Local domains of motor cortical activity revealed by fiber-optic calcium recordings in behaving nonhuman primates

Abstract

Brain mapping experiments involving electrical microstimulation indicate that the primary motor cortex (M1) directly regulates muscle contraction and thereby controls specific movements. Possibly, M1 contains a small circuit "map" of the body that is formed by discrete local networks that code for specific movements. Alternatively, movements may be controlled by distributed, larger-scale overlapping circuits. Because of technical limitations, it remained unclear how movement-determining circuits are organized in M1. Here we introduce a method that allows the functional mapping of small local neuronal circuits in awake behaving nonhuman primates. For this purpose, we combined optic-fiber-based calcium recordings of neuronal activity and cortical microstimulation. The method requires targeted bulk loading of synthetic calcium indicators (e.g., OGB-1 AM) for the staining of neuronal microdomains. The tip of a thin (200 µm) optical fiber can detect the coherent activity of a small cluster of neurons, but is insensitive to the asynchronous activity of individual cells. By combining such optical recordings with microstimulation at two well-separated sites of M1, we demonstrate that local cortical activity was tightly associated with distinct and stereotypical simple movements. Increasing stimulation intensity increased both the amplitude of the movements and the level of neuronal activity. Importantly, the activity remained local, without invading the recording domain of the second optical fiber. Furthermore, there was clear response specificity at the two recording sites in a trained behavioral task. Thus, the results provide support for movement control in M1 by local neuronal clusters that are organized in discrete cortical domains.

Keywords: local circuits; microendoscopy; monkey; neurophysiology.

Fig. 1.

Experimental arrangement and calcium dye loading in monkey brain in vivo. (A) Scheme of the recording setup with two optical fibers combined with two microelectrodes for stimulation of the neurons and detection of spike activity. APD, avalanche photo diode; ND, neutral density. (B) Arrangement for multicell bolus loading of the tissue consisting of the staining pipette for application of the Ca²⁺-sensitive dye solution and an optical fiber. The green area indicates the region stained with the dye and the blue area, the volume of the tissue illuminated by the excitation light delivered from the optical fiber. (C) Fluorescence emission (Em) of the tissue after application of the dye containing solution monitored with the optical fiber. The excitation energy (Ex) given in arbitrary units (a.u.) was identical at all time points. (D) Plot of the increase of the fluorescence versus time after dye application. (E) Background fluorescence signal of the stained tissue before and after fixation of the optical fiber with agarose and acrylic glue.

Fig. 2.

Relation between distance of arm movement and magnitude of local cortical activity. (A) Experimental arrangement for electrical microstimulation (Stim) and the simultaneous recording of Ca²⁺ signals with an optical fiber (OF). (B) Consecutive trials of microstimulation-evoked neuronal Ca²⁺ signals in M1. Ca²⁺ signals represent local network activity and were evoked by burst stimulation (eight pulses at 100 Hz, 30 µA). Site of recording/stimulation was in M1 at 3 mm below the cortical surface. (C) Relation between arm movements and Ca²⁺ signals. The image indicates the resting position of the arm. The stimulation strength of 17 µA evoked a small Ca²⁺ signal (Upper Right) but no arm movement (Upper Left). At higher stimulation strengths, there is a proportional increase of the movement vector (of constant direction) with the amplitude of the Ca²⁺ signals. Each stimulus consisted of a burst of four pulses at 100 Hz. (D and E) Plots indicating the linear relations between the stimulation strengths and the amplitudes of the Ca²⁺ signal (D) and the relations between the Ca²⁺ signal and the distance of the arm movements (E), respectively. (F) The rise times and decay times of Ca²⁺ signals were constant for each stimulation pulse. Each data point represents the average of five experimental sessions. The red arrows indicate the stimulus threshold for movements. Note that for some points the error bars are smaller than the symbols.

Fig. 3.

Ca²⁺ signals restricted to local cortical domains determine specific movements. (A) Scheme of the experimental arrangement for applying current pulses and recording Ca²⁺ signals simultaneously from two distinct locations of the primary motor cortex. (B, Upper) Delivery of stimulation light through optical fiber 1 (OF1, 488 nm, black trace), produced Ca²⁺ dye-dependent fluorescence signal (blue), but no response was detected by optical fiber 2 (OF2, red). (Lower) Opposite result when stimulation through OF2. (C) Distinct arm movements produced by microstimulation at two sites (Stim1 and Stim2) in M1. Stim1: burst of eight pulses at 100 Hz, 25 µA; Stim 2: burst of eight pulses at 100 Hz, 30 µA. The image indicates the resting position of the arm. (D) The local Ca²⁺ signals corresponding to the movements shown in C. (E and F) The average movement vectors (n = 4) and the corresponding superposition of four consecutive local domain Ca²⁺ signals, as in C and D, respectively.

Fig. 4.

Local domain-specific Ca²⁺ signals recorded during trained complex movements. (A) Schematic of the setup for the vibrotactile detection task (Materials and Methods). (B) Absence of Ca²⁺ signals, both in OF1 and OF2, in the absence of arm movements. (C) Trajectory of a simple voluntary movement of the left arm to push buttons (up) and back to the resting position (down). (D) Ca²⁺ signals corresponding to the movement indicated in C. Superimposed individual traces (four each) and the average traces (n = 7, thick lines) recorded simultaneously at two distinct sites in M1 with the optical fibers OF1 (blue) and OF2 (red). (E) Trajectory of a task-related complex movement of the left arm from rest to the key, then to the push buttons and then back to rest. (F) Ca²⁺ signals corresponding to the movement indicated in E. Superimposed individual traces (three each) and the average traces (n = 10, thick lines). The schemes at the bottom of the traces indicate the sequential steps of the behavioral task.