Silke Kreitz1, Alice Zambon2, Marianne Ronovsky2, Lubos Budinsky3, Thomas Helbich3, Spyros Sideromenos2, Claudiu Ivan1, Laura Christina Konerth1, Isabel Wank1, Angelika Berger4, Arnold Pollak4, Andreas Hess1, and Daniela D. Pollak2
1Institute of Experimental and Clinical Pharmacology and Toxicology, Friedrich Alexander University Erlangen-Nuremberg, Erlangen, Germany, 2Department of Neurophysiology and Neuropharmacology, Medical University of Vienna, Vienna, Austria, 3Department of Biomedical Imaging and Image-Guided Therapy, Division of Molecular and Gender Imaging, Medical University of Vienna, Vienna, Austria, 44Department of Pediatrics and Adolescent Medicine, Medical University of Vienna, Vienna, Austria
Synopsis
The infection of the
pregnant female and the ensuing induction of maternal immune activation affect
fetal development with long-lasting consequences for health and disease.
Specifically aberrant neural wiring may contribute in the manifestation of
psychiatric disorders such as depression. Here, we investigated altered resting
state functional connectivity using fMRI in adult mice after prenatal immune
activation. While the overall flow of information was intact, especially the
cortico-limbic connectivity was disrupted in resting state networks of adult
offspring. We propose that these altered connectivity patterns may lead to behavioral
and emotional abnormalities with relevance for neuropsychiatric disorders.
Introduction
Gestational infection
constitutes a risk factor for the occurrence of psychiatric disorders in
offsprings1. Activation of the maternal immune system with
subsequent impact on the development of the fetal brain is considered to form
the neurobiological basis for aberrant neural wiring and the psychiatric
manifestations later in offspring life2. Here we used a validated
animal model to investigate the impact of maternal immune activation (MIA) particularly
on adult offspring resting state functional connectivity.Methods
Pregnant C57Bl6/N mice were injected intraperitoneally at embryonic day 12.5 either with Poly(I:C) (=
polyriboinosinic-polyribocytidilic acid), a synthetic analog of virus-specific
double-stranded RNA, to induce MIA or with 0.9% NaCl (Saline) as vehicle
control. Functional connectivity was assessed with resting state functional MRI
(RS-fMRI) in the 3 month old offspring.
RS-fMRI data were acquired with a T2*-weighted
single-shot gradient echo-based Echo Planar Imaging sequence (GE-EPI) covering
22 axial slices of the brain in 2 seconds (600 volumes, total time 10 minutes,
TEef=15 ms, TR=2000 ms, matrix 94x64, FOV 15x15 mm, slice thickness
0.5 mm). The measured matrix was reconstructed by zero filling to 128x128
pixels resulting in a final resolution of 117x117x500 µm.
Standard preprocessing was performed including inter-slice time and motion
correction, spatial gaussian smoothing (FWHM 0.58 mm), low pass filtering at
0.1 Hz and regression of the global mean. Brain voxels were
labeled individually for each animal as belonging to 211 pain related brain
structures based on the mouse atlas from Franklin and Paxinos3. For
RS data analysis the average time course of each seed region was correlated
with every voxel in the brain. After defining the FDR corrected significant
correlation voxels an asymmetric correlation matrix was created for each
subject using for each seed region the mean significant correlation values of
all brain structures (MSRA)4. Resulting group average networks were
thresholded at K=10 (resulting in same density, on average 10 connections per
node). Group average network communities were detected using a heuristic method
based on modularity optimization5. Additionally, on those networks small
world index σ6 and the node specific graph-theoretical parameters
degree, clustering coefficient, average shortest path length7,and
hub score8 were calculated per animal. To assess group specific
effects a two factor ANOVA (treatment, brain structures) with interaction was
performed on node specific graph-theoretical parameters. Differences in
connectivity strength between groups were calculated using network based
statistics (NBS)9.Results
First, it appears that the overall effectiveness of information flow in
Poly(I:C) mice was intact. We did not observe differences in σ, and also the hub
functionality was similar between both groups. However, analysis of node
specific community association (Fig. 1) and connectivity strength (Fig. 2)
revealed significant differences. Significant increases in connectivity
strength could be found mainly in limbic circuits (amygdala, habenulae/septum,
basal ganglia, and hypothalamus), brainstem-cerebellum connections, and between
visual/auditory cortex and structures of the posterior association as well as
somatosensory cortex. However, the thalamic connections, especially to the
cortex, and somatosensory to cerebellum were significantly weakened in MIA
offspring (Fig. 2). These changes strongly suggest a weakening in
cortico-limbic connectivity. Referring to resting state networks derived from
ICA analysis known especially from human10 but also from rodent11
studies, dominant connectivity modulation could be observed within the saliency,
the sensorimotor and the default mode network (Fig. 2A).
