Growth differentiation factor 5 is a key physiological regulator of dendrite growth during development

Abstract

Dendrite size and morphology are key determinants of the functional properties of neurons. Here, we show that growth differentiation factor 5 (GDF5), a member of the bone morphogenetic protein (BMP) subclass of the transforming growth factor β superfamily with a well-characterised role in limb morphogenesis, is a key regulator of the growth and elaboration of pyramidal cell dendrites in the developing hippocampus. Pyramidal cells co-express GDF5 and its preferred receptors, BMP receptor 1B and BMP receptor 2, during development. In culture, GDF5 substantially increased dendrite, but not axon, elongation from these neurons by a mechanism that depends on activation of SMADs 1/5/8 and upregulation of the transcription factor HES5. In vivo, the apical and basal dendritic arbours of pyramidal cells throughout the hippocampus were markedly stunted in both homozygous and heterozygous Gdf5 null mutants, indicating that dendrite size and complexity are exquisitely sensitive to the level of endogenous GDF5 synthesis.

Keywords: Bone morphogenetic protein; Dendrite; Growth differentiation factor 5; Hippocampus; Mouse.

Fig. 1.

Expression of GDF5, BMPR1B and BMPR2 in the developing mouse hippocampus. (A-C) Relative levels of Gdf5 (A), Bmpr1b (B) and Bmpr2 (C) mRNA in the hippocampus (mean ± s.e.m., n=4 per age). (D-F) Representative western blots probed for GAPDH and either GDF5 (D), BMPR1B (E) or BMPR2 (F) and graphs of pro-GDF5, mature GDF5, BMPR1B and BMPR2 relative to GAPDH in the hippocampus (mean ± s.e.m., n=3 per age). Lysates of E18 midbrain were used as positive control.

Fig. 2.

Localisation of GDF5, BMPR1B and BMPR2 in the developing mouse hippocampus. (A) Sections of P10 hippocampus triple labelled with TOTO-3, anti-MAP2 and either anti-GDF5, anti-BMPR1B or anti-BMPR2. Upper panels show low-power images of CA regions and dentate gyrus and the lower panels show high-power images of CA1. Gdf5^bp hippocampus was used as a control for anti-GDF5. Controls for anti-BMPR1B and anti-BMPR2 received no primary antibodies. (B) E18 hippocampal neurons double stained with anti-MAP2 and either anti-GDF5, anti-BMPR1B or anti-BMPR2 after 7 days in culture. PL, pyramidal layer; DG, dentate gyrus; sr, stratum radiatum. Scale bars: 200 μm (A, upper panels); 25 μm (A, lower panels; B).

Fig. 3.

GDF5 selectively enhances dendrite growth from cultured mouse hippocampal pyramidal cells. (A-C) Axonal growth from E18 neurons after 3 days. (A) Representative neurons incubated with or without 100 ng/ml GDF5. (B) Axon length in control cultures and cultures treated with GDF5. (C) Camera lucida drawings of representative control neurons and neurons treated with 100 ng/ml GDF5. (D-F) Dendrite growth from E18 neurons after 7 days. (D) Representative neurons incubated with and without 100 ng/ml GDF5. (E) Percentage of dendrites longer than 50 μm in control and GDF5-treated cultures. (F) Camera lucida drawings of representative control and GDF5-treated neurons. Mean ± s.e.m. of data from >150 neurons per condition from at least three separate experiments are shown. ^**P<0.001, ^***P<0.0001, statistical comparison with control. Scale bars: 100 μm (A,C); 50 μm (D,F).

Fig. 4.

BMPR1B and BMPR2 are required for GDF5-promoted dendrite growth. (A) Representative E18 neurons transfected after 6 days with plasmids expressing either GFP alone or GFP plus either constitutively active (CA) BMPR1B or dominant-negative (DN) BMPR1B. Six hours after transfection, the neurons were cultured for 18 hours with or without 100 ng/ml GDF5. (B) Percentage of dendrites longer than 50 μm under these experimental conditions. (C) Representative neurons transfected after 6 days with plasmids expressing either GFP alone or GFP plus either wild-type (WT) BMPR2 or dominant-negative BMPR2. After transfection, the neurons were cultured for 18 hours with or without GDF5. (D) Percentage of dendrites longer than 50 μm under these experimental conditions. Mean ± s.e.m. of data from >150 neurons per condition from at least three separate experiments are shown ^**P<0.001, ^***P<0.0001, statistical comparison with control. Scale bars: 50 μm.

