Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Aug 1.
Published in final edited form as: Neuroimage. 2018 May 5;176:380–389. doi: 10.1016/j.neuroimage.2018.05.009

Development of brain-wide connectivity architecture in awake rats

Zilu Ma 1, Yuncong Ma 1, Nanyin Zhang 1,*
PMCID: PMC6345182  NIHMSID: NIHMS1515328  PMID: 29738909

Abstract

Childhood and adolescence are both critical developmental periods, evidenced by complex neurophysiological changes the brain undergoes and high occurrence rates of neuropsychiatric disorders during these periods. Despite substantial progress in elucidating the developmental trajectories of individual neural circuits, our knowledge of developmental changes of whole-brain connectivity architecture in animals is sparse. To fill this gap, here we longitudinally acquired rsfMRI data in awake rats during five developmental stages from juvenile to adulthood. We found that the maturation timelines of brain circuits were heterogeneous and system specific. Functional connectivity (FC) tended to decrease in subcortical circuits, but increase in cortical circuits during development. In addition, the developing brain exhibited hemispheric functional specialization, evidenced by reduced inter-hemispheric FC between homotopic regions, and lower similarity of region-to-region FC patterns between the two hemispheres. Finally, we showed that whole-brain network development was characterized by reduced clustering (i.e. local communication) but increased integration (distant communication). Taken together, the present study has systematically characterized the development of brain-wide connectivity architecture from juvenile to adulthood in awake rats. It also serves as a critical reference point for understanding circuit- and network-level changes in animal models of brain development-related disorders. Furthermore, FC data during brain development in awake rodents contain high translational value and can shed light onto comparative neuroanatomy.

Keywords: Brain development, Adolescence, Resting-state functional connectivity, Awake rat

Introduction

Childhood and adolescence are phases characterized by rapid social, emotional, and cognitive growth. During these developmental periods, the brain undergoes a complex and dynamic maturational process that contains multiple stages (Andersen, 2003; Paus, 2005; Paus et al., 2008; Spear, 2000), marked by notable neuronal changes such as myelination of axons and synaptic pruning (Huttenlocher, 1979; Huttenlocher et al., 1982; Petanjek et al., 2008). In addition, cell morphology, neural transmitter and receptor density also change drastically during early brain development (Sisk and Foster, 2004; Sisk and Zehr, 2005). Protracted development during childhood and adolescence makes the brain particularly vulnerable to adverse early-life experience (Bremner et al., 1997; Stein et al., 1997; Vythilingam et al., 2002). In fact, neuropsychiatric disorders often emerge during childhood and persist across the lifespan (Insel, 2010; Paus et al., 2008). Therefore, understanding postnatal brain development is crucial for understanding brain function in health and disease.

Both human (Fair et al., 2010; Fair et al., 2008; Power et al., 2010) and animal (Calabrese et al., 2013; Choi et al., 2015) studies have revealed that brain development during childhood and adolescence is spatiotemporally heterogeneous, with different brain regions exhibiting unique time courses of ontogeny. For instance, the measurement of cortical thickness suggests that the sensory cortex matures relatively early in life (Blakemore, 2012; Gogtay et al., 2004), while the development of the prefrontal cortex has late onset and is long lasting throughout the entire adolescent period and into early adulthood (Bourgeois et al., 1994). Accordingly, the maturation of sensory functions is completed at early ages (Antonini and Stryker, 1993; Wiesel and Hubel, 1963) whereas the development of prefrontal cortex-related function such like emotion regulation and cognition continues during early adulthood (Huttenlocher and Dabholkar, 1997).

The effort to characterize developmental trajectories of individual neural circuits has benefited tremendously from various structural and functional neuroimaging approaches (Gogtay et al., 2004). In particular, resting-state functional magnetic resonance imaging (rsfMRI) has emerged as a popular technique for assessing the development of neural circuits/networks in humans (Biswal et al., 1995). This non-invasive imaging method measures large-scale functional connectivity (FC) by quantifying temporal correlations of spontaneous blood-oxygenation-level dependent (BOLD) signal between brain regions. Using this method, Fair and colleagues showed that regions in the default-mode network (DMN) were only sparsely connected in children, but strongly connected in adults (Fair et al., 2008), indicating that childhood and adolescence are critical periods for the development of this key brain network. Also since rsfMRI has a global field of view, it allows the developmental changes of the whole-brain network organization to be studied (Stevens et al., 2009).

Despite that rsfMRI has provided critical insight into the functional development of neural circuits in humans, this knowledge is sparse in animals. In particular, our understanding of the development of whole-brain connectivity architecture in rodents remains limited. A major factor limiting rsfMRI to be applied to investigating brain network development in animals is that most animal imaging experiments rely on anesthesia to immobilize animals, while anesthesia is a major confounder to FC measurement as it disrupts neural and vascular activities (Gao et al., 2016; Hamilton et al., 2017; Liang et al., 2012b). Lack of knowledge in developmental changes of whole-brain connectivity architecture during juvenile and adolescence in animals has highlighted a critical gap in comprehensively understanding animal’s post-natal brain development. It also hinders the characterization of systems-level mechanisms underlying animal models of brain development-related disorders.

To comprehensively trace the developmental trajectories of functional organization of largescale brain circuits, here we longitudinally acquired rsfMRI data in awake rats during five developmental stages: juvenile (P30-P31), early adolescence (P34-P35), adolescence (P41-P42), late adolescence (P48-P49) and adulthood (P70-P90). We discovered that the development of individual neural circuits was spatiotemporally heterogeneous. Besides, the developing brain exhibited hemispheric functional specialization. Finally, we showed that whole-brain network development was characterized by reduced local clustering but increased integration from juvenile into adulthood.

Materials and Methods

Subjects

Male Long Evans rats (n=59) were used in this study. 43 of them underwent longitudinal imaging from the juvenile period to adulthood, while the other 16 rats were scanned only on P41 and used as the control for the potential effects of repeated imaging. Pregnant Long Evans rats obtained from Charles River Laboratory (Wilmington, MA) at embryonic day 15 (E15) upon arrival were singly housed. Food and water were provided ad libitum. A 12h:12h light:dark cycle and constant room temperature were maintained. Offsprings were held in their mother’s cage until weaning. Approval for the study was obtained from the Institutional Animal Care and Use Committee of the Pennsylvania State University.

