doi: 10.1093/schbul/sbab027.
Hyperactivity is a Core Endophenotype of Elevated Neuregulin-1 Signaling in Embryonic Glutamatergic Networks
Tilmann Götze 1 2, Maria Clara Soto-Bernardini 1, Mingyue Zhang 3, Hendrik Mießner 2, Lisa Linhoff 1 4, Magdalena M Brzózka 5, Viktorija Velanac 1, Christian Dullin 6 7 8, Fernanda Ramos-Gomes 7, Maja Peng 3, Hümeyra Husseini 3, Eva Schifferdecker 3, Robert Fledrich 9, Michael W Sereda 1 4, Katrin Willig 10 11, Frauke Alves 6 7, Moritz J Rossner 5, Klaus-Armin Nave 1, Weiqi Zhang 3, Markus H Schwab 1 2 12
Affiliations
- PMID:33871014
- PMCID: PMC8379540
- DOI: 10.1093/schbul/sbab027
Item in Clipboard
Hyperactivity is a Core Endophenotype of Elevated Neuregulin-1 Signaling in Embryonic Glutamatergic Networks
Tilmann Götze et al. Schizophr Bull..
Display options
Format
Display options
Format
Abstract
The neuregulin 1 (NRG1) ErbB4 module is at the core of an "at risk" signaling pathway in schizophrenia. Several human studies suggest hyperstimulation of NRG1-ErbB4 signaling as a plausible pathomechanism; however, little is known about the significance of stage-, brain area-, or neural cell type-specific NRG1-ErbB4 hyperactivity for disease-relevant brain endophenotypes. To address these spatiotemporal aspects, we generated transgenic mice for Cre recombinase-mediated overexpression of cystein-rich domain (CRD) NRG1, the most prominent NRG1 isoform in the brain. A comparison of "brain-wide" vs cell type-specific CRD-NRG1 overexpressing mice revealed that pathogenic CRD-NRG1 signals for ventricular enlargement and neuroinflammation originate outside glutamatergic neurons and suggests a subcortical function of CRD-NRG1 in the control of body weight. Embryonic onset of CRD-NRG1 in glutamatergic cortical networks resulted in reduced inhibitory neurotransmission and locomotor hyperactivity. Our findings identify ventricular enlargement and locomotor hyperactivity, 2 main endophenotypes of schizophrenia, as specific consequences of spatiotemporally distinct expression profiles of hyperactivated CRD-NRG1 signaling.
Keywords: ErbB4 receptor; conditional transgenic mice; schizophrenia; ventricular enlargement.
© The Author(s) 2021. Published by Oxford University Press on behalf of the Maryland Psychiatric Research Center.
Figures
Glutamatergic network-specific stimulation of CRD-NRG1/ErbB4 signaling. (A) CRD-NRG1 protein structure. Arrowhead, proteolytic cleavage in juxtamembrane “stalk” region. C, C-terminus; CRD, cysteine rich domain; cyt, cytoplasm; EGF, epidermal growth factor-like domain; ex, extracellular; ICD, intracellular domain; N, N-terminus; TM, transmembrane domain. (B) (top) Schedule of Cre-mediated recombination inNEX-Cre mice. (bottom)NEX-Cre-mediated removal of “floxed”Stop element permits HA-CRD-NRG1 expression fromStop-Nrg1 transgene. (C) HA-CRD-NRG1 expression in the cortical plate ofNC-Nrg1 mice. Fluorescent immunostainings on coronal brain sections (E16). Scale bar, 10 µm. (D) HA-CRD-NRG1 expression in hippocampus and cortex is restricted to glutamatergic projection neurons. Immunostaining on brain sections fromNC-Nrg1 mice (4 months) for HA epitope, glutamatergic (neurogranin, NRGN), and GABAergic markers (parvalbumin, PV; GAD67). Arrowhead (left panel) marks NeuN+ neuron without HA epitope expression (most likely GABAergic neuron). Arrowhead (right panel) points to HA epitope-negative GABAergic neuron. cx, cortex. Scale bars, 20 µm. (E) Western blotting of hippocampal protein lysates from controls (NEX-Cre;Stop-Nrg1) andNC-Nrg1 mice (4 months). CTF, C-terminal fragment; FL, full length. (F) Densitometric quantification of pErbB4 bands (Tyr1284). Integrated density values normalized to ErbB4/β-actin.n = 3 each, ***P < .001; one-way ANOVA with Bonferroni’s multiple comparison test (F (2) = 41.62, P = .0003). (G) Schematic temporal profile of HA-CRD-NRG1 overexpression inNC-Nrg1 mice.
Cortical-restricted vs brain-wide CRD-NRG1-mediated ErbB4 hyperactivation in transgenic mice. (A) Immunostainings for HA-CRD-NRG1 (age 4 months). Scale bar, 500 μm. a, amygdala; hip, hippocampus; hy, hypothalamus; ncx, neocortex; st, striatum; th, thalamus. (B) Western blotting of cortical protein lysates (P10). Actin served as loading control. Asterisk, unspecific band. FL, full-length; NTF, N-terminal fragment. (C) Densitometric quantification of full-length CRD-NRG1 normalized to actin.n = 3; **P < .01.t-test. AU, arbitrary units. (D) Western blot analysis of pErbB4 (Tyr1284) and ErbB4 in cortical protein lysates (P10). Actin served as loading control. (E–G) Densitometric quantification of (E, F) pErbB4 bands normalized to ErbB4 and actin (F (2) = 9.515, P = .0077), (G) ErbB4 bands normalized to actin (F (2) = 18.18, P = .0011).n = 3–4; *P < .05, ***P < .001. One-way ANOVA with Bonferroni’s multiple comparison test,t-test; AU, arbitrary units. (H) Body weight (wt, T-Nrg1, NC-Nrg1; 3 months) (F (2) = 7.937, P = .0007). (I) Body weight ofCC-Nrg1 mice (15–18 weeks). Values, mean ± SEM. **P < .01. One-way ANOVA with Bonferroni’s multiple comparison test.
