Research ArticleNEURODEVELOPMENT

Excitatory neuron–specific SHP2-ERK signaling network regulates synaptic plasticity and memory

See allHide authors and affiliations

Science Signaling  05 Mar 2019:
Vol. 12, Issue 571, eaau5755
DOI: 10.1126/scisignal.aau5755
  • Fig. 1 Expressing SHP2D61G in excitatory neurons impairs spatial memory.

    (A) AAV constructs encoding Cre-dependent double-floxed inversed open reading frame HA-tagged αCaMKII-Cre::SHP2D61G or αCaMKII-Cre::EYFP. ITR, inverted terminal repeat sequence; WPRE, Woodchuck hepatitis virus (WHV) posttranscriptional regulatory element. (B) HA staining of SHP2D61G-expressing hippocampal slices from αCaMKII-Cre or vGAT-IRES-Cre mice. 4′,6-diamidino-2-phenylindole (DAPI) staining was used to identify nuclei. Scale bars, 100 μm. (C) Performance of αCaMKII-Cre::SHP2D61G and αCaMKII-Cre::EYFP mice in the MWM task. Data are means ± SEM from n = 7 mice per group; F1,12 = 0.01, P = 0.928 by two-way repeated measures analysis of variance (ANOVA). (D) Quadrant occupancy analysis for the probe test with mice described in (C). F3,36 = 5.459, **P < 0.01 by two-way repeated measures ANOVA with Bonferroni posttest. RQ, right quadrant; LQ, left quadrant; OQ, opposite quadrant. (E) Proximity to target platform (the average distance to the platform’s former location during the probe trial) by the mice described in (C). *P < 0.05 by unpaired t test. (F) Time exploring the relocated (new) object in OPR test by the mice described in (C), but with 13 (EYFP) and 10 (SHP2D61G) mice; **P < 0.01 and P = 0.486, respectively, compared to a hypothetical 50% (equal preference for new and old object), by one-sample t test. (G to J) As in (C) to (F) in mice overexpressing SHP2D61G in inhibitory neurons. Data are means ± SEM from n = 16 (G to I) or 21 (J) vGAT-IRES-Cre::SHP2D61G mice and n = 11 (G to I) or 17 (J) vGAT-IRES-Cre::EYFP mice. (G) F1,25 = 0.362, P = 0.553 by two-way ANOVA; (H) F3,75 = 0.257, P = 0.856 by two-way repeated measures ANOVA; (I) P = 0.523 by unpaired t test; (J) *P < 0.05 and **P < 0.01 by paired t test.

  • Fig. 2 Expressing SHP2D61G in excitatory neurons impairs LTP.

    (A) Time course of the fEPSP slope. LTP induced by theta burst stimulation (TBS; four bursts, each burst consisting of four stimuli at 100 Hz, 200-ms interburst interval) in αCaMKII-Cre::EYFP or αCaMKII-Cre::SHP2D61G slice. The fEPSP slopes were normalized to the average baseline. (B) The average fEPSP slope of 51 to 60 min after LTP induction. Average of last 10 min of LTP, αCaMKII-Cre::EYFP, n = 16 slices from 10 mice; αCaMKII-Cre::SHP2D61G, n = 16 slices from 10 mice; unpaired t test, ***P < 0.001. (C) Time course of the fEPSP slope. LTP induced by TBS in vGAT-IRES-Cre::SHP2D61G or vGAT-IRES-Cre::EYFP slices. The fEPSP slopes were normalized to the average baseline. (D) The average fEPSP slope of 51 to 60 min after LTP induction. vGAT-IRES-Cre::EYFP, n = 9 slices from seven mice; vGAT-IRES-Cre::SHP2D61G, n = 13 slices from eight mice; unpaired t test, P = 0.523.

  • Fig. 3 SHP2D61G selectively activates RAS-ERK signaling in excitatory neurons.

    (A) Representative IHC images from αCaMKII-Cre::SHP2D61G-HA and αCaMKII-Cre::EYFP mice. Slices were immunostained for p-ERK1/2 (red) and HA (green). Arrows indicate double labeling of p-ERK1/2 and SHP2D61G-HA and double labeling of p-ERK1/2 and EYFP. Higher-magnification images of boxed CA1 region are also shown (the fourth column). (B) Proportion of hippocampal neurons from the mice described in (A) that were p-ERK positive. αCaMKII-Cre::EYFP, n = 14 slices from four hippocampi; αCaMKII-Cre::SHP2D61G, n = 15 slices from four hippocampi; unpaired t test, **P < 0.01. (C and D) As described in (A) and (B) for vGAT-IRES-Cre::SHP2D61G and vGAT-IRES-Cre::EYFP hippocampi. n = 8 and 6 slices, respectively, from four hippocampi; unpaired t test, P = 0.104. (E and F) As described in (A) and (B) for vGAT-IRES-Cre::KRASG12V and vGAT-IRES-Cre::EYFP mouse hippocampal CA1 region. n = 6 and 7 slices, respectively, from four hippocampi; unpaired t test, ***P < 0.005. Scale bars, 20 μm. Data are means ± SEM.