Acclimation

All animals were acclimated to the scanner environment and scanning noise for seven days to minimize stress and motion during imaging (Gao et al., 2016; Liang et al., 2011). Before set-up, EMLA cream (2.5% lidocaine and 2.5% prilocaine) was applied to nose, jaw, and ear areas to relieve any discomfort that may be induced by the head restrainer. During setup, rats were first briefly anesthetized with 3% isoflurane and placed in a head restrainer with teeth secured over a bite bar, ears secured by two adjustable ear pads, and nose secured by a nose bar. To achieve optimal imaging results, we adjusted the design of the restrainer for each age group by adapting for the animal’s head size and shape. The secured animal was guided into a plexi-glass body tube. After waking up (usually < 10 mins), the animal was placed into a black opaque chamber where the prerecorded sound of different MRI sequences was played. Animals were acclimated in this setup for seven days with increasing exposure time each day up to 60 mins (i.e. 15, 30, 45, 60, 60, 60 and 60 mins, respectively). Previous studies have confirmed that this procedure provides adequate acclimation in animals by diminishing motion and stress during imaging (Liang et al., 2014; Ma et al., 2016) and has been used by several other groups (Becerra et al., 2011; Bergmann et al., 2016; Chang et al., 2016; Ferris et al., 2006; Huang et al., 2011; Martin et al., 2006; Yoshida et al., 2016).

Image Acquisition

rsfMRI data were acquired on a 7 Tesla Bruker scanner (Bruker, Germany). Similar to the acclimation setup procedure, rats were placed in a head restrainer embedded in a birdcage head coil. Anatomical images were first acquired using a high-resolution T1-weighted imaging sequence with the parameters: repetition time (TR) = 1500 ms, echo time (TE) = 8 ms, field of view (FOV) = 3.2 × 3.2cm2, in-plane matrix size = 256 × 256, 1mm slice thickness. Subsequently, rsfMRI scans were collected using the single-shot gradient-echo echo-planar imaging (GE-EPI) sequence with the parameters: TR = 1000 ms, TE = 15 ms, FOV = 3.2 × 3.2cm2, flip angle = 60 degree, in-plane matrix size = 64 × 64, 1mm slice thickness, 600 volumes each scan. Example EPI volumes for all age groups were shown in Supplementary Information (SI Figure 1).

Data Preprocessing

rsfMRI data were preprocessed using the pipeline developed in our lab as described in previous publications (Liang et al., 2012a; Liang et al., 2013). Briefly, each rat was manually registered to a rat brain atlas for its own age group based on Swanson atlas for the adult group (i.e. the P70 group), and the histology atlas of the developing rat brain (Calabrese et al., 2013) for younger groups (i.e. P30, P34, P41 and P48). This procedure was performed using the Medical Image Visualization and Analysis software (MIVA, http://ccni.wpi.edu). The first 10 volumes of each scan were removed to ensure the MR signal to reach steady state. EPI scans were then segmented into 200-volume segments to ensure the same degree of freedom (Zhang et al., 2010). The animal’s motion was estimated using the frame-wise displacement (FD) described in (Ma et al., 2016; Power et al., 2015). Volumes with FD > 0.25mm and their immediate neighboring volumes were scrubbed. Segments with more than 5% of volumes removed were excluded from further analysis. Subsequently, rsfMRI images were motion corrected using SPM12 (University College London, United Kingdom), and spatially smoothed using a Gaussian kernel (FWHM = 1 mm). Nuisance signals from the white matter and ventricles, as well as six motion parameters were regressed out. Finally, data were temporally smoothed using a band-pass filter with the cutoff frequencies of 0.01 and 0.1 Hz.

Data Analysis

Depending on specific questions to answer, the whole rat brain was parcellated into either 67 bilateral or 134 unilateral regions of interest (ROIs) based on the anatomic definition. For each scan, the time course of each ROI was obtained by regionally averaging rsfMRI time courses of all voxels inside the ROI. The FC of each pair of ROIs was derived by calculating the Pearson correlation coefficient between their time courses. Correlation coefficients (i.e. r values) were transformed to Z values using Fisher’s Z transformation. All ROIs were grouped into one of nine brain systems including the sensorimotor cortex, polymodal association cortex, hippocampus, retrohippocampal regions, amygdala complex, striatum, pallidum, thalamus and hypothalamus (Liang et al., 2012b). To examine the overall developmental changes, age-related effects were statistically tested using one-way ANOVA with FC strength of each rat (i.e. z score averaged across segments in one rsfMRI scan) at each developmental stage as the independent variable. To test FC changes between successive ages, two sample t-test was applied using the linear mixed-effect model with the random effect of rats and the fixed effect of z score of each segment. The statistical significance was set at p < 0.05. False discovery rate correction (Genovese et al., 2002) was further applied when multiple comparisons were involved.

The graph theory analysis was applied to examine the developmental changes of the topological organization of the whole-brain network. For each age group, the averaged clustering coefficient of the whole-brain network graph was quantified as function of graph density in the range between 0.05 to 0.35 with the step size of 0.05. At each density, the averaged clustering coefficient was normalized to a reference network generated by randomizing the connectivity between ROIs but maintaining the same degree distribution. The process was repeated for 1000 times. These 1000 normalized clustering coefficients obtained were then averaged, and reported as the final averaged clustering coefficient at the given density.

Results

In the present study, we systematically characterized the developmental changes of whole-brain connectivity architecture from juvenile into adulthood using FC in awake rats. rsfMRI data were longitudinally collected during five developmental stages: juvenile (P30-P31), early adolescence (P34P35), adolescence (P41-P42), late adolescence (P48-P49) and adulthood (P70-P90).

Motion level was consistent among age groups.

Motion is a critical factor that can affect the FC quantification (Power et al., 2012; Power et al., 2015), and any difference in motion levels among separate age groups can potentially induce artifactual difference in FC. Therefore, to rule out this confounding factor, we first examined the motion level in all five age groups. There was no significant difference in motion level across all age groups, measured by six motion parameters as well as FD estimated from rsfMRI data (Fig. 1A and B, one-way ANOVA, P > 0.05 for three translational and three rotational motion parameters, as well as FD). In addition, the level of motion scrubbing was consistent across age groups, with < 1% difference in the portion of volumes removed across all age groups (Fig. 1C). Consistent motion levels among age groups were probably due to the stringent motion control procedures in our image preprocessing (see Data Preprocessing), and the design of the restrainer optimized for each age group (see Methods). Furthermore, the average FD (~0.15 mm, Fig. 1B) was considerably smaller than the voxel size (0.5 mm), also indicating that motion was well controlled in this experiment.

Figure 1. Motion levels in different age groups.

Figure 1.

