Research ArticleNeuroscience

Pharmacologic inhibition of LIMK1 provides dendritic spine resilience against β-amyloid

See allHide authors and affiliations

Science Signaling  25 Jun 2019:
Vol. 12, Issue 587, eaaw9318
DOI: 10.1126/scisignal.aaw9318
  • Fig. 1 ROCK1 and ROCK2 regulate dendritic spine length and density through isoform-specific mechanisms.

    (A) Representative maximum-intensity wide-field fluorescent images, after deconvolution, of hippocampal neurons expressing vector, ROCK1, or ROCK2 compared with the Lifeact-GFP control (top). Scale bar, 5 μm. 3D digital reconstructions of dendrites (bottom). Reconstructions were generated in Neurolucida 360. n = 10 to 17 neurons (one dendrite per neuron) were analyzed per experimental condition in three independent cultures. (B) Dendritic spine length in hippocampal neurons expressing vector, wild-type human ROCK1, or ROCK1-L105G and treated with blebbistatin or SR7826. Controls were transfected with Lifeact-GFP and treated with DMSO. Data are means ± SEM of three experiments. ****P < 0.0001 and *P < 0.05 (versus vector, actual P = 0.0230; P = 0.0285 versus DMSO) by one-way analysis of variance (ANOVA) with Šidák’s test. (C) Dendritic spine density in hippocampal neurons expressing vector, wild-type human ROCK2, or ROCK2-L121G and treated with blebbistatin or SR7826. Data are means ± SEM of three experiments. ****P < 0.0001, ***P < 0.001, **P < 0.01 (versus DMSO, actual #P = 0.0083), and P = 0.0207 versus SR7826 by one-way ANOVA wih Šidák’s test. Related data and analyses are shown in fig. S1.

  • Fig. 2 Aβ-induced dendritic spine degeneration is mediated by the ROCK2-LIMK1 pathway.

    (A) Representative maximum-intensity wide-field fluorescent images of hippocampal neurons after deconvolution. Scale bar, 5 μm. N = 9 to 17 neurons (one dendrite per neuron) were analyzed per experimental condition in three independent cultures. (B) Dendritic spine density in hippocampal neurons transduced with lentivirus expressing scramble (SCR) or ROCK1 (R1)–targeted, or ROCK2 (R2)–targeted shRNA and exposed to DMSO or oligomeric Aβ42 (500 nM). Data are means ± SEM of three experiments. ****P < 0.0001 (Aβ42 versus DMSO controls) and **P < 0.001 (versus SCR and Aβ42, actual #P = 0.0069) by one-way ANOVA with Šidák’s test. (C) Representative maximum-intensity wide-field fluorescent images (after deconvolution) of hippocampal neurons exposed to SR7826 (10 μM) with or without Aβ42 (500 nM). Scale bar, 5 μm. Data (right) are means ± SEM of three experiments. N = 6 to 17 neurons (one dendrite per neuron) were analyzed per experimental condition in three independent cultures. ****P < 0.0001 and **P < 0.01 (actual P = 0.0072) by one-way ANOVA with Šidák’s test. (D) Representative wide-field live-cell fluorescent images of hippocampal neurons over time, exposed to DMSO, Aβ42, SR7826, or SR7826 and Aβ42. Asterisks highlight loss (red), maintenance (yellow), or formation (green) of dendritic spines. Scale bar, 5 μm. (E) Representative spine density counts in hippocampal neurons for 6 hours with the indicated treatments. Dots represent the spine density (spines per 10 μm) for a single dendrite at 15 min intervals for 6 hours. n = 3 to 5 neurons (one dendrite per neuron) were analyzed per experimental condition in three independent cultures. Related data are shown in fig. S2.

  • Fig. 3 LIMK1 inhibition protects against Aβ-induced neuronal hyperexcitability.

    (A) Representative bright-field image of primary hippocampal neuron cultures grown on a MEA plate. (B) Representative traces (left) and raster plots from three units (right) after exposure to DMSO, Aβ42, SR7826, or SR7826 and Aβ42. N = 17 to 24 wells per condition, which includes four to six neurons per well from three independent cultures. (C and D) Mean action potential frequency (C) and mean bursts frequency (D) over baseline in hippocampal neurons treated with DMSO, Aβ42, or SR7826 with or without Aβ42. Data are means ± SEM of three experiments. ***P < 0.001 [actual P = 0.0009 (C) and 0.0003 (D)] and **P < 0.01 [actual P = 0.0062 (C) and 0.0016 (D)] by one-way ANOVA with Šidák’s test.

