Neurotrophin and FGF Signaling Adapter Proteins, FRS2 and FRS3, Regulate Dentate Granule Cell Maturation and Excitatory Synaptogenesis

Abstract

Dentate granule cells (DGCs) play important roles in cognitive processes. Knowledge about how growth factors such as FGFs and neurotrophins contribute to the maturation and synaptogenesis of DGCs is limited. Here, using brain-specific and germline mouse mutants we show that a module of neurotrophin and FGF signaling, the FGF Receptor Substrate (FRS) family of intracellular adapters, FRS2 and FRS3, are together required for postnatal brain development. In the hippocampus, FRS promotes dentate gyrus morphogenesis and DGC maturation during developmental neurogenesis, similar to previously published functions for both neurotrophins and FGFs. Consistent with a role in DGC maturation, two-photon imaging revealed that Frs2,3-double mutants have reduced numbers of dendritic branches and spines in DGCs. Functional analysis further showed that double-mutant mice exhibit fewer excitatory synaptic inputs onto DGCs. These observations reveal roles for FRS adapters in DGC maturation and synaptogenesis and suggest that FRS proteins may act as an important node for FGF and neurotrophin signaling in postnatal hippocampal development.

Keywords: FGF; FRS; hippocampus; neurogenesis; neurotrophin; synaptogenesis.

Figure 1. FRS regulates body weight and contributes to survival in a fraction of animals

(A) Mating scheme for the generation of double mutant mice. (B) Reduced body weight in 33.3% of mutant mice at P20 (asterisks) and 20% of mutant mice at P30 (asterisk). Solid and broken lines respectively indicate average body weights of control and mutants. n= 9 (P20) and n=5 (P30). About 13.3% of all mutant mice (with lower body weights) died between P20 and P30 and the other 20% (also with lower body weights) died ~P60–P75. None of the mutant mice with normal body weights or control group displayed any sign of mortality.

Figure 2. Expression of the hGFAP-Cre transgene is not affected by the loss of Frs3

(A–C) IHC analyses of P7 brain sections for the detection of Cre expression in cortex (A), hippocampus (B) and cerebellum (C) on a Frs2^fl/fl;Frs3^+/− (Frs3 control) and Frs2^fl/fl;Frs3^−/− (Frs3 mutant) genetic backgrounds. Counterstain, DAPI. n=2. Cx, Cortex; DG, dentate gyrus; Cb, cerebellum. (D-E) Cre expression in the DG. IHC analyses of P7 hGFAP-Cre;Frs2^fl/fl;Frs3^+/− brain sections through DG showing Cre expression in GFAP+ (D, arrow) but not in NeuN+ cells (E). Counterstain, DAPI.

Figure 3. FRS2 and FRS3 collaborate during postnatal brain development

(A–C) Photomicrographs of sagittal H&E stained brain sections from P7 mice showing smaller brains along the A/P axis in double mutants compared with either control or single mutants. Despite a reduced size along the A/P axis, the cortical thickness was largely normal in double mutants (A). The DG was particularly reduced in double mutants (B). Cerebellar foliation deficits (lobes VI/VII and IX) were observed in Frs3 single as well as in double mutants (C, arrows). Note, Frs2 single mutants displayed a largely normal brain development even on a Frs3^+/− background. n=3. Cx, Cortex; DG, dentate gyrus; Cb, cerebellum.

Figure 4. FRS is required for DG morphogenesis

(A,C,E) Photomicrographs of sagittal H&E stained brain sections from P7 mice showing a smaller hippocampus with altered DG morphogenesis (A), a normal CA1 neuronal field area (C), and a reduced DGC field area (E) in mutants. (B,D,F) Quantitation of fields of 5× (A) or 40× (C,E) images from three consecutive equivalent sagittal planes derived from serially cut 5 µm thick sections from three mice per genotype was carried out for comparison. A.U., arbitrary units. Average ± SEM. Unpaired two-tailed Student’s t-test was used. n.s., non-significant; *p<0.05. Con, control; DM, double mutant.

Figure 5. FRS promotes DG neural stem cell proliferation and DGC maturation

(A–E) Double mutant mice exhibited reduced neural stem cell proliferation and a reduced number of mature neurons. IHC analyses of P7 brain sections through the DG for markers of proliferation, Ki67 (A); cell death, active caspase-3 (B); neural stem cells, GFAP (C); immature neurons, DCX and TBR2 (D); and mature DGCs, NeuN (E). Counterstain, hematoxylin (A) and DAPI (B–E). (F–J) Quantitation for (A–E). Twenty-four different fields from three mice per genotype were analyzed. Average ± SEM. Unpaired two-tailed Student’s t-test was used. n.s., non-significant; *p<0.05 and **p<0.01.

Figure 6. FRS is required for DGC dendritic branching

(A–B) Double mutant mice exhibited reduced DGC dendritic arborization. Mutant mice with lower body weights at P20 were excluded from these analyses. Soma of single DGCs located at least 30 µm above the hilar border in acute hippocampal slices from P21–P25 mice were patched-loaded with Alexa Fluor-594 (30 µM) for two-photon imaging of morphology. Traces of reconstructed DGC dendritic trees by concentric 40 µm bins from the soma using Neurolucida neuron tracing software were used for Sholl analyses. Imaged cells (A). Sholl analysis quantitation (B). Data points corresponding to 200 µm distance from cell body were excluded from statistical analyses due to the lack of sufficient data points in mutants. Average ± SEM. Two-way ANOVA was used (B). F_1,33=4.58; p=0.04. (C) Double mutant mice exhibited reduced total DGC dendritic tree length. At least ten different DGCs from five mice per genotype were analyzed. Average ± SEM. Unpaired two-tailed Student’s t-test was used (C). *p<0.05.

Figure 7. FRS is required for excitatory synaptogenesis in DGCs

(A) Double mutant mice exhibited reduced DGC dendritic spine density. Mutant mice with lower body weights at P20 were excluded from these analyses. Spines located in the dendrites at a distance 50 µm from the soma were imaged at higher resolution and magnification as described in the Materials and Methods. Spine density was assessed using the Image J software. Twenty different dendrites of ten different cells from five mice per genotype were used for spine density analyses. Average ± SEM. Unpaired two-tailed Student’s t test was used. (B-C) Double mutant DGCs exhibited a reduction in frequency but not amplitude of mEPSCs. Synaptic events were recorded from DGCs of P21–P25 mice in the presence of 100 µM picrotoxin and 0.5 µM TTX. Four representative mEPSC sweeps from control and mutant DGCs (B, Upper). Composite average of mEPSC events (B, Lower). Cumulative probability plots of mEPSC amplitudes (left) and inter-event intervals (right) (C). Insets, summary histograms for mean amplitude (left) and frequency (right). At least eight different cells from three mice per genotype were considered. Average ± SEM. One-way ANOVA was used. n.s., non-significant; **p<0.01 and ***p<0.001. Amp, Amplitude; Freq, frequency; Con, Control; DM, double mutant.