIP and V4, while the low frequency synchronization between LIP spikes and pulvinar phase increased. These results indicate a causal role for pulvinar in synchronizing neural activity between interconnected cortical nodes of a large-scale network, even in the absence of an active task state.

Keywords: cortico-cortical interactions; macaque electrophysiology; pulvinar causal role; reversible-inactivation with muscimol; transthalamic pathways.

Figure 1.. Anatomical localization of inactivation zone.

(A) Structural MRI scan visualizing the initial site of a Gadolinium MRI contrast agent injection (0.5μl) in monkey B and monkey R. MR images of individual animals were non-linearly wrapped onto the D99 digital template atlas to approximate the boundaries of pulvinar subdivisions. PM: Medial pulvinar, PL: Lateral pulvinar, PI: Inferior pulvinar. The injection site was localized within dorsal portions of the lateral pulvinar (dPL) for both monkeys. (B) Serial MRI images showing the spread of Gadolinium contrast agent over a duration of 3 hours, following injection into dPL. The inserts visualize the spread of the contrast agent at 15, 60, and 180 minutes, respectively, after the end of the contrast injection, during one session in monkey B. The orange bar indicates the time window (15-minutes, starting from 15 minutes after the end of injection) used for the post-injection analyses of recordings.

Figure 2.. dPL inactivation did not significantly change spike rates and local field potential power in LIP and V4.

(A) Left: Changes in LIP spike rates before (green) and after (black) muscimol or control injections into dPL. Right: Percentage change in local field potential (LFP) power in LIP across a frequency range of 4-90 Hz during post vs. pre-injection windows, for muscimol (magenta) and saline control (black) injections. (B) Left: Changes in V4 spike rates before (green) and after (black) muscimol or control injections into dPL. Right: Percentage change in local field potential (LFP) power of V4 across a frequency range of 4-90 Hz during post vs. pre-injection windows, for muscimol (magenta) and saline control (black) injections. Combined data from both monkeys. (Error bars/shaded areas: s.e.m.).

Figure 3.. LFP-LFP coherence between LIP and V4 significantly decreased following pulvinar inactivation.

Strength of LFP-LFP phase coherence (measured as phase locking value – PLV) between LIP and V4 across a 4-60 Hz frequency range during pre- (green) and post- (black) injection windows, shown separately for muscimol (left) and control (right) sessions. The post-injection LFP-LFP phase coherence between LIP and V4 was significantly lower compared to the pre-injection coherence in theta (4-7 Hz) and alpha/low beta (7-15 Hz) frequency ranges for muscimol, but not for control sessions. Combined data from both monkeys. Shaded areas: s.e.m..

Figure 4.. Changes in spike-field coherence (SFC) between pulvinar and cortical areas following dPL injections.

(A) Left: Change in SFC between pulvinar spikes and LIP LFP phase over a frequency range of 4-60 Hz during post vs. pre-injection windows for muscimol (magenta) and control (black) sessions. The dPL spike-LIP phase coherence decreased significantly in the 4-15 Hz range following muscimol injections into dPL. Right: Change in SFC between LIP spikes and pulvinar LFP phase during post vs. pre-injection windows for muscimol (magenta) and control (black) sessions. The LIP spike-pulvinar phase coherence increased significantly in the 4-15 Hz range following muscimol injections. (B) Left: change in SFC between pulvinar spikes and V4 LFP phase during post vs. pre-injection windows for muscimol (magenta) and control (black) sessions. The dPL spike-V4 phase coherence decreased significantly in the 4-15 Hz range following muscimol injections into dPL. Right: Change in SFC between V4 spike and pulvinar LFP phase during post vs. pre-injection windows for muscimol (magenta) and control (black) sessions. Shaded areas: s.e.m..

Figure 5.. Thalamo-cortical interactions under normal conditions (A) and during pulvinar inactivation (B).

Typically, pulvinar influences cortical regions (here LIP and V4) to coordinate cortico-cortical interactions (A). After muscimol injections into dPL, pulvinar influences on LFP phase of LIP and V4 decreased significantly in low frequencies (4 - 15 Hz). At the same time, low frequency coherence from LIP, but not V4, to pulvinar significantly increased following dPL inactivation, suggesting that the transthalamic pathway is under parietal cortex control (B).