ROCK1 and ROCK2 inhibition alters dendritic spine morphology in hippocampal neurons

Abstract

Communication among neurons is mediated through synaptic connections between axons and dendrites, and most excitatory synapses occur on actin-rich protrusions along dendrites called dendritic spines. Dendritic spines are structurally dynamic, and synapse strength is closely correlated with spine morphology. Abnormalities in the size, shape, and number of dendritic spines are prevalent in neurologic diseases, including autism spectrum disorders, schizophrenia, and Alzheimer disease. However, therapeutic targets that influence spine morphology are lacking. Rho-associated coiled-coil containing protein kinases (ROCK) 1 and ROCK2 are potent regulators of the actin cytoskeleton and highly promising drug targets for central nervous system disorders. In this report, we addressed how pharmacologic inhibition of ROCK1 and ROCK2 affects dendritic spine morphology. Hippocampal neurons were transfected with plasmids expressing fluorescently labeled Lifeact, a small actin binding peptide, and then incubated with or without Y-27632, an established pan-ROCK small molecule inhibitor. Using an automated 3D spine morphometry analysis method, we showed that inhibition of ROCK1 and ROCK2 significantly increased the mean protrusion density and significantly reduced the mean protrusion width. A trending increase in mean protrusion length was observed following Y-27632 treatment, and novel effects were observed among spine classes. Exposure to Y-27632 significantly increased the number of filopodia and thin spines, while the numbers of stubby and mushroom spines were similar to mock-treated samples. These findings support the hypothesis that pharmacologic inhibition of ROCK1 and ROCK2 may convey therapeutic benefit for neurologic disorders that feature dendritic spine loss or aberrant structural plasticity.

Introduction

Neurons engage in a large number of contacts with other neurons through axons and dendrites. These connections, known as synapses, are the mechanism of information processing and storage in the brain.¹ In mammals, the majority of excitatory synapses occur on actin-rich protrusions along dendrites called dendritic spines. Structural plasticity of spines is tightly coordinated with synaptic function, and subtle alterations in spine biology can induce marked effects on connectivity patterns of neuronal circuits and subsequent behavior.^2,3 Aberrations in dendritic spine number, size, and shape accompany many neurologic disorders that involve deficits in cognition and information processing, including Alzheimer disease (AD), schizophrenia, and autism spectrum disorders.⁴

Synapse strength is tightly correlated with dendritic spine morphology, and over the course of life synaptic activity influences the number and shape of spines, notably in brain development, behavioral learning, and aging.^5-7 Although the functional delineation of the spine structure-synapse relationship remains elusive, a mounting body of evidence predicts that spine morphology influences excitatory neurotransmission and is critical for neuronal development and plasticity.^3,8,9 Live imaging studies have shed light on the dynamic structural plasticity of dendritic spines, indicating that spines can change size and shape over timescales of seconds to minutes and hours to days.¹⁰ Although spine morphology is grossly heterogeneous, spines can be generally classified on the basis of their 3-dimensional structure as stubby, mushroom, or thin.^11-14 Stubby spines are theorized to be transitional structures that will enlarge, possibly to mushroom spines, which represent more stable structures with a wide head and thin neck. Thin spines are more dynamic, and lack the wide, stable head indicative of mature mushroom-shaped spines. Moreover, the volume of the spine-head is directly proportional to the density of receptors at the postsynaptic tip, while a smaller spine-head size regulates calcium equilibrium by promoting efficient diffusion of calcium through the neck of the spine.^15-18 Dendritic filopodia are actin-rich protrusions that are widely considered the precursors of dendritic spines, and this hypothesis is supported by results in primary hippocampal neuron cultures that demonstrate filopodia initiate contact with axons.¹⁹ Together, these findings suggest that dendritic spine morphology can directly reflect spine function.

Filamentous actin is highly concentrated in dendritic spines, and spine shape, stability, and plasticity involves actin cytoskeleton remodeling.^20-22 Therefore, signaling cascades or protein complexes that modify actin dynamics or bind to the actin cytoskeleton are candidate regulators of spine morphology. RhoA and other Rho GTPases are extensively studied regulators of actin dynamics and heavily influence dendritic spine biology and synaptic plasticity.^23-27 Active, GTP-bound RhoA is a potent inhibitor of spine outgrowth through its principle downstream effectors, Rho-associated coiled-coil containing protein kinases (ROCK) 1 and ROCK2.^28-31 ROCK1 and ROCK2 are ubiquitous serine/threonine kinases that share 65% similarity in their amino acid sequences and 92% identity in their kinase domains.^31,32 Notably, ROCK1 and ROCK2 are highly promising drug targets for the treatment of central nervous system (CNS) disorders, including spinal cord injury, stroke, and AD.^30,33-35 Pharmacologic inhibition of ROCK1 and ROCK2 improved learning and working memory in aged rats.³⁶ Whether these beneficial effects of ROCK inhibition on cognitive behavior in rats were the result of dendritic spine alterations in the prefrontal cortex or hippocampus remains to be elucidated, however recent studies indicate that small molecule inhibitors of ROCKs can increase dendritic spine density.^37-39 In this report, we examine how ROCK1 and ROCK2 inhibition influences dendritic spine morphology in hippocampal neurons.