ANOVA on node specific graph-theoretical parameters revealed a treatment
effect only in average shortest path length and only degree showed significant
interactions (Table 1). Therefore, clustering coefficient and hub score did not
account for group differences between RS of Poly(I:C) offspring and control
animals. Degree was enhanced in olfactory input, brain stem, amygdala, nucleus
accumbens, and in thalamus and decreased in paraventricular thalamus, primary
somatosensory cortex, insula, parietal association and retrosplenial cortex
(Fig. 3A). Path length was especially enhanced in structures of brain stem
(medulla) and sensory cortex indicating enhanced segregation of these brain
areas. In general dorsal thalamic structures showed increased and ventral
decreased path length (Fig. 3B). Again, all these findings point towards higher
segregation of cortical-limbic functional connectivity due to MIA.Discussion
The remarkable dysfunction in cortical-limbic
connectivity circuits in Poly(I:C) mice
is particularly noteworthy in light of the well-documented
alterations in the cortical-limbic
mood-regulating circuitry in patients suffering from mood disorders12,13.
The observed hyper-connectivity within limbic circuits, encompassing the
amygdala, habenulae/septum, basal ganglia, and hypothalamus in Poly(I:C) mice
may be considered a result of deficient top-down inhibition by cortical areas
resulting from the aberrant cortico-limbic connectivity. The strengthened
brainstem-cerebellum, intra-brainstem and intra-cortical connections could
reflect adaptive responses indicating functional rearrangements in response to
the neurodevelopmental insult of MIA.Acknowledgements
This work was supported by the Austrian Science
Fund (stand-alone project P 27520 to D.D.P), BMBF NeuroRad (02NUK034D) and BMBF
NeuroImpa (01EC1403C) to A.H. and “Verein unser Kind” to A.B. We thank Johannes
Kaesser, Jutta Prade and Sandra John for their excellent technical support.References
1. Meyer U, Nyffeler M, Yee BK, Knuesel I, Feldon J.
Adult brain and behavioral pathological markers of prenatal immune challenge
during early/middle and late fetal development in mice. Brain Behav Immun.
2008;22(4):469–486.
2. Reisinger SN, Khan D, Kong E, Berger A, Pollak A,
Pollak DD. The poly(I:C)- induced maternal immune activation model in
preclinical neuropsychiatric drug discovery. Pharmacol Ther. 2015;149:213–226.
3. Franklin KBJ, Paxinos G. The Mouse Brain in stereotaxic coordinates. Academic Press, New
York, 3 Ed. 2008
4. Kreitz S, de Celis Alonso B, Uder M, Hess A. A New
Analysis of Resting State Connectivity and Graph Theory Reveals Distinctive
Short-Term Modulations due to Whisker Stimulation in Rats. Front Neurosci. 2018;12:334.
5. Blondel VD, Guillaume JL, Lambiotte R, Lefebvre E.
Fast unfolding of communities in large networks. J Stat Mech. 2008:10008–10020.
6. Watts DJ, Strogatz SH. Collective dynamics of
'small-world' networks. Nature
1998;393(6684):440-442.
7. Sporns O. Network attributes for segregation and integration
in the human brain. Curr Opin Neurobiol. 2013; 23:162–171.
8. Kleinberg JM. Hubs, authorities, and communities. Acm
Comput Surv. 1999; 31.
9. Zalesky A, Fornito A, Bullmore ET. Network-based
statistic: identifying differences in brain networks. NeuroImage 2010;53(4):1197-1207.
10. Raichle ME. The restless brain. Brain Connect. 2011;1:3–12
11. Becerra L, Bishop J, Barmettler G, Kainz V, Burstein
R, Borsook D. Brain network alterations in the inflammatory soup animal model
of migraine. Brain Res. 2017;1660:36–46.
12. Ketter TA, Kimbrell TA, George MS, Dunn RT, Speer AM,
Benson BE, Willis MW, Danielson A, Frye MA, Herscovitch P, Post RM. Effects of
mood and subtype on cerebral glucose metabolism in treatment-resistant bipolar
disorder. Biol Psychiatry. 2001;49(2):97–109.
13. Anand A, Li Y, Wang Y, Lowe MJ, Dzemidzic
M. Resting state corticolimbic connectivity abnormalities in unmedicated
bipolar disorder and unipolar depression. Psychiatry Res. 2009;171(3):189–198.