Fig. 5.

GDF5 promotes dendrite growth by activating SMAD signalling. (A) Representative western blot probed for phospho-SMADs 1/5/8, total SMAD1 and β3 tubulin in lysates of E18 hippocampal neurons of treated with 100 ng/ml GDF5 after culturing the neurons for 7 days. (B) Levels of phospho-SMAD1/5/8 relative to total SMAD1 from three separate western blot experiments (mean ± s.e.m.). (C) Reporter signal from E18 hippocampal neurons transfected with the SMAD1/5/8-responsive BRE-luciferase reporter after 6 days in vitro and cultured for a further 24 hours with or without 100 ng/ml GDF5 (mean ± s.e.m. from three independent experiments, signal normalised to 1 for controls). (D) Representative neurons transfected after 6 days in vitro with plasmids expressing either GFP alone or GFP and either SMAD6, SMAD7 or dominant-negative (DN) SMAD4. After transfection, the neurons were cultured for a further 18 hours with or without 100 ng/ml GDF5. (E) Percentage of dendrites longer than 50 μm under these experimental conditions. Mean ± s.e.m. of data from >150 neurons per condition from at least three separate experiments are shown. ^***P<0.0001, statistical comparison with control. Scale bar: 50 μm.

Fig. 6.

GDF5 promotes hippocampal dendrite growth by inducing HES5 expression. (A) Levels of Hes1 and Hes5 mRNAs relative to reference mRNAs in hippocampal neurons treated with 100 ng/ml of GDF5 for 4 or 18 hours after 6 days in culture. Hes mRNA levels are normalised to untreated cultures (controls). (B) Luciferase reporter signal from E18 hippocampal neurons transfected with either the pHes5-pGL3 or pHes5-SM-pGL3 reporter after 6 days in vitro and cultured for a further 24 hours with or without 100 ng/ml GDF5. The reporter signal in GDF5-treated cultures is normalised to the signal in untreated cultures (controls). (C) Representative western blot probed for HES5, GFP and GAPDH in lysates of HEK 293T cells transfected with either an empty expression plasmid or plasmids expressing either Hes5 shRNA or a scrambled RNA sequence. (D) Percentage of dendrites >50 μm growing from E18 neurons that were transfected after 6 days in vitro with a GFP plasmid together with a plasmid that either expresses Hes5 shRNA or a Hes5 shRNA scrambled sequence. After transfection, the neurons were cultured for a further 18 hours with or without GDF5. (E) Percentage of dendrites longer than 50 μm from E18 neurons that were transfected after 6 days in vitro with either pGFP alone or pGFP and plasmids that express either wild-type HES5 or truncated HES5. Six hours after transfection, the neurons were cultured for 18 hours with or without 100 ng/ml GDF5. The mean ± s.e.m. of three experiments is shown for each dataset. (A) ^*P<0.05, ^**P<0.01, compared with control, REST test. (B,D,E) ^***P<0.0001, compared with control, Mann-Whitney U-test. n.s, not significant.

Fig. 7.

Reduced size and complexity of pyramidal dendrites in Gdf5^bp mice. (A,B) Representative images of Golgi-stained preparations of the hippocampi of P10 wild-type and Gdf5^bp mice showing pyramidal cells in CA1 (A) and CA3 (B). Sholl plots, total length and number of branch points in the apical (C) and basal (D) dendrites of the CA1 field of P10 hippocampi of P10 wild-type, Gdf5^bp and heterozygous mice. The dendrites of at least 50 neurons from at least six separate mice of each genotype were analysed (mean ± s.e.m., ^**P<0.001, ^***P<0.0001, statistical comparison with wild type). Scale bar: 50 μm.