Results and Discussion

To assess how dendritic spine morphology is influenced by pharmacologic inhibition of ROCK1 and ROCK2, rat hippocampal neurons were isolated at E18 and cultured at high-density on glass coverslips as previously described.⁴⁰ At 16 d in vitro (DIV), neurons were transfected with plasmid encoding Lifeact-ruby, a fluorescently-tagged small actin binding peptide.⁴¹ Previous studies have demonstrated that Lifeact-expressing neurons display normal, physiological actin dynamics and dendritic spine morphology.^41,42 Additionally, images of Lifeact-expressing neurons can be used to efficiently generate automated identification and traces of dendrites, dendritic protrusions, and spines. This automated analysis uses universal parameters, needs no manual editing, and can be conducted in analysis programs such as NeuronStudio and Imaris Filament Tracer (freeware and commercially available 3D tracing software, respectively).^43,44 This methodology allows for the accurate study of dendritic spine morphology, incorporating non-disruptive neuron labeling with unbiased spine detection and measurement.⁴³

Twenty-four hours post-transfection, neurons were treated for 6 hours with H₂O (mock) or 10 µM Y-27632, a well-characterized pan-ROCK inhibitor.⁴⁵ A 6 hour time point was chosen because delivery of pan-ROCK small molecule inhibitor for 6 hours in aged rats improved learning and working memory.³⁶ Neurons were fixed and imaged on a Nikon Widefield microscope. Z-series images were acquired incrementally through the entire visible dendrite. Images were deconvolved in AutoQuant X and subjected to 3D spine morphometry analyses using Imaris Filament Tracer Surpass mode according to published methods.⁴³ A single region of interest (ROI) was selected per neuron and 12-15 neurons were evaluated per condition per experiment. Spine density and classification were measured for each ROI.⁴³ Incubation with Y-27632 dramatically altered spine morphology compared to mock treated samples (Fig. 1 and Fig. 2A). Quantitative analysis indicated that inhibition of ROCKs significantly increased the mean protrusion density (protrusions per 10 µm), which was consistent with previous findings (Fig. 2B).^37-39 Statistical analysis of cumulative distributions indicated that mean protrusion width of spines was significantly reduced following Y-27632 treatment compared to mock (Fig. 3A). A trending increase in mean protrusion length was observed in neurons exposed to Y-27632; however no global change in spine head width was observed (Fig. 3B-C).

Figure 1.

Representative maximum intensity projection deconvolved widefield microscopy images from mock or Y-27632-treated hippocampal neurons. Scale bar is 5 µm.

Figure 2.

Y-27632 increases dendritic protrusion density in hippocampal neurons. (A) Representative images from mock or Y-27632-treated neurons. Scale bar is 5 µm. (B) Mean number of dendritic protrusions per 10 µm (among all ROIs that were analyzed) are increased in Y-27632 (6.810 ± 0.53) treated neurons compared to mock (4.825 ± 0.55). *, P = 0.0162. Data representative of 3 independent experiments. All data expressed as the percentage of the mean ± SEM.

Figure 3.

Cumulative distribution analyses of protrusion width, length, and spine head width. (A) Mean protrusion width in Y-27632 treated neurons is significantly reduced. Kolmogorov-Smirnov (KS) test: mean protrusion width mock vs. Y-27632: P = 0.003. (B) Trending increase in mean protrusion length in Y-27632-treated neurons compared to mock. KS test: mean protrusion length mock vs. Y-27632: P = 0.369. (C) Mean spine head width is similar between mock and Y-27632 groups. KS test: mean spine head width mock vs. Y-27632: P = 0.672. Data representative of 3 independent experiments.

Next, numbers of protrusions classified on the basis of their structure as stubby, mushroom, or thin spines or filopodia were measured (Fig. 4A-B). This analysis indicates the changes in overall number of each spine type. Significant increases in the number of thin spines and filopodia were observed after treatment with Y-27632, whereas trending increases in the number of stubby and mushroom spines were not significant (Fig. 4C).

Figure 4.