Open in a new tab

(A) Averaged motion parameters. Left column: translational movement (μm). Right column: rotational movement (in rad). Bars: STD. One-way ANOVA, p > 0.05 for all 6 parameters across age groups. (B) Averaged FD of rsfMRI data for different age groups. Bars: STD. (C) Percentage of volumes scrubbed during image preprocessing. Bars: STD.

General connectivity patterns remained stable, but specific connections could change significantly during development.

After ruling out the potential impact of motion, we examined the brain-wide connectivity architecture during development. The rat brain was parcellated into 67 bilateral ROIs. The overall patterns of FC between these ROIs were consistent across separate developmental stages. Fig. 2A showed similar ROI-to-ROI connectivity matrices across all age groups, reflected by high spatial Pearson correlations between them (mean ± std = 0.76 ± 0.06). These results suggest that the overall brain connectivity structure was already established before juvenile. They also demonstrate consistent data quality across all age groups.

Figure 2. Whole-brain connectivity patterns at different developmental stages.

Figure 2.

Open in a new tab

(A) Averaged connectivity matrix between 67 bilateral ROIs at postnatal 31, 35, 41, 49 and 70 days, respectively. ROIs are organized into 9 anatomical brain systems. Red: sensorimotor cortex; orange: polymodal association cortex; blue: retrohippocampal regions; green: hippocampus; black: amygdala complex; purple: striatum; pink: pallidum; brown: thalamus; gray: hypothalamus; yellow: all other ROIs. (B) Connections exhibiting significantly changed FC during development (one-way ANOVA, p < 0.05 after FDR correction). (C) Connections showing significant FC changes, overlaid on the rat brain (sagittal view) and organized into 9 brain systems. The node size is proportional to the percentage of the total number of significantly changed within-system connections relative to the total number of all possible within-system connections, and the edge width is proportional to the percentage of the total number of significantly changed connections between two systems relative to the total number of all possible between-system connections. Top: connections showing FC increase; bottom: connections showing FC decrease during development.

Despite a relatively stable whole-brain connectivity structure, individual circuits exhibited significant changes during development. Fig. 2B showed connections that displayed significant FC changes across five developmental stages (293 connections, One-way ANOVA, p < 0.05 after FDR correction). These connections were overlaid on a glass rat brain (sagittal view, Fig. 2C), displayed based on the brain system they belonged to. To preserve the quantitative information, the node size was proportional to the total number of significantly changed within-system connections normalized to the total number of all possible within-system connections, and the edge width was proportional to the total number of significantly changed connections between two systems normalized to the total number of all possible between-system connections. All subcortical systems (i.e. hippocampus, retrohippocampal regions, amygdala complex, striatum, pallidum, thalamus and hypothalamus) exhibited decreased within-system FC during development (i.e. larger node sizes in the negative change map than the positive change map), but the sensorimotor and polymodal association cortical systems displayed increased FC during development. In addition, ortico-cortical connections and cortico-subcortical connections generally showed FC increases across brain development periods (i.e. thicker edge in the positive change map than the negative change map).

Neural circuit development was more pronounced during juvenile and adolescence

We further examined FC changes during transitions between successive developmental stages in these 293 connections (Fig. 2B). Fig. 3A showed connections with significant FC changes between every two successive development stages (i.e. P31 - P35, P35 - P41, P41 - P49 and P49 – P70), respectively (two-sample t-tests, linear mixed model, P<0.05, FDR corrected), which were overlaid on a glass rat brain (sagittal view, Fig. 3B). These matrices demonstrated that FC changes were more pronounced during transitions from juvenile to late adolescence, suggesting that juvenile to adolescence are critical periods for neural circuit development. In addition, the spatial patterns of these significantly changed connections were considerably different between these transitions, which indicates that different circuits might have distinct maturation timelines. Finally, neural circuit development stabilized from late adolescence to adulthood, evidenced by few connections exhibiting significant FC changes during this transition.

Figure 3. FC changes during transitions between successive developmental stages.

Figure 3.

Open in a new tab

(A) Connections with significant FC change between successive development stages (P31–35, P35–41, P41–49 and P49–70, respectively, two-sample t-test, linear mixed model, p < 0.05, FDR corrected). Color indicates the value of FC change. (B) Connections (shown in A) overlaid on the rat brain, organized into the same 9 brain systems shown in Fig. 2. Left column: connections showing FC increase, right column: connections showing FC decrease.

Individual neural circuits displayed distinct developmental trajectories

To investigate the patterns of developmental changes of FC for individual circuits, we regressed the FC strength against age (in postnatal day) for each connection, and used the slope of regression as a measure of FC change during development (Fig. 4A). Our data showed strong bias in developmental changes across different anatomical systems. Specifically, significant FC increases (i.e. positive slopes) were found in cortico-cortical connections involving regions in sensorimotor (SM) and polymodal-association (PM) cortices, as well as cortico-subcortical connections. By contrast, significant FC decreases (i.e. negative slopes) were observed in connections in subcortical (SC) systems including retrohippocampal, hippocampal, thalamic, striatum, pallidum and hypothalamic regions, with some exceptions in the hippocampus. In addition, the extent of FC strength changes, measured by the magnitude of the slope, varied considerably across connections.

Figure 4. FC trajectories as a function of age for individual connections.

Figure 4.

Open in a new tab

(A) Connections exhibiting significant FC strength change across age groups. Color indicates the slope value when regressing FC strength against age (postnatal days). Only connections showing significant FC changes (two sample t-test, linear mixed model, p<0.05, FDR corrected) with the slope >0.004 were shown. (B) Developmental trajectories of representative connections in each category. GU: gustatory area; TT: tenia tecta; SSp: primary somatosensory area; AUD: auditory area; PL: prelimbic area; MOp: primary somatomotor area; ACA: anterior cingulate area; VIS: visual area; PTL: parietal region; Tev: ventral temporal association area; RSP: retrosplenial area; ORB: orbital area; IL: infralimbic area; ECT: ectorhinal area; DG: dentate gyrus; ATN: anterior nuclei, dorsal thalamus; LAT: lateral nuclei, dorsal thalamus; CP: caudoputamen; AI: agranular insular area; SI: substantia innominate; MED: medial nuclei, dorsal thalamus; LZ: lateral zone of hypothalamus; PVZ: periventricular zone of hypothalamus; BST: bed nuclei stria terminalis; ZI: zona incerta; FS: striatal fundus.

