On the co-occurrence of startles and hippocampal sharp waves in newborn rats

Abstract

Hippocampal sharp waves (SPWs) are among the earliest neural population patterns observed in infant mammals. Similarly, startles are among the earliest behavioral events observed. Here we provide evidence indicating that these two events are linked mechanistically soon after birth in freely moving and head-fixed 1 to 4-day-old rats. EMG electrodes and intrahippocampal silicon depth electrodes were used to detect the presence of startles and SPWs, respectively. In intact pups, the majority of sharp waves were preceded by startles (average latency: 161 ms). When the hippocampal formation was surgically separated from the brainstem, however, sharp waves and startles still occurred, but now independently. In addition, unrelated to startles or SPWs, gamma oscillations were detected in several subjects, as were neocortical "spindles" that propagated passively into the hippocampus. The co-occurrence of sharp waves and startles provides the opportunity for Hebbian changes in synaptic efficacy and, thus, is poised to contribute to the assembly of neural circuits early in development.

Figure 1

(A) Representative recording of SPWs preceded by startles in a head-fixed P2 rat. Traces from top: CA1 multiunit activity (MUA; filter: 500-5000 Hz); local field potentials (LFP) recorded from CA1 and stratum radiatum (filter: >100 Hz); EMG activity in the nuchal and vastus lateralis muscles. Individual instances of SPWs and startles are indicated. (B) Shaded area from (A) expanded to reveal temporal structure of the startle-SPW complex in detail.

Figure 2

(A) Peristimulus histogram of mean (+ s.e.) number of startles, determined from EMG records and triggered by SPWs, in intact and head-fixed P1-4 rats (n = 4). The inset shows the averaged nuchal EMG amplitude in relation to the peak of a SPW (t = 0; vertical line). N = 92 SPWs. (B) Same as in (A) accept pups were decerebrated before testing (n = 6). N = 83 SPWs. The inset figures are produced by averaging nuchal EMG after downsampling to 250 Hz.

Figure 3

(A) Representative gamma burst in a head-fixed P2 rat during sleep. Traces from top: CA1 multiunit activity (MUA; filter: 500-5000 Hz); local field potentials (LFP) recorded from CA1 and stratum radiatum (filter: >100 Hz). Note the phase reversal across the two hippocampal layers and the concurrent increase in multiunit activity. (B) Depth profile during a gamma burst (shaded region) for the same P2 subject. Note that the highest amplitude is recorded in the middle portion of the stratum radiatum. Each recording site has a vertical separation of 100 μm. (C) Depth profile of a startle-related SPW (left) and a neocortical “spindle” (right) from another P2 rat during sleep. For the SPW, note the phase reversal between CA1 and stratum radiatum (asterisk). For the spindle, note the absence of phase reversals and the diminishing amplitude as the wave travels from neocortex toward hippocampus.