  • Fig. 4 LIMK1 inhibition rescues hippocampal thin spine loss in hAPP mice.

    (A and B) Representative immunoblots (A) and densitometry analysis (B) of ROCK2 and pLIMK1 protein abundance in the hippocampus of hAPPJ20 mice relative to each in NTG littermates. For densitometry, pLIMK1 was normalized to levels of LIMK1. Immunoblot for amyloid precursor protein (APP) identifies human APP in hAPPJ20 mice. Data are means ± SEM of seven mice (three males and four females per genotype). *P < 0.05 (actual P = 0.0191) and ***P < 0.001 (actual P = 0.0007) by an unpaired t test. (C) Representative maximum-intensity image of a CA1 pyramidal neuron in the hippocampus iontophoretically filled with Lucifer yellow. Blue signal is 4′,6-diamidino-2-phenylindole (DAPI). Scale bar, 50 μm. (D) Representative maximum-intensity high-resolution confocal microscope images of dye-filled dendrites, from mock- or SR7826-treated hAPPJ20 and NTG mice, after deconvolution and corresponding 3D digital reconstruction models of dendrites. Scale bar, 5 μm. Colors in digital reconstructions correspond to dendritic protrusion classes: blue, thin spines; orange, stubby spines; green, mushroom spines; and yellow, dendritic filopodia. (E and F) Mean number of apical (E) and basal (F) spines per 10 μm in dendrites from mock- or SR7826-treated hAPPJ20 and NTG mice. Data are means ± SEM. Apical dendrite conditions are as follows: N = 26 dendrites from five NTG mock mice (three females, two males), N = 26 dendrites from five NTG SR7826 mice (one female, four males), N = 33 dendrites from five hAPPJ20 mock mice (two females, three males), and N = 30 dendrites from five hAPPJ20 SR7826 mice (two females, three males) for a total of 3546 μm analyzed. Basal dendrite conditions are as follows: N = 22 dendrites from five NTG mock mice (three females, two males), N = 23 dendrites from five NTG SR7826 mice (one female, four males), N = 26 dendrites from five hAPPJ20 mock mice (two females, three males), and N = 26 dendrites from five hAPPJ20 SR7826 mice (two females, three males) for a total of 3126 μm analyzed. *P < 0.05 (actual P = 0.0313) and ***P < 0.001 (actual P = 0.0005) by one-way ANOVA with Šidák’s test. (G and H) Mean number of thin, stubby, or mushroom spines per 10 μm of apical (G) or basal (H) dendrites from mock- or SR7826-treated hAPPJ20 and NTG mice. Data are means ± SEM; N as given in (E) and (F). *P < 0.05 (actual P = 0.0398), **P < 0.01 (actual P = 0.0098), ***P < 0.001 (actual P = 0.0002), and ****P < 0.0001 by two-way ANOVA with Tukey’s test. (I and J) Mean spine length of apical (I) and basal (J) spines among CA1 pyramidal neurons in the hippocampus from mock- or SR7826-treated hAPPJ20 and NTG mice. Data are means ± SEM; N as given in (E) and (F). ***P < 0.001 (actual P = 0.0003) and ****P < 0.0001 by one-way ANOVA with Šidák’s test.

Supplementary Materials

  • stke.sciencemag.org/cgi/content/full/12/587/eaaw9318/DC1

    Fig. S1. Expression of human ROCKs in hippocampal neurons.

    Fig. S2. Aβ42-induced spine loss is prevented by fasudil.

    Fig. S3. Fasudil protects against Aβ-induced neuronal hyperexcitability.

    Fig. S4. SR7826 reduces the abundance of phosphorylated cofilin in the hippocampus.

    Fig. S5. SR7826 alters mean spine head diameter.

    Fig. S6. Aβ deposits are not substantially altered by SR7826.

  • This PDF file includes:

    • Fig. S1. Expression of human ROCKs in hippocampal neurons.
    • Fig. S2. Aβ42-induced spine loss is prevented by fasudil.
    • Fig. S3. Fasudil protects against Aβ-induced neuronal hyperexcitability.
    • Fig. S4. SR7826 reduces the abundance of phosphorylated cofilin in the hippocampus.
    • Fig. S5. SR7826 alters mean spine head diameter.
    • Fig. S6. Aβ deposits are not substantially altered by SR7826.

    [Download PDF]

Stay Connected to Science Signaling

Navigate This Article