In addition to the sign of developmental FC changes, individual circuits exhibited distinct time courses of ontogeny. Fig 4 plotted the developmental trajectories of FC strength in 4 representative connections for each category (SM-SM, SM-PA, PA-PA, SM-SC, PA-SC and SC-SC connections). FC development in the thalamus and sensori-motor regions plateaued around juvenile/adolescent periods (e.g. SM-SM and SM-SC connections), whereas connections involving frontal regions had late onset and continued to develop into adulthood (e.g. PA-PA connections). To further examine the lagged and prolonged developmental trajectories associated with high-tier polymodal association regions (e.g. frontal regions), we mapped the default-mode network (DMN) in all age groups (SI Fig. 2), generated using the seed-based correlational analysis with the posterior cingulate cortex (PCC) as the seed. The DMN showed virtual no connectivity with the prefrontal cortex during the juvenile and early adolescent periods (P31 and P35). However, this connectivity started to display during adolescence (P41), strengthened during late adolescence/early adulthood (P49), and full matured during adulthood (P70).

The rodent brain exhibited hemispheric specialization during development

It has been shown that inter-hemispheric connectivity is monotonically reduced throughout the development from childhood to adulthood in humans (Anderson et al., 2011). To investigate whether a similar developmental pattern exists in rodents, we measured cross-hemispheric FC at all five ages in our rats. Each of 67 bilateral ROIs used above was further split to the left and right sides, providing 134 unilateral ROIs. Fig. 5A showed between-ROI connectivity matrices for all five age groups. Similar to the results shown in Fig. 2A, the general brain connectivity architecture showed high consistency between age groups (mean spatial correlation ± std = 0.75 ± 0.05), again indicating the whole-brain connectivity structure was already established before juvenile. These FC matrices also showed that the brain connectivity pattern was relatively bilaterally symmetric at all five ages.

Figure 5. Connectivity patterns between unilateral ROIs at different developmental stages.

Figure 5.

Open in a new tab

(A) Connectivity matrix between 134 unilateral ROIs at postnatal 31, 35, 41, 49 and 70 days, respectively. ROIs are organized into the same 9 anatomical brain systems shown in Fig. 2 (first half: left hemisphere, second half: right hemisphere). (B) Left: averaged inter-hemispheric homotopic FC value as a function of age. Bar: SEM. Mid: spatial correlation of within-hemispheric ROI-to-ROI connectivity patterns between the two hemispheres (i.e. two dashed white triangles) as a function of age. Right, spatial correlation of cross-hemispheric ROI-to-ROI connectivity patterns between the two hemispheres (i.e. two dashed white triangles) as a function of age. (C) Age-related changes in inter-hemispheric homotopic FC in individual regions. The strength of inter-hemispheric homotopic FC of each individual region is displayed for all ages. The 6th column shows the slope in the regression of the FC strength against age.

Interestingly, like humans, interhemispheric FC between homotopic ROIs monotonically decreased during development in rodents. Fig. 5B (left) showed a monotonic decrease in averaged interhemispheric homotopic FC across all ROIs. To closely examine the spatial distribution of age-related changes in interhemispheric FC, we displayed the strength of interhemispheric homotopic FC for individual ROIs across the rat brain for all ages (columns 1–5 in Fig. 5C), as well as the regression slope of the interhemispheric homotopic FC against age (the last columns in Fig. 5C). The vast majority of ROIs displayed negative slopes, indicating a reduction of interhemispheric FC during development. Only a few ROIs showed increasing interhemispheric FC during development, including the primary and secondary somatosensory/somatomotor areas as well as the cortical amygdalar nucleus.

Decreasing interhemispheric FC during development might suggest a more specialized functional organization with each hemisphere. To test this notion, between the left and right hemispheres we compared the similarity of their within-hemisphere connectivity patterns at all ages, defined by the spatial correlation of the two dashed triangles in the whole-brain connectivity matrix shown in Fig. 5B (middle). This similarity of within-hemispheric ROI-to-ROI connectivity patterns decreased monotonically over age. Furthermore, we observed that the similarity of cross-hemispheric ROI-to-ROI connectivity patterns between the two hemispheres (the two dashed triangles in Fig. 5B, right) also decreased over ages. Taken together, these results demonstrated that the FC strength, as well as the similarity of the connectivity patterns between the two hemispheres monotonically decreased during development, suggesting that both hemispheres are functionally more specialized as the brain matures.

Brain development was characterized by reduced local clustering but increased distant integration

Increased between-system FC and decreased within-system FC during development shown in Fig. 2 indicate that developmental changes of neural circuitries might be characterized by enhanced network integration (increased long-distance FC) and reduced local clustering (decreased short-distance FC) in the awake rat brain. To test this hypothesis, we plotted the distribution of FC in significantly changed connections (Fig. 2B) as a function of their physical distance (Fig. 6A). Our data clearly showed that reduced FC for short-distance connections but increased FC for long-distance connections. Fig. 6B displayed short- (top, physical distance ≤ 4mm) and long-distance (bottom, physical distance ≥ 6mm) connections with significant FC changes (One-way ANOVA, p < 0.05, FDR corrected). These data demonstrate that the vast majority of short-distance connections exhibited reduced FC changes (i.e. negative slopes), whereas long-distance connections predominantly exhibited increased FC changes during development (i.e. positive slopes). This feature was also corroborated by the decreased topological measure of averaged clustering coefficient (Fig. 6C) during brain development. These results collectively indicate that the rat neural network development is characterized by reduced local clustering but increased distant integration.

Figure 6. Relationship between FC change and physical distance.

Figure 6.

Open in a new tab

(A) Connections with significant slopes between FC and age, plotted against their physical distance. (B) Short- vs long-distance connections separately displayed in a glass rat brain (axial, sagittal and coronal views). Blue connections indicate reduced FC over age (i.e. negative slopes). Red connections indicate increased FC over age (i.e. positive slopes). (C) Normalized average clustering coefficient for all age groups.

Controlling for the potential effects of repeated imaging

To rule out the potential effects of repeated fMRI experiments during early developmental stages on subsequent imaging data, we performed a control imaging experiment on rats that were only scanned on P41 after the acclimation (i.e. no imaging was conducted on P31 and P35, n = 16). All data were processed using the same pipeline described in Methods. We compared the whole-brain connectivity matrix to the P41 group that also underwent imaging on P31 and P35 (Fig. 2A), and found very high spatial Pearson correlation between the two connectivity matrices (Fig. 7A, r = 0.86). In addition, very few connections showed significant FC difference between these two groups (Fig. 7B, p < 0.05, FDR corrected). Highly consistent data between the two P41 groups with and without prior repeated imaging demonstrated that earlier scans on P31 and P35 had minimal effects on subsequent imaging data. Importantly, the negligible difference between these two P41 groups was appreciably less than the number of connections showing significant FC changes between P31 and P35, as well as between P35 and P41 (Fig. 3), indicating that these changes reported were not attributed to repeated imaging.