CRD-NRG1 signals that affect the ventricular system originate from cells other than glutamatergic neurons. (A) µCT analysis (3 months). (B) Volume rendering of µCT data sets. (C) Quantification of total ventricular volume/brain volume normalized to wt (F (2) = 14.13, P = .0004). Error bars, normalized mean ± SEM. One-way ANOVA with Bonferroni’s multiple comparison test. *P < .05; ***P < .001. Scale bars, 1 mm. (D) µCT analysis ofT-Nrg1 mice (P35) andDC-Nrg1 mice (P30). (E) Quantification of ventricular volume as in (C) (F (2) = 119.9, P < .0001); ****P < .0001. (F) H&E histology and GFAP immunostaining on coronal brain sections following µCT analysis. (left) Upper dashed line separates cortical layers I–IV from V/VI, lower line marks border to corpus callosum (cc). (right) Zoomed images of boxed area in left panel. Scale bar, 100 µm. (G) GFAP+ area (normalized to HE+ area) calculated as ratio lower/upper cortex. Values normalized to wt. Error bars, normalized mean ± SEM. Mann–WhitneyU-test, **P < .01. (H) GFAP immunostaining of brain sections following µCT analysis. Upper dashed line inDC-Nrg1 sample separates equally sized cortical areas. Scale bar, 100 µm. (I) Quantification as in (H). Statistical analysis not performed due ton = 2Dhh-Cre controls.
Inhibitory synaptic transmission, synaptic plasticity, and theta oscillations are affected inT-Nrg1 andNC-Nrg1 mice. (A, B) Top: sample traces of responses before and after HFS in CA1 ofT-Nrg1 (A) andNC-Nrg1 mice (B). Below: reduced LTP inT-Nrg1 (A) andNC-Nrg1 mice (B). (C) Top: sIPSC traces in PFC ofWT andT-Nrg1 mice; below: averaged sIPSC frequency was increased in PrL ofT-Nrg1 mice. (D) Top: mIPSC traces in the PFC ofWT andT-Nrg1 mice; below: increased averaged mIPSC frequency in PrL ofT-Nrg1 mice. (E) Top: sIPSC traces in PFC of control andNC-Nrg1 mice; below: reduced averaged sIPSC amplitude in PrL ofNC-Nrg1. (F) Top: mIPSC traces in PFC of control andNC-Nrg1 mice; below: averaged mIPSC frequency was reduced in PrL ofNC-Nrg1 mice. Bars, group means (±SEM). *p < .05; **p < .01; ***p < .001. (G, H) Top: recordings of sIPSCs in a CA1 pyramidal cell before and following addition of 5 µM carbachol inT-Nrg1 (G) andNC-Nrg1 mice (H). Below: autocorrelation and relative power of 10-s stretches from recordings after carbachol treatment inT-Nrg1 (G) andNC-Nrg1 mice (H).
Cortical-restricted CRD-NRG1 hyperstimulation in glutamatergic neurons causes novelty-induced hyperactivity. (A) Tendency for reduced freezing ofNC-Nrg1 mice when tested for cued fear memory in fear conditioning (context baseline, F (3) = 2.035, P = .1188; context, F (3) = 2.798, P = .0479; cue baseline, F (3) = 1.390; P = .2546; cue, F (3) = 3.674, P = .017). (B–D) Hole board test. Number of visits was unaltered (F (3) = 0.8934, P = .45). (B–D) Hole board test. Number of visits was unaltered (F (3) = 0.8934, P = .45) (B), but distance traveled was increased (F (3) = 5.385, P = .0024) (C), most pronounced during early stages when distance was plotted against time (F (3) = 26.99, P < .0001) (D). (E–G) Open field test. Number of corners visited (F (3) = 9.073, P < .0001) (E) and distance traveled (F (3) = 9.519; P < .0001) (F) were increased, with increased distance traveled most pronounced during early stages (F (3) = 39.09, P < .0001) (G). (H)NC-Nrg1 mice exhibited increased total time active in the tail suspension test (F (3) = 8.343, P = .0001). Data were analyzed using ordinary one-way ANOVA with Sidak’s multiple comparisons posttest (B, C; E, F, H), 2-way ANOVA with Tukey’s posttest (A) or 2-way ANOVA with Dunnett’s posttest for repeated measurements (D, G). Individual data points are shown with means ±SD (A–C, E, F, H) or ±SEM (D, G) (for color figure refer online version). n.s., not significant; ***P < .001, **P < .01, *P < .05.
References
- Birnbaum R, Weinberger DR. Genetic insights into the neurodevelopmental origins of schizophrenia. Nat Rev Neurosci. 2017;18(12):727–740. - PubMed
- Pardiñas AF. Common schizophrenia alleles are enriched in mutation-intolerant genes and maintained by background selection. bioRxiv, doi:10.1101/068593, 9 August 2016, preprint: not peer reviewed. - DOI
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources
Medical