Millisecond-scale differences in neural activity in auditory cortex can drive decisions

Abstract

Neurons in the auditory cortex can lock with millisecond precision to the fine timing of acoustic stimuli, but it is not known whether this precise spike timing can be used to guide decisions. We used chronically implanted microelectrode pairs to stimulate neurons in the rat auditory cortex directly. Here we demonstrate that rats can exploit differences in the timing of cortical activity as short as three milliseconds to guide decisions.Animals can detect the fine timing of some stimuli. For example, interaural time differences of less than one millisecond are used for the spatial localization of sound 1 . It is also clear that cortical neurons can lock with millisecond precision to the fine timing of some stimuli in the auditory cortex 2, 3 , the visual cortex 4 , somatosensory cortex 5,6 and in vitro 7 . Furthermore, spike generation in the auditory cortex is controlled by a stereotyped and precisely timed sequence of excitatory input followed approximately three milliseconds later by inhibitory input 8 . However, although it has recently been established that even a few cortical spikes are sufficient to drive decisions 9, 10 , it has been difficult to establish whether the fine timing of cortical activity can suffice.We therefore set out to probe the precision with which the fine timing of neural activity in the auditory cortex could guide behaviour in the rat. For the spatial localization of sound, the relevant sub-millisecond interaural time difference cues are extracted by specialized subcortical structures. To ensure that we were probing cortical rather than subcortical mechanisms, we bypassed subcortical auditory pathways and trained animals to respond to direct intracortical electrical stimulation. We used transient biphasic current trains delivered via two chronically implanted intracortical microelectrodes 11-13 to stimulate two populations of neurons in primary auditory cortex (area A1; Fig. 1a). We designed the stimulation patterns so that the only cue available to guide behaviour was the relative timing of the activity elicited in the two cortical populations.We first trained adult male Long Evans rats to perform a simple auditory two alternative choice task (2-AC) 14 . The animal initiated a trial by inserting its nose into the centre port of a three port operant chamber, triggering one of two acoustic stimuli. These stimuli indicated whether the left or right goal port would be rewarded with water ( Fig. 1A; for details, see Supplemental material). Chance performance was 50%. After animals reached criterion performance (> 90%), we implanted electrodes at two sites (A and B; Fig. 1A) about 1.1 mm apart along the rostro-caudal axis in the rat's left primary auditory cortex (Fig. S3). We then substituted electrical stimulation (< 30 µA) through these