Figure 7. Effect of repeated imaging on FC.

Figure 7.

Open in a new tab

(A) Whole-brain connectivity matrices for the P41 group with prior scans on P31 and P35 (left), and without any prior scans (right). (B) Connections exhibiting significant FC difference between the two P41 groups (two sample t-test, linear mixed model, p<0.05, FDR corrected). ROIs are organized by 9 anatomical brain systems, which are coded on the left and top of matrices. The color code is the same as that defined in Figure 2.

Discussion

For the purpose of elucidating developmental changes of brain-wide connectivity architecture in rodents, here we employed the awake animal rs-fMRI approach to trace FC changes in individual neural circuits from juvenile to adulthood. We discovered that although the general architecture of FC matrices remained stable across ages (Fig. 2), individual neural circuits could exhibit significant changes during development (Figs. 3 and 4). We also found that different neural circuits displayed distinct developmental trajectories (Fig. 4). Furthermore, our data indicate that the rat brain development is characterized by increased hemispheric functional specialization (Fig. 5), decreased local clustering and increased distant integration (Fig. 6).

The significance of the present study can be found in two folds. First, to date, there is lack of network-level knowledge of brain development in rodents, particularly in the measure of FC of neural circuits across the whole brain. A major obstacle is the confounding effects of anesthesia on FC measurement in most animal imaging experiments (Hamilton et al., 2017; Liang et al., 2012b; Liang et al., 2015a; Liang et al., 2015b; Ma et al., 2017; Smith et al., 2016). This knowledge gap has been filled by the present study, which employed the awake animal rsfMRI approach established in our lab (Gao et al., 2016; Ma et al., 2016). FC data collected in awake rodents also contain high translational value and can shed light on comparative neuroanatomy (see one example in SI Fig. 2). Second, establishing normal maturation timelines for different brain circuits/networks can be used as developmental markers for further investigations of brain development-related disorders in animal models. Numerous studies have suggested that early life adversity experienced during crucial developmental periods from childhood to adulthood might have significant influences on the courses of neural maturation, and thus increase the potential of psychopathology and exacerbate the consequence of any adverse experiences in the future. For instance, children with ADHD display approximately a 3-year delay in the cortical maturation, especially in prefrontal regions but slight earlier maturation in the primary motor cortex (Shaw et al., 2010). Therefore, data of the present study are complementary to molecular and behavioral data on brain development and can serve as a reference for fully understanding animal models of development-related brain disorders.

Neural circuit development was spatiotemporally heterogeneous and system specific

Our data show that the pattern of neural circuit development was distinct between cortical and subcortical systems. The FC strength between cortical regions generally increased, whereas subcortical connections were weakened during development. More specifically, significant increases in FC were found in connections within and between the sensorimotor and polymodal-association cortices, as well as in thalamo-cortical connections. By contrast, significant decreases of FC were observed in connections involving hippocampal, retrohippocampal, thalamic, striatal and hypothalamic regions. These data are consistent with the human literature reporting stronger cortico-cortical and thalamocortical but weaker subcortical connectivity in young adults compared to children (Fair et al., 2010; Supekar et al., 2009). Collectively, these results suggest that neural circuit development is system specific and might be characterized by a rebalance between cortical and subcortical circuits.

In addition to system-specific FC changes, individual brain regions exhibited distinct developmental trajectories. Our data showed that regions reported to mature early, such as subcortical regions of the thalamus and hippocampus, as well as sensorimotor cortices plateaued around adolescence (P41~P49) or earlier (Fig. 4), whereas FC between late developing regions such as polymodal association cortices continued to change into adulthood (Fig. 4; SI Fig. 2). These results were similar to the rodent literature reports that the volume of anterior commissure showed constant increase until postnatal 70 days, while hippocampus and amygdala plateaued during juvenile/adolescence (Calabrese et al., 2013). In humans, the cortical thickness in sensory and motor regions have been shown to mature the earliest (Blakemore, 2012), followed by the frontal and parietal lobes, which peak around early adolescence (Shaw et al., 2010). The maturation in the superior temporal cortex appeared to be the latest (Gogtay et al., 2004). In addition, human fMRI study showed that the frontal lobe connectivity was relatively weak in childhood, but was significantly strengthened by adulthood (Fair et al., 2008). In particular, the frontal cortex in the DMN had sparse connections with the posterior brain (e.g. PCC), but such connectivity was significantly stronger in adults (Fair et al., 2008). This age-related difference in DMN connectivity patterns was also observed in our rat data (SI Fig. 2). Taken together, these results indicate that in both humans and rodents the maturation timelines are heterogeneous across neural regions/circuits.

The functional significance of heterogeneous ontogeny of separate neural circuits remains not entirely clear, but is likely related to the biological demand for the brain function. A general principle of neural development is that phylogenetically “newer” brain functions (e.g. cognition) are associated with brain regions/circuits with longer developmental trajectories. As shown by literature studies and our data, early maturation of sensorimotor cortices might reflect the importance of sensorimotor functions during early stages of life. By contrast, late and long-lasting development of the frontal cortex allows the environment to interact with and shape the maturation of this region, and thereby influence cognition-related functions. Interestingly, our previous study showed that the genetic influences on sensorimotor networks are stronger than cognition-related networks in humans (Fu et al., 2015), suggesting a possible evolutionary role of heterogeneous developmental trajectories of different regions/circuits.

Brain development exhibited hemispheric functional specialization

An interesting finding of the present study is decreased interhemispheric FC between homotopic ROIs. This data is consistent with a previous human rsfMRI study, in which healthy subjects exhibited a decrease in interhemispheric connectivity across an age range of 8 to 42 years, whereas this reduction trend was not as pronounced in autism patients (Anderson et al., 2011). These results suggest that decreasing interhemispheric FC might be a characteristic developmental feature in the mammalian brain, and can be a marker for brain development-related disorders.

