Effect of sedatives or anesthetics on the measurement of resting brain function in common marmosets

Abstract

Common marmosets are promising laboratory animals for the study of higher brain functions. Although there are many opportunities to use sedatives and anesthetics in resting brain function measurements in marmosets, their effects on the resting-state network remain unclear. In this study, the effects of sedatives or anesthetics such as midazolam, dexmedetomidine, co-administration of isoflurane and dexmedetomidine, propofol, alfaxalone, isoflurane, and sevoflurane on the resting brain function in common marmosets were evaluated using independent component analysis, dual regression analysis, and graph-theoretic analysis; and the sedatives or anesthetics suitable for the evaluation of resting brain function were investigated. The results show that network preservation tendency under light sedative with midazolam and dexmedetomidine is similar regardless of the type of target receptor. Moreover, alfaxalone, isoflurane, and sevoflurane have similar effects on resting state brain function, but only propofol exhibits different tendencies, as resting brain function is more preserved than it is following the administration of the other anesthetics. Co-administration of isoflurane and dexmedetomidine shows middle effect between sedatives and anesthetics.

Keywords: anesthetic effects; common marmoset; resting state network; resting-state functional magnetic resonance imaging.

Fig. 1

RSNs detected in each condition resulting group-ICA. Voxels with z-scores >2.5 are labeled. The 3 brain images in each network diagram show, from left to right, sagittal, coronal, and axial planes of the marmoset brain. DAN, dorsal attention network; DSN, dorsal somatosensory network; VSN, ventral somatomotor network; AN, auditory network; PVN, primary visual network; LVN, lateral visual network; DVN, dorsal visual network; BGN, basal ganglia network.

Fig. 2

The temporal SNR and the head motion plot. The temporal SNR with the post-denoising data (left) or pre-denoising data (middle) was calculated with dividing mean value of signal by its standard deviation. The motion plot shows a scan-by-scan plot of the head motion in the 6 directions estimated by realignment: x (left/right), y (front/back), z (up/down), pitch (rotational direction of nodding and looking up), roll (rotational direction of moving the ear closer to the shoulder), and yaw (rotational direction of looking left/right). All data were calculated for each individual using 1 of the 12 sessions that were imaged and averaged across individuals.

Fig. 3

Betweenness centrality of the 8 regions determined to be significantly different between groups. The bar plots show betweenness centrality, calculated from each hemisphere in each individual, 8 hemispheres in total, in regions where statistically significant difference was observed with both one-way ANOVA and Tukey–Kramer test with P < 0.05. OFC, orbitofrontal cortex; IPS, intraparietal sulcus; PPA, postal parietal area; IPL, inferior parietal lobe; S1, primary somatosensory cortex; AC, auditory cortex; MT, middle temporal area; RC, retrosplenial cortex.

Fig. 4

DMN detected in each condition and results of the dual regression analysis. The left column shows voxels with z values higher than 2.5, identified as red to yellow in ICA, and the right column shows voxels whose timecourses were significantly affected (P < 0.002) by sedatives or anesthetics in dual regression analysis, and these voxels are identified as blue to light blue.

Fig. 5

Partial correlate coefficients and betweenness centrality among DMN constituent regions. a) Partial correlate coefficients between the regions comprising the DMN, from left to right in the first row: awake, Mida, Dex, and IsoDex; from left to right in the second row: Propo, Alfa, Iso, and Sevo. The lower triangle of the matrix shows the calculated partial correlation coefficients, and the upper triangle shows the absolute difference from the awake condition. The upper left of the matrix shows the correlation within the left hemisphere, the lower right shows the correlation within the right hemisphere, and the others show the correlation between hemispheres. These matrices show the constituent areas of the DMN, respectively, from left or top, PM, IPS, PPA, IPL, Prec, PCC, RC, V2, V3, V6, Subi, ProA, and SuC in left and right hemispheres. The histogram shows the distribution of partial correlation coefficients, and the box located at the upper right of the histgram  shows the expanded distribution of partial correlation coefficients that are higher than the threshold calculated for the awake state. b) Betweenness centrality calculated among regions comprising the DMN. Only the regions where statistically significant difference was observed with both one-way ANOVA and Tukey–Kramer test with P < 0.05 are shown. OFC, orbitofrontal cortex; IPS, intraparietal sulcus; PPA, postal parietal area; PCC, posterior cingulate cortex; V3, third visual cortex.