  • Fig. 4 Cell type–specific transcriptome analyses reveal differential expressions of RAS signaling molecules.

    (A) Workflow for cell type–specific transcriptome analysis. cDNA, complementary DNA. (B) Scatterplot illustrating genes enriched in αCaMKII+ neurons (red, 1679 transcripts) or in vGAT+ neurons (blue, 1803 transcripts) out of 11,554 transcripts. (C) Unsupervised hierarchical clustering analysis based on Pearson’s correlation of normalized fragments per kilobase million (FPKM) values shows clear segregation between two cell types. (D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of 3482 DEGs. Venn diagram indicates a comparison of KEGG pathways enriched in two neuronal types. The bar graph indicates the top five KEGG pathways that are enriched in αCaMKII+ neurons (red bar) and in vGAT+ neurons (blue bar). The numbers of genes in each pathway are indicated in the bars. (E) Validation of RNA-seq data. Expression of genes were represented by log2 FC (fold change, αCaMKII/vGAT) of FPKM value and by log2 FC of ΔCt values normalized to β-actin level. The qRT-PCR and RNA-seq results represent the means of biological duplicates, which have technical triplicates. Red shading indicates enriched genes in αCaMKII+ neurons, and blue shading indicates enriched genes in vGAT+ neurons. (F) A schematic of RAS-ERK signaling in excitatory and inhibitory neurons based on the transcriptome data (genes take the place of proteins in the pathway). Genes in red are enriched in αCaMKII+ neurons, genes in blue are enriched in vGAT+ neurons, and gray represents similarly expressed genes.

  • Fig. 5 Coexpressing GAB1Y627F reverses the SHP2D61G-mediated ERK hyperactivation in excitatory neurons.

    (A) Immunostaining for p-ERK1/2 (red), HA (green), and MYC (white) in EYFP, GAB1Y627F-MYC, SHP2D61G-HA, and SHP2D61G-HA/GAB1Y627F-MYC expressing hippocampal slices from αCaMKII-Cre mice. Arrows indicate double labeling of p-ERK1/2 and EYFP, GAB1Y627F-MYC, SHP2D61G-HA, or SHP2D61G-HA/GAB1Y627F-MYC. Scale bars, 20 μm. (B) Proportion of p-ERK–positive neurons in each of the samples described in (A). Data are means ± SEM. EYFP, n = 11 slices from 10 hippocampi; GAB1Y627F, n = 10 slices from 10 hippocampi; SHP2D61G, n = 29 slices from 26 hippocampi, SHP2D61G/GAB1Y627F, n = 10 slices from 12 hippocampi; one-way ANOVA (P < 0.001) with Bonferroni’s multiple comparison test, *P < 0.05, **P < 0.01, and ***P < 0.005. (C and D) Representative IHC images, staining for p-ERK1/2 (red), HA (SHP2D61G; magenta), MYC (GAB1; cyan), and green fluorescent protein (GFP; GRB2, green) in hippocampal slices from vGAT-IRES-Cre::EYFP, vGAT-IRES-Cre::SHP2D61G/GAB1WT, and vGAT-IRES-Cre::SHP2D61G/GAB1WT/GRB2WT mice. Arrows indicate quadruple labeling of p-ERK1/2, SHP2D61G-HA, GAB1-MYC, and GRB2-GFP. Scale bars, 40 μm. (D) Proportion of p-ERK–positive neurons in each of the samples described in (C). Data are means ± SEM. vGAT-IRES-Cre::EYFP, n = 10 slices from four hippocampi; vGAT-IRES-Cre::SHP2D61G/GAB1WT, n = 10 slices from six hippocampi; vGAT-IRES-Cre::SHP2D61G/GAB1WT/GRB2WT, n = 12 slices from six hippocampi; one-way ANOVA, F2,29 = 4.235; P < 0.05, unpaired t test, *P < 0.05; EYFP versus SHP2D61G/GAB1WT, P = 0.916.

  • Fig. 6 GAB1Y627F coexpression in excitatory neurons restores SHP2D61G-mediated LTP and memory deficits.