Decreased interhemispheric FC might suggest functional specialization of the two hemispheres. To further examine this notion, for both hemispheres we calculated their within-hemisphere ROI-to-ROI FC patterns and cross-hemisphere ROI-to-ROI FC patterns. We then compared the spatial similarities of these patterns between the two hemispheres. Our data showed that both within- and cross-hemisphere connectivity patterns became more dissimilar between the two hemispheres during development, which suggests that nodes within each hemisphere become more specialized during development. This result is consistent with previous research reporting changing brain asymmetry during development (Geschwind and Galaburda, 1985a, b, c). It has been shown that the small-world characteristics were similar between two hemispheres in neonates. However, prior to adulthood, both hemispheres were organized in a more efficient but asymmetric manner (Ratnarajah et al., 2013). For example, the left hemisphere was associated with language and memory formation whereas the right hemisphere was involved in emotional processes (Geschwind and Galaburda, 1985a, b, c; Vallortigara et al., 1999). These results collectively indicate that reduced interhemispheric communication is accompanied by functional specialization within each hemisphere during brain development, and this feature is characteristic to both rodents and humans.

Brain development was characterized by reduced local clustering and increased distant integration

Our data showed a strong trend of decreased short-distance FC but increased long-distance FC during development. This result is consistent with human studies, in which a similar relationship between the FC strength and ROI wiring distance was demonstrated (Fair et al., 2008; Stevens et al., 2009; Supekar et al., 2009). Fair and colleagues showed that in humans, distributed regions in the default-mode network were only sparsely connected at 7–9 years old, whereas the connectivity between these regions was substantially stronger by adulthood with more long-range connections (Fair et al., 2008). In addition, our topological measures of the whole brain network showed decreased averaged clustering coefficient (Fig. 6C) during brain development. This result is also consistent with the report that there is an age-dependent increase in communication efficiency in distributed networks (Stevens et al., 2009). Importantly, these results provide strong evidence supporting a long-standing hypothesis in the literature that the brain development is accompanied by changed local segregation and distant integration (Fair et al., 2009; Fair et al., 2007). However, this notion has been challenged, as difference in motion levels between young and adult subject groups can cause bias in FC quantification, which may lead to artifactual decrease in short-distance FC and increase in long-distance FC (Fair et al., 2007). Notably, our results were not attributed to motion differences during scans, as the motion level was consistent across all age groups (Fig. 1).

Decreased short-distance but increased long-distance FC during development can be explained by two possible neurobiological processes: synaptic pruning and myelination. Systematic synapse pruning during late adolescence can lead to weakening of local connections and formation of more specialized regions. In parallel, although by infant age, long-range anatomical connectivity is adult-like, myelination continues into young adulthood. Increased efficiency of signal propagation with the myelin sheath may support increased functional integration and energy efficiency for long-distance connections. Indeed, in previous rodent positron emission tomography (PET) studies, connections in retrosplenial and medial prefrontal cortex showed increase in energy efficiency over development. This result suggests that although distant connections already physically exist at early age, there is a redistribution of energy consumption during brain development, which contributes to the functional maturation of long-distance connectivity.

Potential Pitfalls

There are a few potential pitfalls. First, brain sizes were different at different ages. As the brain volume changed during development, the same ROIs could contain different numbers of voxels at different ages, and thus there could be more partial volume effect for smaller brains. Second, the hemodynamic response might be different for different ages due to the developmental changes of the hemodynamic response function (Colonnese et al., 2008; Kozberg et al., 2013). Different hemodynamic responses at different ages could affect the quantification of FC. However, the earliest age selected in this study was postnatal 31 days. Studies have shown that in rats the local field potential (LFP) and BOLD responses were already adult like by P30 (Colonnese et al., 2008). Therefore, our results are unlikely affected by different BOLD responses during the developmental period we studied. Third, it is likely that larger FC changes during earlier developmental stages observed resulted from the priming effects of prior scans, particularly considering that scanning days were closer during earlier stages. Another possibility is cumulative desensitization to imaging-related stress during longitudinal scans. To rule out these possibilities, we performed a control experiment on a separate group of rats that were only scanned on P41, and the data in this control group were highly consistent with the P41 data collected with prior imaging on P31 and P35 (Fig. 7). These data demonstrated that prior scans on P31 and P35 had minimal effects on the imaging data on P41. Importantly, the negligible difference between these two P41 groups was appreciably less than the number of connections showing significant FC changes between P31 and P35, as well as between P35 and P41, indicating that these FC changes were not due to priming effects and/or desensitization to stress for longitudinal scans, but indeed represent development-related neural circuitry changes. Furthermore, our previously reported data also showed minimal effects of multiple times of imaging on the stress response in animals (Gao et al., 2016). Specifically, we compared the anxiety level measured by the elevated plus maze (EPM) between animals with no manipulation (group 1), animals with four-time repeated imaging (group 2), animals exposed to predator odor stress but did not undergo any imaging (group 3). The data showed no significant difference between the no-manipulation group and the repeated imaging group, but the predator odor exposed group exhibited significantly heightened anxiety level (Gao et al., 2016), which also demonstrate that multiple repeated imaging had minimal effects on the stress level of rats. Taken together, these results indicate that priming effects and/or desensitization to stress during longitudinal scans are not dominant factors attributing to the FC changes between different developmental stages we observed. This result is actually not surprising considering the fact that before the first imaging session, all animals were acclimated to the scanning environment 7 days during the acclimation procedure. Therefore, any potential priming effects and stress responses should have already been diminished before the first scan (see (Gao et al., 2016) for more detailed discussion).

Conclusion

In the present study, we systematically characterized neural circuit development from juvenile to adulthood using FC in awake rodents. Our data showed heterogeneous developmental trajectories across brain systems. In addition, the developing brain exhibited hemispheric specialization, local segregation and distant integration. These data have provided a critical reference point for further understanding the emergence, course, and severity of brain development-related disorders in animal models.

Supplementary Material

1

Acknowledgments

The present study was supported by National Institute of Neurological Disorders and Stroke (R01NS085200, PI: Nanyin Zhang, PhD) and National Institute of Mental Health (R01MH098003 and RF1MH114224, PI: Nanyin Zhang, PhD).

