Intramembranous valine linked to schizophrenia is required for neuregulin 1 regulation of the morphological development of cortical neurons - PubMed (original) (raw)

Intramembranous valine linked to schizophrenia is required for neuregulin 1 regulation of the morphological development of cortical neurons

Yachi Chen et al. J Neurosci. 2010.

Abstract

Neuregulin 1 (NRG1) signaling is critical to various aspects of neuronal development and function. Among different NRG1 isoforms, the type III isoforms of NRG1 are unique in their ability to signal via the intracellular domain after gamma-secretase-dependent intramembranous processing. However, the functional consequences of type III NRG1 signaling via its intracellular domain are mostly unknown. In this study, we have identified mutations within type III NRG1 that disrupt intramembranous proteolytic processing and abolish intracellular domain signaling. In particular, substitutions at valine 321, previously linked to schizophrenia risks, result in NRG1 proteins that fail to undergo gamma-secretase-mediated nuclear localization and transcriptional activation. Using processing-defective mutants of type III NRG1, we demonstrate that the intracellular domain signaling is specifically required for NRG1 regulation of the growth and branching of cortical dendrites but not axons. Consistent with the role of type III NRG1 signaling via the intracellular domain in the initial patterning of cortical dendrites, our findings from pharmacological and genetic studies indicate that type III NRG1 functions in dendritic development independent of ERBB kinase activity. Together, these results support the proposal that aberrant intramembranous processing and defective signaling via the intracellular domain of type III NRG1 impair a subset of NRG1 functions in cortical development and contribute to abnormal neuroconnectivity implicated in schizophrenia.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

Type III NRG1 regulates dendritic development of cortical neurons in vivo. Confocal image stacks of YFP-labeled neurons in the cortical plate of E19 type III Nrg1 knock-out (KO/YFP+) mice as well as their wild-type (WT/YFP+) littermates were captured and analyzed for the total length and number of basal dendritic branches. A, Shown are images from collapsed image stacks of WT/YFP+ and KO/YFP+ cortical neurons. YFP expression is shown in green. KO/YFP+ cortical neuron shows decreased complexity of basal dendrites. B, C, Results from KO/YFP+ neurons were compared with that of WT/YFP+ neurons. KO/YFP+ neurons had reduced total basal dendritic length and fewer basal dendritic branches. The results are based on three independent experiments: WT/YFP+ neurons (60 neurons, 5 animals), KO/YFP+ neurons (55 neurons, 3 animals). Scale bar, 10 μm. Values represent means ± SEM. **p < 0.001.

Figure 2.

Figure 2.

Type III NRG1 is expressed in cortical neurons and functions in the development of dendrites and axons in vitro. A, Cortical neurons (9 DIV) from E18 WT mice were processed for immunostaining of type III NRG1 (green). Endogenous type III NRG1 expression was detected in the cell bodies and neuronal processes (arrowheads). A single optical section through the center of the _z_-stack is shown. B, Dispersed P0 WT cortical neurons (3 DIV) were immunostained with antibodies for type III NRG1 (red) and the somatodendritic marker MAP2 (green). The nuclear morphology is revealed by DAPI staining (blue). Collapsed images of _z_-stacks are shown. Type III NRG1 is present in the cell bodies as well as MAP2-positive processes (arrowheads) of cells. C, Dispersed cortical neurons (3 DIV) from an E19 type III Nrg1 heterozygous mouse were immunostained with antibodies against type III NRG1 (green), MAP2 (blue), and the glutamatergic axonal marker VGLUT1 (red). Type III NRG1 was detected in the axonal branches (i.e., MAP2-negative but VGLUT1-positive processes; yellow arrows) and the dendritic branches (i.e., MAP2- and VGLUT1-positive processes; white arrowheads). Collapsed images of _z_-stacks are shown. D–H, EGFP (green) was transiently expressed in dispersed WT and KO neurons. Dendrites and axons of EGFP/MAP2 positive neurons were analyzed for the total length and number of branch points. KO neurons showed shorter total dendritic length (D, E), fewer dendritic branch points (D, F), and shorter total axonal length (G) than WT neurons. Axonal branch point numbers (H) were not significantly different between WT versus KO neurons. Only partial view of the axon (i.e., the longest branch) is shown in D. Note that the total length and number of branch points of dendrites or axons of WT versus KO neurons are also shown in Figure 4. See Figure 4 legend for additional information. Scale bars, 10 μm. Values represent means ± SEM. *p < 0.05; n.s., not significant.

Figure 3.

Figure 3.