Y-27632 increases the number and proportion of thin spines and filopodia. (A) Illustration of dendritic spine classifications (not to scale). (B) Representative image indicating examples of mushroom, stubby, and thin spines and filopodia. Scale bar is 5 µm. (C) Mean number of stubby, mushroom, or thin spines and filopodia per 10 µm (among all ROIs that were analyzed). Stubby spines (mock 1.38 ± 0.224; Y-27632 1.98 ± 0.213); mushroom spines (mock 1.93 ± 0.332; Y-27632 2.21 ± 0.29); thin spines (mock 1.21 ± 0.177; Y-27632 1.85 ± 0.179; *P = 0.0211); filopodia (mock 0.27 ± 0.06; Y-27632 0.75 ± 0.066; ***P< 0.0001). (D) Analysis of mean percent of stubby, mushroom, or thin spines and filopodia per neuron. Stubby spines (mock 28.00 ± 1.95; Y-27632 29.00 ± 2.25); mushroom spines (mock 42.33 ± 3.1; Y-27632 31.83 ± 2.47); thin spines (mock 23.33 ± 3.15; Y-27632 27.89 ± 1.88); filopodia (mock 5.917 ± 1.28; Y-27632 11.83 ± 1.3). Data representative of 3 independent experiments. All data expressed as the percentage of the mean ± SEM.

Additionally, proportion of protrusions classified on the basis of their structure as stubby, mushroom, or thin spines or filopodia were measured. Proportion or percent of protrusions reveals the relative levels of spine types per neuron or across a culture. This percent is calculated for each cell independently, and then averaged across cells. Inhibition of ROCKs decreased the proportion of mushroom spines but increased the proportion of thin spines and filopodia per neuron. No change in proportion of stubby spines was observed in Y-27632-treated samples compared to mock (Fig. 4D). In addition, the fold change in spine classes was compared by dividing the spine density per class for each of the Y-27632-treated cells by the average spine density for each class in the mock samples (Table 1). Y-27632 caused insignificant 1.43- and 1.15-fold increases in stubby and mushroom spine density, respectively, but robust 1.53- and 2.78-fold significant increases in thin spine and filopodia density, respectively. These analyses suggest that the proportion of mushroom spines decreases as a result of the dramatic increase in filopodia as well as thin spines following exposure to Y-27632.

Table 1.

Fold change in the density of spine classes per 10 µm with Y-27632 treatment in Figure 4. To calculate fold change, the spine density per class for the samples treated with Y-27632 was divided by the average spine density for each class in the mock group.

–	stubby	mushroom	thin	filopodia
Mean	1.43	1.15	1.53	2.78
SEM	0.15	0.15	0.15	0.25

These findings confirm previous observations by Kang et al. that spine density and length are increased in hippocampal neurons following treatment with Y-27632.³⁸ Kang et al treated neurons with 100 µM Y-27632 for 3 days, and our findings indicate that similar results are observed following exposure to 10 µM Y-27632 for 6 hours. Moreover, studies by Hodges et al. demonstrated that production of filopodia increased in neurons treated with 120 µM Y-27632 for 2 hours, and we show similar findings in this report using 10 µM Y-27632 for 6 hours.³⁷ In future studies, it will be important to determine whether a single treatment of ROCK inhibitor confers prolonged structural plasticity changes in cultured neurons or whether chronic ROCK inhibition is required to maintain effects on spine morphology. Furthermore, future studies will be necessary to comprehensively evaluate how inhibition of ROCKs influences the function of spines on hippocampal neurons. Notably, O'Kane et. al. demonstrated that 10 µM Y-27632 increased the magnitude of long-term potentiation in adult rat brain hippocampal slices, however whether this was due to changes in dendritic spine numbers or dendritic protrusion morphology was not addressed.⁴⁶

Dendrite and dendritic spine loss correlates with age-related memory loss and cognitive decline due to AD.^47,48 Our results coupled with previous findings indicate that inhibition of ROCKs increases spine density and generates filopodia or precursor spines.^37,38 Notably, Huentelman et. al. demonstrated that peripheral delivery of low and high dose (0.1875 mg and 0.3750 mg, respectively) hydroxyfasudil, a pan-ROCK inhibitor, improved learning and working memory in aged rats.³⁶ However, the concentration of hydroxyfasudil to reach the brain was not measured. Additionally, the molecular structures of Y-27632 and hydroxyfasudil are different and the pharmacologic properties of the compounds may not be identical.⁴⁹ Therefore, it is difficult to compare or correlate the studies by Huentelman et. al. to the work presented here.