Footnotes

Conflict of interest: none.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References:

  1. Andersen SL, 2003. Trajectories of brain development: point of vulnerability or window of opportunity? Neurosci Biobehav Rev 27, 3–18. [DOI] [PubMed] [Google Scholar]
  2. Anderson JS, Druzgal TJ, Froehlich A, DuBray MB, Lange N, Alexander AL, Abildskov T, Nielsen JA, Cariello AN, Cooperrider JR, Bigler ED, Lainhart JE, 2011. Decreased interhemispheric functional connectivity in autism. Cereb Cortex 21, 1134–1146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Antonini A, Stryker MP, 1993. Rapid remodeling of axonal arbors in the visual cortex. Science 260, 1819–1821. [DOI] [PubMed] [Google Scholar]
  4. Becerra L, Pendse G, Chang PC, Bishop J, Borsook D, 2011. Robust reproducible resting state networks in the awake rodent brain. PLoS One 6, e25701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bergmann E, Zur G, Bershadsky G, Kahn I, 2016. The Organization of Mouse and Human Cortico-Hippocampal Networks Estimated by Intrinsic Functional Connectivity. Cereb Cortex 26, 4497–4512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Biswal B, Yetkin FZ, Haughton VM, Hyde JS, 1995. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn Reson Med 34, 537–541. [DOI] [PubMed] [Google Scholar]
  7. Blakemore SJ, 2012. Imaging brain development: the adolescent brain. Neuroimage 61, 397–406. [DOI] [PubMed] [Google Scholar]
  8. Bourgeois JP, Goldman-Rakic PS, Rakic P, 1994. Synaptogenesis in the prefrontal cortex of rhesus monkeys. Cereb Cortex 4, 78–96. [DOI] [PubMed] [Google Scholar]
  9. Bremner JD, Randall P, Vermetten E, Staib L, Bronen RA, Mazure C, Capelli S, McCarthy G, Innis RB, Charney DS, 1997. Magnetic resonance imaging-based measurement of hippocampal volume in posttraumatic stress disorder related to childhood physical and sexual abuse--a preliminary report. Biol Psychiatry 41, 23–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Calabrese E, Badea A, Watson C, Johnson GA, 2013. A quantitative magnetic resonance histology atlas of postnatal rat brain development with regional estimates of growth and variability. Neuroimage 71, 196–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chang PC, Procissi D, Bao Q, Centeno MV, Baria A, Apkarian AV, 2016. Novel method for functional brain imaging in awake minimally restrained rats. J Neurophysiol 116, 61–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Choi H, Choi Y, Kim KW, Kang H, Hwang DW, Kim EE, Chung JK, Lee DS, 2015. Maturation of metabolic connectivity of the adolescent rat brain. Elife 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Colonnese MT, Phillips MA, Constantine-Paton M, Kaila K, Jasanoff A, 2008. Development of hemodynamic responses and functional connectivity in rat somatosensory cortex. Nat Neurosci 11, 72–79. [DOI] [PubMed] [Google Scholar]
  14. Fair DA, Bathula D, Mills KL, Dias TG, Blythe MS, Zhang D, Snyder AZ, Raichle ME, Stevens AA, Nigg JT, Nagel BJ, 2010. Maturing thalamocortical functional connectivity across development. Front Syst Neurosci 4, 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fair DA, Cohen AL, Dosenbach NU, Church JA, Miezin FM, Barch DM, Raichle ME, Petersen SE, Schlaggar BL, 2008. The maturing architecture of the brain’s default network. Proc Natl Acad Sci U S A 105, 4028–4032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fair DA, Cohen AL, Power JD, Dosenbach NU, Church JA, Miezin FM, Schlaggar BL, Petersen SE, 2009. Functional brain networks develop from a “local to distributed” organization. PLoS Comput Biol 5, e1000381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fair DA, Dosenbach NU, Church JA, Cohen AL, Brahmbhatt S, Miezin FM, Barch DM, Raichle ME, Petersen SE, Schlaggar BL, 2007. Development of distinct control networks through segregation and integration. Proc Natl Acad Sci U S A 104, 13507–13512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ferris CF, Febo M, Luo F, Schmidt K, Brevard M, Harder JA, Kulkarni P, Messenger T, King JA, 2006. Functional magnetic resonance imaging in conscious animals: a new tool in behavioural neuroscience research. J Neuroendocrinol 18, 307–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fu Y, Ma Z, Hamilton C, Liang Z, Hou X, Ma X, Hu X, He Q, Deng W, Wang Y, Zhao L, Meng H, Li T, Zhang N, 2015. Genetic influences on resting-state functional networks: A twin study. Hum Brain Mapp. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gao YR, Ma Y, Zhang Q, Winder AT, Liang Z, Antinori L, Drew PJ, Zhang N, 2016. Time to wake up: Studying neurovascular coupling and brain-wide circuit function in the un-anesthetized animal. Neuroimage. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Genovese CR, Lazar NA, Nichols T, 2002. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage 15, 870–878. [DOI] [PubMed] [Google Scholar]
  22. Geschwind N, Galaburda AM, 1985a. Cerebral lateralization. Biological mechanisms, associations, and pathology: I. A hypothesis and a program for research. Arch Neurol 42, 428–459. [DOI] [PubMed] [Google Scholar]
  23. Geschwind N, Galaburda AM, 1985b. Cerebral lateralization. Biological mechanisms, associations, and pathology: II. A hypothesis and a program for research. Arch Neurol 42, 521–552. [DOI] [PubMed] [Google Scholar]
  24. Geschwind N, Galaburda AM, 1985c. Cerebral lateralization. Biological mechanisms, associations, and pathology: III. A hypothesis and a program for research. Arch Neurol 42, 634–654. [DOI] [PubMed] [Google Scholar]
  25. Gogtay N, Giedd JN, Lusk L, Hayashi KM, Greenstein D, Vaituzis AC, Nugent TF 3rd, Herman DH, Clasen LS, Toga AW, Rapoport JL, Thompson PM, 2004. Dynamic mapping of human cortical development during childhood through early adulthood. Proc Natl Acad Sci U S A 101, 8174–8179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hamilton C, Ma Y, Zhang N, 2017. Global reduction of information exchange during anesthetic-induced unconsciousness. Brain Struct Funct. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Huang W, Heffernan ME, Li Z, Zhang N, Overstreet DH, King JA, 2011. Fear induced neuronal alterations in a genetic model of depression: an fMRI study on awake animals. Neurosci Lett 489, 74–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Huttenlocher PR, 1979. Synaptic density in human frontal cortex - developmental changes and effects of aging. Brain Res 163, 195–205. [DOI] [PubMed] [Google Scholar]
  29. Huttenlocher PR, Dabholkar AS, 1997. Regional differences in synaptogenesis in human cerebral cortex. J Comp Neurol 387, 167–178. [DOI] [PubMed] [Google Scholar]