Valine residues 321 and 322 are involved in type III NRG1 intramembranous processing, nuclear translocation, and transactivation. A, Type III NRG1 undergoes proteolytic cleavage (yellow arrows), first in the extracellular juxtamembrane region and subsequently in the TMc. The second cleavage is γ-secretase dependent and releases NRG1-ICD from the membrane. N, N-terminal; C, C-terminal; CRD, cysteine-rich domain; EGF, EGF-like domain; TMc, C-terminal transmembrane domain; ICD, intracellular domain. B, Schematic of chimeric wild-type NRG1 (NRG1-WT) and mutant NRG1 Gal4DBD fusion constructs used in the study. Positions of alanine substitutions are shown. NLS, Nuclear localization signal. C, NRG1-WT, NRG1-T307A/G308A, NRG1-V321A/V322A, or NRG1-K329A/Q330A Gal4DBD chimera was coexpressed with a luciferase reporter construct to assay nuclear localization and reporter gene expression in cortical neurons. Cells were treated with vehicle or the extracellular domain of ERBB4 (B4-ECD). Firefly/Renilla luciferase luminescence ratios were determined. Luminescence values (means ± SEM) based on four individual experiments are shown: n = 4 vector, 6 NRG1-WT plus B4-ECD, 4 NRG1-T307A/G308A plus B4-ECD, 5 NRG1-V321A/V322A plus B4-ECD, 5 NRG1-K329A/Q330A plus B4-ECD. AU, Arbitrary units. *p < 0.05; n.s., not significant. D, E, NRG1-WT or NRG1-V321A/V322A was transiently expressed in cortical neurons. Cells were treated with vehicle or B4-ECD in the absence or presence of a γ-secretase inhibitor (L685458), processed for immunofluorescent localization of Gal4DBD (Gal4; green), and nuclear stained with DAPI (blue). Transfected neurons were identified based on Gal4 staining. Confocal image stacks of Gal4DBD localization were acquired. In D, a single optical section through the nucleus of a vehicle-treated NRG1-WT-expressing, a B4-ECD-treated NRG1-WT-expressing, or a B4-ECD-treated NRG1-V321A/V322A-expressing cell is shown in each panel. In vehicle-treated NRG1-WT-expressing cells, Gal4 immunoreactivity was detected in the soma and processes and weakly in nuclei. Treatment with B4-ECD enhanced nuclear Gal4 staining in NRG1-WT-expressing cells but not in NRG1-V321A/V322A-expressing cells. Nuclear Gal4 immunoreactivity was quantified in 30 and 71 neurons in two independent experiments, and results are presented in E. Within each experiment, the mean Gal4 intensity from all analyzed neurons per condition was first determined and then normalized against that of vehicle-treated NRG1-WT-expressing cells to obtain “fold induction of nuclear Gal4 intensity.” The average of two values of fold induction from two experiments is plotted in the graph. Stimulation of type III NRG1 processing and translocation via B4-ECD treatment resulted in 1.6-fold increase in the nuclear localization of Gal4DBD-tagged NRG1-ICD. In NRG1-WT-expressing cells, this increase in the nuclear localization of Gal4DBD-tagged NRG1-ICD in response to B4-ECD was eliminated by the γ-secretase inhibitor. In NRG1-V321A/V322A-expressing cells, ERBB binding via B4-ECD treatment failed to stimulate nuclear translocation of Gal4DBD-tagged NRG1-ICD in the absence or presence of the γ-secretase inhibitor. Scale bar, 10 μm.

Figure 4.

Figure 4.

The development of cortical dendrites but not axons specifically requires nuclear targeting of the intracellular domain of type III NRG1. EGFP (green) was transiently expressed in dispersed WT and KO neurons. Rescue was assessed by coexpression of wild-type and mutant forms of NRG1 with EGFP. A, KO neurons show shorter dendritic length and fewer branch points than WT neurons. Coexpression of NRG1-WT in KO neurons rescued dendritic growth and branching pattern. In contrast, coexpression of mutant NRG1-V321A/V322A in KO neurons does not restore normal dendritic phenotypes. Only partial view of the axon (i.e., the longest branch) is shown. B–G, EGFP/MAP2-positive neurons were analyzed for total dendritic length (B, D), number of dendritic branch points (C, E), total axonal length (F), and number of axonal branch points (G). Expression of NRG-WT or mutant NRG1-T307A/G308A in KO neurons rescued dendritic length (B, D), number of dendritic branch points (C, E), and axonal length (F). Mutant NRG1-V321A/V322A expression failed to restore the dendritic phenotypes of KO neurons (B, C) but rescued the axonal length (F). Similarly, expression of NRG1-K329A/Q330A in KO neurons failed to rescue the total dendritic length (D) and number of branch points (E) but restored the axonal length (F). WT neurons, KO neurons, and KO neurons expressing either wild-type or mutant forms of NRG1 exhibited comparable axonal branch points (G). Where indicated in D and E, NRG1-WT-expressing KO neurons were treated with PD158780, an ERBB kinase inhibitor (ERBB inhibitor). Expression of NRG1-WT in KO neurons restored dendritic length (D) and number of branch points (E) even in the presence of the ERBB kinase inhibitor. The dendrite results presented in B and C are based on three independent transfections: WT neurons (47 neurons, 3 animals), KO neurons (64 neurons, 5 animals), NRG1-WT-expressing KO neurons (40 neurons, 3 animals), NRG1-T307A/G308A-expressing KO neurons (51 neurons, 5 animals), and NRG1-V321A/V322A-expressing KO neurons (43 neurons, 4 animals). The dendrite data shown in D and E are based on two independent transfections: WT neurons (16 neurons, 2 animals), KO neurons (48 neurons, 4 animals), NRG1-K329A/Q330A-expressing KO neurons (37 neurons, 4 animals), NRG1-WT-expressing KO neurons (35 neurons, 4 animals), and ERBB inhibitor-treated NRG1-WT-expressing KO neurons (48 neurons, 4 animals). The axon results presented in F and G are based on four independent transfections: WT neurons (39 neurons, 3 animals), KO neurons (88 neurons, 6 animals), NRG1-WT-expressing KO neurons (64 neurons, 6 animals), NRG1-T307A/G308A-expressing KO neurons (45 neurons, 5 animals), NRG1-V321A/V322A-expressing KO neurons (34 neurons, 4 animals), and NRG1-K329A/Q330A-expressing KO neurons (17 neurons, 2 animals). Scale bar, 10 μm. Values represent means ± SEM. *p < 0.05; n.s., not significant.