    (A) LTP, as assessed by the time course (top) of the fEPSP slope (bottom), in hippocampal slices from αCaMKII-Cre mice expressing singly or coexpressing GAB1Y627F and SHP2D61G. Scale bars, 0.5 ms and 5 mV. Data are mean ± SEM; αCaMKII-Cre::EYFP, n = 9 slices from five mice; αCaMKII-Cre::GAB1Y627F, n = 4 slices from three mice; αCaMKII-Cre::SHP2D61G, n = 6 slices from three mice; αCaMKII-Cre::SHP2D61G/GAB1Y627F, n = 11 slices from six mice. (B) The average fEPSP slope of 51 to 60 min after LTP induction shown in (A). Two-way ANOVA with Bonferroni posttest, *P < 0.05 and **P < 0.01. (C) AMPA/NMDA ratio in SHP2D61G-expressing excitatory neurons and those coexpressing GAB1Y627F. Averages of 15 consecutive responses obtained at −70 mV (AMPA) and + 40 mV (NMDA) were used for the AMPA/NMDA ratio calculation. αCaMKII-Cre::EYFP, n = 27 cells from four mice; αCaMKII-Cre::SHP2D61G, n = 41 cells from five mice, αCaMKII-Cre::SHP2D61G /GAB1Y627F, n = 32 cells from four mice. Data are means ± SEM. One-way ANOVA with Bonferroni posttest, *P < 0.05 and **P < 0.01. (D) Latency to the platform in the MWM task during training trials. αCaMKII-Cre::EYFP, n = 8 mice; αCaMKII-Cre::GAB1Y627F, n = 10 mice; αCaMKII-Cre::SHP2D61G, n = 14 mice; and αCaMKII-Cre::SHP2D61G/GAB1Y627F, n = 12 mice. Data are means ± SEM. Two-way repeated measures ANOVA, P = 0.907. (E) Quadrant occupancy analysis for the probe test in mice expressing SHP2D61G and those coexpressing SHP2D61G and GAB1Y627F in hippocampal neurons. Data are means ± SEM. αCaMKII-Cre::EYFP, n = 8 mice; αCaMKII-Cre::GAB1Y627F, n = 10 mice; αCaMKII-Cre::SHP2D61G, n = 14 mice; αCaMKII-Cre::SHP2D61G/GAB1Y627F, n = 12 mice; two-way repeated measures ANOVA with Bonferroni posttest (time spent in TQ), *P < 0.05 and ***P < 0.005. (F) Proximity to target platform occupied by αCaMKII-Cre::GAB1Y627F mice (n = 10), αCaMKII-Cre::SHP2D61G mice (n = 14), αCaMKII-Cre::SHP2D61G/GAB1Y627F mice (n = 12), and an EYFP control group (n = 8). Data are means ± SEM. One-way ANOVA (P < 0.005) with Bonferroni’s multiple comparison test, *P < 0.05 and **P < 0.01.

Supplementary Materials

  • www.sciencesignaling.org/cgi/content/full/12/571/eaau5755/DC1

    Fig. S1. The second probe trials after extended trainings.

    Fig. S2. Short-term memory test in object place recognition.

    Fig. S3. Effects of expressing SHP2D61G in excitatory or inhibitory neurons on basal synaptic transmission and PPF ratio.

    Fig. S4. Effect of SHP2D61G on ERK activation in PV+ neurons.

    Fig. S5. The total number of p-ERK1/2+ neurons and viral vector–expressing cells were not significantly different between EYFP- and SHP2D61G-infected hippocampi.

    Fig. S6. Validation of the quality of cell sorting and bioinformatic workflow.

    Fig. S7. Comparison of GAB1 protein abundance in vGAT+ and vGAT neurons in vGAT-Cre;tdTomato mice.

    Fig. S8. The effect of GAB1Y627F on the interaction of SHP2D61G with GAB1 and ERK activation.

    Fig. S9. The second probe trials after extended trainings in rescue experiments.

    Table S1. List of 3482 DEGs.

    Table S2. Functional annotation of 3482 DEGs.

    Table S3. Expression profile of RASopathy-associated genes.

    Table S4. Primer sequences for qRT-PCR validation.

  • The PDF file includes:

    • Fig. S1. The second probe trials after extended trainings.
    • Fig. S2. Short-term memory test in object place recognition.
    • Fig. S3. Effects of expressing SHP2D61G in excitatory or inhibitory neurons on basal synaptic transmission and PPF ratio.
    • Fig. S4. Effect of SHP2D61G on ERK activation in PV+ neurons.
    • Fig. S5. The total number of p-ERK1/2+ neurons and viral vector–expressing cells were not significantly different between EYFP- and SHP2D61G-infected hippocampi.
    • Fig. S6. Validation of the quality of cell sorting and bioinformatic workflow.
    • Fig. S7. Comparison of GAB1 protein abundance in vGAT+ and vGAT neurons in vGAT-Cre;tdTomato mice.
    • Fig. S8. The effect of GAB1Y627F on the interaction of SHP2D61G with GAB1 and ERK activation.
    • Fig. S9. The second probe trials after extended trainings in rescue experiments.
    • Legends for tables S1 to S4

    [Download PDF]

    Other Supplementary Material for this manuscript includes the following:

    • Table S1 (Microsoft Excel format). List of 3482 DEGs.
    • Table S2 (Microsoft Excel format). Functional annotation of 3482 DEGs.
    • Table S3 (Microsoft Excel format). Expression profile of RASopathy-associated genes.
    • Table S4 (Microsoft Excel format). Primer sequences for qRT-PCR validation.

Stay Connected to Science Signaling

Navigate This Article