Covariations in coordinated preparation in this model could give rise to saccade and reach RT correlations. Analyzing the link between RT and neural activity might reveal shared representations that control both movements together. We trained two monkeys to make either coordinated reaches and saccades (Figure 2A)

or saccades alone (Figure 2B) to a visually cued Caspase activation target. Before coordinated movements, saccade RTs (SRTs) were correlated with reach RTs (RRTs; example in Figure 2C; R = 0.69, mean SRT = 190 ms, mean RRT = 280 ms). Across 105 experimental sessions, SRT-RRT correlations were 0.50 ± 0.24 (mean ± std). Mean SRT across the population was also significantly faster when the saccade was made with a reach (243 ± 0.6 ms, mean ± SEM) than when it was made alone Selleckchem Lonafarnib (252 ± 0.6 ms; p < 0.001). These results demonstrate that correlations exist between RTs for saccades and reaches such that saccades can be initiated more quickly when made with a reach. We recorded spiking and LFP activity from 105 sites in area LIP (74 in Monkey H; 31 in Monkey J), 135 sites in PRR (53 in Monkey H; 82 in Monkey J) and 36 visually responsive sites in V3d (36 in Monkey J; Figures 3A and S1). We first present example activity from a single session recorded in area LIP during the reach and saccade task. Spiking and LFP activity in area LIP showed robust selectivity for the preferred (Figure 3Bi) compared

with the null (Figure 3Ci) direction. Spatial tuning was present in LFP activity with different dynamics at different frequencies. One pattern of power changes was present before movements to the

preferred direction (Figure 3Bii), and another pattern was present before movements to the null direction (Figure 3Cii). LFP power was generally greatest around 15–17 Hz below in the beta-frequency band and decreased relative to baseline for preferred direction trials (Figure 3D). In contrast, LFP power increased at frequencies above ∼30 Hz in the gamma-frequency band, and the opposite pattern was present for trials in the null direction (Figure 3E). Thus, reach and saccade movements influence the rate of spiking as well as LFP power in both gamma- and beta-frequency bands. To build a link between neural activity and coordination, we then related LFP power and spike firing rate to saccade and reach RTs. We started by considering LFP power. We examined whether LFP activity predicts movement RTs by grouping LFP power during trials with the slowest or fastest RTs. We selected LFP activity from 72 sites in area LIP with at least 60 trials in each direction and for each task (Monkey H: 57 sites; Monkey J: 15 sites). Before reach and saccade movements in the preferred direction, beta-band LFP power (15 Hz) was significantly greater during the 33% of trials with the slowest SRTs than for the 33% of trials with the fastest SRTs (Figure 4A; p < 0.05, rank-sum test).