Figure 5.

Figure 5.

Type III NRG1 function in cortical dendritic development requires residue 321-mediated proteolytic processing and signaling via the intracellular domain and does not involve ERBB kinase activation. A, Schematic of wild-type (NRG1-WT) and mutant form of type III NRG1 containing the valine-to-leucine substitution (NRG1-V321L). Note that both constructs do not contain the Gal4DBD fusion protein. B–F, NRG1-WT or NRG1-V321L was coexpressed with EGFP in WT versus KO cortical neurons. In B, KO neuron and NRG1-V321L-expressing KO neuron show comparable total dendritic growth and branching pattern. B4KO neuron from ErbB4 knock-out mouse shows more extensive dendritic development than KO neuron from type III Nrg1 knock-out animal. Only partial view of the axon (i.e., the longest branch) is shown. Expression of NRG1-WT but not NRG1-V321L in KO neurons restored the total dendritic length (C) and the number of dendritic branch points (D). The presence of an ERBB kinase inhibitor (ERBB inhibitor), PD158780, did not affect the ability of NRG1-WT to rescue dendritic phenotypes of KO neurons (C, D). The total dendritic length (C) and the number of dendritic branch points (D) of ErbB4 knock-out cortical neurons (B4KO) were not significantly different from that of WT neurons. Expression of NRG1-WT or NRG1-V321L in KO neurons restored axonal length (E). The total axonal lengths (E) of WT neurons versus B4KO neurons were comparable. WT neurons, KO neurons, KO neurons expressing either NRG1-WT or NRG1-V321L, and B4KO neurons were not significantly different in the number of axonal branch points (F). The dendrite data shown in C and D are based on two independent transfections: WT neurons (42 neurons, 3 animals), KO neurons (51 neurons, 4 animals), NRG1-WT-expressing KO neurons (29 neurons, 2 animals), ERBB inhibitor-treated NRG1-WT-expressing KO neurons (30 neurons, 2 animals), NRG1-V321L-expressing KO neurons (34 neurons, 2 animals), and B4KO neurons (24 neurons, 2 animals). The axon results presented in E and F are based on three independent transfections: WT neurons (31 neurons, 2 animals), KO neurons (54 neurons, 8 animals), NRG1-WT-expressing KO neurons (27 neurons, 3 animals), NRG1-V321L-expressing KO neurons (35 neurons, 3 animals), and B4KO neurons (24 neurons, 2 animals). Scale bar, 10 μm. Values represent means ± SEM. *p < 0.05; n.s., not significant.

Comment in

Similar articles

Cited by

References

    1. Bao J, Wolpowitz D, Role LW, Talmage DA. Back signaling by the Nrg-1 intracellular domain. J Cell Biol. 2003;161:1133–1141. - PMC - PubMed
    1. Bao J, Lin H, Ouyang Y, Lei D, Osman A, Kim TW, Mei L, Dai P, Ohlemiller KK, Ambron RT. Activity-dependent transcription regulation of PSD-95 by neuregulin-1 and Eos. Nat Neurosci. 2004;7:1250–1258. - PubMed
    1. Barros CS, Calabrese B, Chamero P, Roberts AJ, Korzus E, Lloyd K, Stowers L, Mayford M, Halpain S, Müller U. Impaired maturation of dendritic spines without disorganization of cortical cell layers in mice lacking NRG1/ErbB signaling in the central nervous system. Proc Natl Acad Sci U S A. 2009;106:4507–4512. - PMC - PubMed
    1. Chen Y, Ghosh A. Regulation of dendritic development by neuronal activity. J Neurobiol. 2005;64:4–10. - PubMed
    1. Chen YJ, Johnson MA, Lieberman MD, Goodchild RE, Schobel S, Lewandowski N, Rosoklija G, Liu RC, Gingrich JA, Small S, Moore H, Dwork AJ, Talmage DA, Role LW. Type III neuregulin-1 is required for normal sensorimotor gating, memory-related behaviors, and corticostriatal circuit components. J Neurosci. 2008;28:6872–6883. - PMC - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources