Isoflurane and ketamine differentially influence spontaneous and evoked laminar electrophysiology in mouse V1

Abstract

General anesthesia is ubiquitous in research and medicine, yet although the molecular mechanisms of anesthetics are well characterized, their ultimate influence on cortical electrophysiology remains unclear. Moreover, the influence that different anesthetics have on sensory cortexes at neuronal and ensemble scales is mostly unknown and represents an important gap in knowledge that has widespread relevance for neural sciences. To address this knowledge gap, this work explored the effects of isoflurane and ketamine/xylazine, two widely used anesthetic paradigms, on electrophysiological behavior in mouse primary visual cortex. First, multiunit activity and local field potentials were examined to understand how each anesthetic influences spontaneous activity. Then, the interlaminar relationships between populations of neurons at different cortical depths were studied to assess whether anesthetics influenced resting-state functional connectivity. Lastly, the spatiotemporal dynamics of visually evoked multiunit and local field potentials were examined to determine how each anesthetic alters communication of visual information. We found that isoflurane enhanced the rhythmicity of spontaneous ensemble activity at 10-40 Hz, which coincided with large increases in coherence between layer IV with superficial and deep layers. Ketamine preferentially increased local field potential power from 2 to 4 Hz, and the largest increases in coherence were observed between superficial and deep layers. Visually evoked responses across layers were diminished under isoflurane, and enhanced under ketamine anesthesia. These findings demonstrate that isoflurane and ketamine anesthesia differentially impact sensory processing in V1. NEW & NOTEWORTHY We directly compared electrophysiological responses in awake and anesthetized (isoflurane or ketamine) mice. We also proposed a method for quantifying and visualizing highly variable, evoked multiunit activity. Lastly, we observed distinct oscillatory responses to stimulus onset and offset in awake and isoflurane-anesthetized mice.

Keywords: anesthesia; evoked; laminar electrophysiology; spontaneous.

Fig. 1.

Anesthetics alter temporal aspects of resting state population activity. A: representative spike streams recorded from layer IV in the same animal at the same electrode site. Multiunit activity (MUA) under ketamine anesthesia exhibits bursting characteristics. B: local field potentials (LFPs) corresponding to spike streams from A. Large, negative deflections in the ketamine example match bursts in the MUA. C: average firing rates along the length of the electrode shank, computed over 1 s intervals, were slightly greater under ketamine anesthesia. Amplitude of the noise floor did not differ significantly across anesthetics. D and E: LFP peak-to-peak (V_P2P; D) amplitude and sum (V_SUM; E) were significantly greater under anesthesia. Ketamine had significantly greater LFP activity than isoflurane. F: amplitude of the noise floor did not differ significantly across anesthetics (n = 96 electrode sites: *P < 0.001, #P = 0.001, %P = 0.01).

Fig. 2.

Isoflurane and ketamine increase resting-state power at distinct frequencies. A: representative power spectra, recorded from layer IV in the same animal at the same electrode site. B: power spectra from A, normalized by subtracting the awake spectrum from the anesthetized. C and D: anesthesia significantly increased mean and peak power, along the depth of the cortex. E: awake-induced local field potential (LFP) power against depth, averaged across pseudotriggers, and then animals. Depth is relative to layer IV, indicated by the dotted black regions. F: averaged power against depth under isoflurane and ketamine anesthesia, normalized to awake, as in B. Largest increases in power occurred at alpha and beta frequencies under isoflurane, and delta and gamma frequencies under ketamine. G: relative power within frequency bands were differentially affected by isoflurane and ketamine. (n = 96 electrode sites: *P < 0.001, #P = 0.002). P_ANES, anesthetized power spectra; P_AWAKE, awake power spectra.

Fig. 3.

Resting state coherence increases under anesthesia. A: mean pairwise coherence across electrode sites for awake (A), isoflurane-anesthetized (I), and ketamine-anesthetized (K) animals. Each cell represents the mean coherence between two electrode sites, averaged across pseudo triggers and animals. Cells along the diagonal show the average coherence between signals recorded from the same electrode site and are thus equal to one. Dotted black regions indicate layer IV. White rectangles border supragranular-infragranular (SG-IG), granular-infragranular (G-IG), and granular-supragranular (G-SG) regions. B: laminar SG-IG, G-IG, and G-SG coherence against frequency. Interlaminar coherence for each animal was calculated as the average coherence in each region. Bold lines indicate the mean coherence across animals, and dim lines indicate standard error. Shaded regions encompass alpha-, beta-, and gamma-frequency bands and are used in the calculations for C. C: mean coherence between 7 and 90 Hz, averaged across animals. Error bars denote SE (n = 6 mice: *P < 0.05).

Fig. 4.