  30. Huttenlocher PR, de Courten C, Garey LJ, Van der Loos H, 1982. Synaptogenesis in human visual cortex--evidence for synapse elimination during normal development. Neurosci Lett 33, 247–252. [DOI] [PubMed] [Google Scholar]
  31. Insel TR, 2010. Rethinking schizophrenia. Nature 468, 187–193. [DOI] [PubMed] [Google Scholar]
  32. Kozberg MG, Chen BR, DeLeo SE, Bouchard MB, Hillman EM, 2013. Resolving the transition from negative to positive blood oxygen level-dependent responses in the developing brain. Proc Natl Acad Sci U S A 110, 4380–4385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Liang Z, King J, Zhang N, 2011. Uncovering intrinsic connectional architecture of functional networks in awake rat brain. J Neurosci 31, 3776–3783. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Liang Z, King J, Zhang N, 2012a. Anticorrelated resting-state functional connectivity in awake rat brain. Neuroimage 59, 1190–1199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Liang Z, King J, Zhang N, 2012b. Intrinsic organization of the anesthetized brain. J Neurosci 32, 10183–10191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Liang Z, King J, Zhang N, 2014. Neuroplasticity to a single-episode traumatic stress revealed by resting-state fMRI in awake rats. Neuroimage 103, 485–491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Liang Z, Li T, King J, Zhang N, 2013. Mapping thalamocortical networks in rat brain using resting-state functional connectivity. Neuroimage 83, 237–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Liang Z, Liu X, Zhang N, 2015a. Dynamic resting state functional connectivity in awake and anesthetized rodents. Neuroimage 104, 89–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Liang Z, Watson GD, Alloway KD, Lee G, Neuberger T, Zhang N, 2015b. Mapping the functional network of medial prefrontal cortex by combining optogenetics and fMRI in awake rats. Neuroimage 117, 114–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ma Y, Hamilton C, Zhang N, 2017. Dynamic Connectivity Patterns in Conscious and Unconscious Brain. Brain Connect 7, 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Ma Z, Perez P, Ma Z, Liu Y, Hamilton C, Liang Z, Zhang N, 2016. Functional atlas of the awake rat brain: A neuroimaging study of rat brain specialization and integration. Neuroimage. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Martin C, Martindale J, Berwick J, Mayhew J, 2006. Investigating neural-hemodynamic coupling and the hemodynamic response function in the awake rat. Neuroimage 32, 33–48. [DOI] [PubMed] [Google Scholar]
  43. Paus T, 2005. Mapping brain maturation and cognitive development during adolescence. Trends Cogn Sci 9, 60–68. [DOI] [PubMed] [Google Scholar]
  44. Paus T, Keshavan M, Giedd JN, 2008. Why do many psychiatric disorders emerge during adolescence? Nat Rev Neurosci 9, 947–957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Petanjek Z, Judas M, Kostovic I, Uylings HB, 2008. Lifespan alterations of basal dendritic trees of pyramidal neurons in the human prefrontal cortex: a layer-specific pattern. Cereb Cortex 18, 915–929. [DOI] [PubMed] [Google Scholar]
  46. Power JD, Barnes KA, Snyder AZ, Schlaggar BL, Petersen SE, 2012. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage 59, 2142–2154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Power JD, Fair DA, Schlaggar BL, Petersen SE, 2010. The development of human functional brain networks. Neuron 67, 735–748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Power JD, Schlaggar BL, Petersen SE, 2015. Recent progress and outstanding issues in motion correction in resting state fMRI. Neuroimage 105, 536–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Ratnarajah N, Rifkin-Graboi A, Fortier MV, Chong YS, Kwek K, Saw SM, Godfrey KM, Gluckman PD, Meaney MJ, Qiu A, 2013. Structural connectivity asymmetry in the neonatal brain. Neuroimage 75, 187–194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Shaw P, Gogtay N, Rapoport J, 2010. Childhood psychiatric disorders as anomalies in neurodevelopmental trajectories. Hum Brain Mapp 31, 917–925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sisk CL, Foster DL, 2004. The neural basis of puberty and adolescence. Nat Neurosci 7, 1040–1047. [DOI] [PubMed] [Google Scholar]
  52. Sisk CL, Zehr JL, 2005. Pubertal hormones organize the adolescent brain and behavior. Front Neuroendocrinol 26, 163–174. [DOI] [PubMed] [Google Scholar]
  53. Smith JB, Liang Z, Watson GD, Alloway KD, Zhang N, 2016. Interhemispheric resting-state functional connectivity of the claustrum in the awake and anesthetized states. Brain Struct Funct. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Spear LP, 2000. The adolescent brain and age-related behavioral manifestations. Neurosci Biobehav Rev 24, 417–463. [DOI] [PubMed] [Google Scholar]
  55. Stein MB, Koverola C, Hanna C, Torchia MG, McClarty B, 1997. Hippocampal volume in women victimized by childhood sexual abuse. Psychol Med 27, 951–959. [DOI] [PubMed] [Google Scholar]
  56. Stevens MC, Pearlson GD, Calhoun VD, 2009. Changes in the interaction of resting-state neural networks from adolescence to adulthood. Hum Brain Mapp 30, 2356–2366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Supekar K, Musen M, Menon V, 2009. Development of large-scale functional brain networks in children. PLoS Biol 7, e1000157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Vallortigara G, Rogers LJ, Bisazza A, 1999. Possible evolutionary origins of cognitive brain lateralization. Brain Res Brain Res Rev 30, 164–175. [DOI] [PubMed] [Google Scholar]
  59. Vythilingam M, Heim C, Newport J, Miller AH, Anderson E, Bronen R, Brummer M, Staib L, Vermetten E, Charney DS, Nemeroff CB, Bremner JD, 2002. Childhood trauma associated with smaller hippocampal volume in women with major depression. Am J Psychiatry 159, 2072–2080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wiesel TN, Hubel DH, 1963. Single-Cell Responses in Striate Cortex of Kittens Deprived of Vision in One Eye. J Neurophysiol 26, 1003–1017. [DOI] [PubMed] [Google Scholar]
  61. Yoshida K, Mimura Y, Ishihara R, Nishida H, Komaki Y, Minakuchi T, Tsurugizawa T, Mimura M, Okano H, Tanaka KF, Takata N, 2016. Physiological effects of a habituation procedure for functional MRI in awake mice using a cryogenic radiofrequency probe. J Neurosci Methods 274, 38–48. [DOI] [PubMed] [Google Scholar]
  62. Zhang N, Rane P, Huang W, Liang Z, Kennedy D, Frazier JA, King J, 2010. Mapping restingstate brain networks in conscious animals. J Neurosci Methods 189, 186–196. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1515328-supplement-1.pdf (794.7KB, pdf)

RESOURCES