Studies in animal models suggest that neuronal circuits that degenerate in AD are also vulnerable to cognitive impairment associated with aging. Moreover, substantial loss of dendritic spines is a shared phenotype of both memory loss due to aging and AD.⁴⁷ Using experimental models of age-related memory loss, Dumitriu et al. demonstrated that selective loss of thin spines correlated with age-related cognitive impairment in male and female rhesus monkeys.² Results presented here reveal that Y-27632 elevates the number and proportion of thin spines as well as filopodia per neuron. The increase in thin spines and filopodia, which are dynamic structures considered as precursors to mature spines, suggests that ROCK inhibition may enhance the capacity for synapse formation and structural plasticity in neurons. These findings show that pharmacologic inhibition of ROCK1 and ROCK2 may be a rational therapeutic strategy to counteract the reduction of spine density and loss of thin spines associated with age-related memory loss.

Materials and Methods

Primary neuron culture, transfection, and drug treatment

Rat hippocampal neurons were prepared from E18 embryos and cultured at high-density for widefield fluorescence microscopy imaging as previously described.⁴³ Briefly, neurons were cultured in Neurobasal medium (Invitrogen) supplemented with NS21, grown on 15 mm glass coverslips, and co-cultured with glia. At 16–17 DIV neurons were transfected with plasmid encoding Lifeact-ruby (a generous gift from Dr. Gary Bassell, Emory University School of Medicine, Atlanta, GA, USA) using Lipofectamine 2000 (Invitrogen) according to manufacturer instructions. All experiments were conducted in compliance with the National Institutes of Health Guidelines for the Care and Use of Experimental Animals and approved by the Institutional Animal Care and Use Committee at the University of Alabama at Birmingham and Emory University School of Medicine.

To inhibit ROCKs, Y-27632 (Calbiochem) was dissolved in H₂O and used at 10 µM. The vehicle control (labeled mock) was H₂O for all experiments. Twenty-four hours after transfection, neurons were treated with 10 μM Y-27632 or an equivalent volume of H₂O for 6 hrs total.

Widefield fluorescence microscopy

Hippocampal neurons were fixed with 4% paraformaldehyde in PBS, washed 3 times with PBS, and coverslips were mounted on microscope slides with propyl gallate containing polyvinyl alcohol. A Nikon Eclipse Ti microscope with a Nikon Intensilight and Photometrics Coolsnap HQ2 camera was employed to image Lifeact-ruby using a 545/30 excitation filter, a 620/60 emission filter, and a 570 dichroic. Images were captured using a 60× oil-immersion objective (Nikon Plan Apo, N.A. One.40). Z-series images were acquired at 0.15 μm increments through the entire visible dendrite.

Image processing and automated image analysis

Automated image analysis was performed exactly as previously described.⁴³ Briefly, Z-stack image series were deconvolved in AutoQuant X (MediaCybernetics) using the blind algorithm that employs an iteratively refined theoretical point spread function. For preparation of figures, maximum intensity Z-projections were generated in Imaris or ImageJ. For each image a new filament was generated using Filament Tracer (Imaris, Bitplane, Inc.) in the Autopath mode. An ROI was selected containing a dendritic segment 40–60 μm in length that was distal to a dendritic branch point and void of crossing neurites or any additional dendritic branch points. Automatic thresholds were employed to assign dendrite end points and dendrite surface rendering. The maximum spine length and minimum spine end diameter were set at 15 μm and 0.215 μm, respectively. A trace was generated and a filter was applied to ensure that all dendritic protrusions ≤ 15 μm were assigned as spines. Data were exported into Excel (Microsoft), where they were compiled and graphed.

Dendritic protrusions were classified into the following 4 groups termed stubby, mushroom, and thin spines and filopodia. These groups were defined as follows: stubby (length ≤1 μm and neck width/head width < 1.5), mushroom (neck width/head width ≥ 1.5 and length ≤5 μm), thin (1 < length ≤ 5 μm and neck width/head width < 1.5) and filapodia (5 < length ≤ 15 μm).¹⁴ Classifications for Filament Tracer were computed in Excel as previously described.⁴³

Statistical analysis

Statistics were conducted using PASW Statistics 18 (SPSS, Inc.). Datasets were analyzed for equal variance using Levene's test and normality using the Kolmogorov Smirnov test. Normally distributed data sets were compared using Student's t-test or ANOVA and post-hoc tests. The experimenter was blind to treatment during all image analysis.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors would like to thank Drs. Gary Bassell and Travis Rush for help comments.

Funding

This work was supported by the National Institutes of Health through NIA 5R00AG043552-04 to J.H.H. Additionally, this research was supported by a New Investigator Research Grant 2015-NIRG-339422 from the Alzheimer Association to J.H.H.