Quantification of evoked multiunit activity. A: representative peristimulus time histograms of evoked responses, recorded from layer V. Each row shows examples from the same mouse at the same electrode site across anesthetics. Evoked responses exhibited dynamic temporal variability across anesthetics. B: example diagram for how multiunit yield was calculated. Multiunit activity (MUA) was quantified by comparing spike counts, X_S, within a bin of duration, B, at some latency, L, after each stimulus. This distribution of spike counts was then compared with the spike counts within the same bin duration, at some latency, L′, before stimulus presentation, using a paired t-test. Analysis was repeated for varying bin sizes and latencies to capture dynamic changes in the evoked response and then performed on all channels and all animals. C–E: pseudocolor plots demonstrate the resultant multiunit yield and signal to noise firing rate ratio (SNFRR, insets). For ease of visualization, all plots show negative latency (L′) equal to 0, such that all bin sizes and poststimulus latencies are compared with the same duration immediately before the stimulus. Awake animals had a consistent, strong transient response, followed by ~100 ms of slightly elevated MUA. Alternatively, isoflurane-anesthetized animals did not produce a strong transient response, and firing rates were only slightly elevated compared with prestimulus intervals. Ketamine anesthesia produced a strong transient response, but periodic bursts of MUA were still observed in the prestimulus interval (A and E). Note that the negative SNFRR indicates that firing rates during the prestimulus interval exceeded those in the poststimulus interval, for the parameters specified by B and L. F and G: parameter values that optimized multiunit yield are shown without the negative latency parameter, as in C–E (F), and with the negative latency parameter (G). Iso, isoflurane; KX, ketamine/xylazine.

Fig. 5.

Laminar responses to visual stimulus further demonstrate temporal variability. A: multiunit firing rate across depth and time, averaged across 64 stimuli and then across animals. Stimulus presentation is shown by gray bar. Strong transient responses were observed in awake and ketamine-anesthetized mice. Isoflurane reduced the temporal synchrony and prolonged the evoked response. Bursting multiunit activity observed in the spontaneous condition under ketamine anesthesia were apparent in the visually evoked response. B: induced current source densities, averaged across 64 stimuli and then across animals. Awake animals demonstrate succinct laminar processing, compared with ketamine. Isoflurane current source density is the weakest and shortest. C–F: quantification of local field potential (LFP) showed synchronous population activity in the awake case (A; C and D) and synchronous and sustained population activity under ketamine (K; C–F). Isoflurane (I) exhibited a weak evoked response (n = 96 electrode sites: *P < 0.001, %P = 0.02). CSD, current source density; MUA, multiunit activity; V_P2P, LFP peak-to-peak amplitude; V_SUM, LFP sum amplitude; V_ON, amplitude after stimulus onset; V_SPON, spontaneous amplitude.

Fig. 6.

Visually evoked laminar coherence increases under anesthesia. A: normalized pairwise coherence, measured as the difference between coherence during ON periods and spontaneous coherence, for awake (A), isoflurane-anesthetized (I), and ketamine-anesthetized (K) animals. B: evoked supragranular-infragranular (SG-IG), granular-infragranular (G-IG), and granular-supragranular (G-SG) coherence (means ± SE) against frequency. Coherence for each animal was calculated as the average coherence in each region encompassed by the white borders in A. C: mean coherence between 7 and 90 Hz (shaded region in B), averaged across animals. Error bars denote SE (n = 6 mice: #P < 0.05, *P < 0.01).

Fig. 7.

Anesthetics alter rhythmic properties of the evoked response. A: representative power spectra recorded from layer V at the same electrode site in the same mouse. Spontaneous power spectra at the same electrode site are shown as dotted lines. Broadband increases in power were observed under anesthesia. B: power spectra in A, normalized by subtracting the spontaneous (P_SPON) power spectrum from the evoked. A pronounced, visually evoked gamma peak was observed under isoflurane at slower frequencies than in the awake state. Largest increases in power after visual stimulation (P_ON) under ketamine occurred at delta frequencies and across high frequencies. P_OFF, power before stimulus onset. C and D: isoflurane and ketamine significantly increased power across the depth of the cortex. E: power across the depth of the cortex, averaged across 64 stimuli and then across animals, normalized to resting state as in B.

Fig. 8.

The evoked before stimulus onset (OFF) response has similar structure of laminar coherence as after stimulus (ON) response. A: normalized pairwise coherence changes during OFF periods for awake (A), isoflurane-anesthetized (I), and ketamine-anesthetized (K) animals. B: coherence (means ± SE) between supragranular-infragranular (SG-IG), granular-infragranular (G-IG), and granular-supragranular (G-SG) layers. C: average coherence from 7 to 90 Hz, averaged across animals. Error bars indicate SE (n = 6 mice: *P < 0.05).