C-terminal fragments of amyloid precursor proteins increase cofilin phosphorylation by LIM kinase in cultured rat primary neurons
Amyloid precursor proteins (APPs) are processed by β-, γ-, and ε-secretases and caspase-3 to generate C-terminal fragments of APP (APP-CTFs), which may contribute to the pathology of Alzheimer’s disease (AD). In addition to amyloid plaques and neurofibrillary tangles, AD brains contain Hirano bodies, which are rod-like structures mostly composed of actin and the actin-binding protein, cofilin.
However, the mechanisms underlying the formation of cofilin-actin rods are still unknown. In this study, we aim to elucidate the effects of APP-CTFs on the actin- depolymerizing factor [(ADF)/cofilin]. Our data indicate that transfection with APP-CT99 and APP-CT57 may increase the phosphorylation level of Ser3 of ADF/cofilin and Thr508 of LIM-kinase 1 in rat primary cortical neuronal cultures. S3 peptide, a synthetic peptide competitor of LIM-kinase 1 for ADF/cofilin phosphorylation and an inhibitor of APP-CTFs, induced ADF/cofilin phosphorylation. In comparison with the wild-type mouse, the APP-CT transgenic mouse showed increased immunoreactivity of phosphorylated cofilin (p-cofilin) in the brain. Treatment with DAPT, an inhibitor of γ-secretase, resulted in a decrease in p-cofilin protein level in the group transfected with full-length APP-695. Transfection with the mutant APP-CTF with a deleted YENPTY domain resulted in no significant increase in p-cofilin level. Thus, APP-CTFs induced cofilin phosphorylation to facilitate nuclear translocation. These results suggest a relationship between APP-CTFs and ADF/ cofilin that may be suggestive of a new toxic pathway in the pathology of AD.
Introduction
The pathogenesis and mechanism of Alzheimer’s disease (AD) are still incompletely understood. Various factors such as cholinergic system dysfunction, inflammation or immune response, genetic mutation, role of amyloid precursor protein (APP), oxidative stress and excitotoxi- city, and apoptosis have been hypothesized to be asso- ciated with AD progression [1–3]. Of these, the role of APP has always received much attention [4–6].
APP, a cell surface protein of unknown function, is implicated in AD pathogenesis [7]. APP topology resembles that of a membrane receptor protein: it has a large extracellular portion, single transmembrane seg- ment, and cytoplasmic tail domain, which interacts with several proteins, including Fe65 [5]. Although the func- tion of APP is not understood completely, it has been implicated in various processes, including signal trans- duction, cell migration, and axonal elongation [7].
APP is cleaved by β-secretase or α-secretase, which sheds the extracellular portion and generates membrane- associated COOH-terminal fragments (APP-CTFs, C99 and C83; C99 includes Aβ domain and C57) that are further cleaved by γ-secretase within the plane of the membrane [7]. The action of γ-secretase results in the extracellular secretion of P3 or 40/42 residue-long Aβ peptides along with the simultaneous release of the APP intracellular domain (AICD, C57, C50) within the cell [3]. APP is also cleaved by ε-secretase and caspases,
wherein each of these enzymes generates C50 and C31 [2–4]. APP-CTFs contribute to the pathology of AD [4,5, 8–11].
APP-CTFs (C57 and C31) translocate to the nucleus to form a ternary complex with Fe65 and CP2/LSF/LBP1, followed by an increase in glycogen synthase kinase 3β (GSK-3β) mRNA and protein levels. Consequences of these events include tau phosphorylation and apoptosis.The deletion of the YENPTY domain of mutated APP- CTFs (C57 and C31) may prevent the nuclear translo- cation of the protein [9–11]; the function of AICD and its role in AD pathology are still largely unknown.
In addition to amyloid plaques and neurofibrillary tangles, AD brains contain Hirano bodies, which are rod-like structures mostly composed of actin and the actin- binding protein, cofilin [12–14]. Cofilin and its family members are ubiquitously expressed proteins, best char-
acterized as actin-binding and actin-modulating proteins. These proteins have now been shown to be involved in multiple facets of cellular biology, independent of their function with respect to actin tread-milling. These func- tions are as diverse as their involvement in membrane and lipid metabolism, mitochondrial-dependent apoptosis, and regulation of transcription and chromatin structure [15–17]. Cofilin 1, the nonmuscle-specific isoform of cofilin, is of particular interest to the field of neurodegeneration. Studies have shown the involvement of nuclear cofilin- actin rods in a Huntington’s disease (HD) cell model and changes in cofilin protein expression profile in lym- phocytes from patients with HD; furthermore, the cyto-plasmic cofilin-actin rods are involved in AD progression [18–21]. Improper regulation of the actin cytoskeleton is thought to be related to neurodegenerative diseases, as actin dynamics are critical for the maintenance of healthy synapses and dendrites [22]. Thus, the understanding of actin dynamics in disease conditions and aging-related stress will give important insights into the mechanisms underlying neurodegeneration [23]. Treatment of hippo- campal neurons with Aβ increases the phosphorylation level of Ser3 of actin-depolymerizing factor (ADF)/cofilin and Thr508 of LIM-kinase 1 (LIMK1) accompanied with the dramatic remodeling of actin filaments, neuritic dys- trophy, and neuronal cell death [24,25]. A recent study suggested the formation of tau neuropil threads in response to abnormal cofilin aggregation [26–28]. Thus, these cofilin-actin inclusions may play an essential role in the pathogenesis of AD. However, the underlying mechanisms of cofilin-actin rod formations are still largely unknown [29]. This study investigates the effect of APP-CTFs on ADF/cofilin phosphorylation to reveal a new toxic pathway underlying AD pathogenesis.
Methods
Materials
DNA constructs and mutagenesis pEGFP-N1-APP-CTFs (C99, C57) and empty vector EGFP-N1 and the deletion mutants of C99 and C57 without YENPTY domain (#C99 and #C57, respec- tively), as well as Flag-APP-695 and Flag empty vector, were kindly provided by Dr Yoo-Hun Suh, College of Medicine, Department of Pharmacology, Seoul National University. The cDNA constructs encoding different lengths of APP carboxyl terminus were designed as shown in Fig. 1. The constructs were generated by PCR from the human APP-695 cDNA to encompass the last 5799 amino acid residues. The deletion mutants #C57 and #C99 were prepared using ExSite PCR mutagenesis kit (Stratagene, La Jolla, California, USA).
Amyloid precursor protein-C-terminal transgenic mouse brain tissues
APPV717I-CT100 transgenic mouse (APP-CT Tg mouse) is an animal model of AD. APP-CT Tg mouse is gener- ated by overexpressing the human APP-CT100 gene carrying the London mutation (V717I). The genotyping, cognitive deficits, and pathological features were pre- viously described [30]. All experiments were performed Schematic representation of the constructs of C99, C57, #C99, and #C57. AICD is generated by γ-secretase cleavage of C99 (major product of β-cleavage). cDNA constructs encoding different lengths of APP carboxyl terminus were as follows: C99 and C57 as well as the YENPTY deletion variant of C99 (#C99) and C57 (#C57). Deletion mutants of #C99 and #C57 were generated using ExSite PCR mutagenesis kit (Stratagene). Constructs were cloned into the pEGFP-N1 vector with GFP at the carboxyl terminus (C99-GFP and #C99-GFP, C57-GFP and #C57-GFP). APP, amyloid precursor protein.
Reagents
Following reagents and antibodies were purchased as sta- ted and used for this study: Trypsin (Sigma, New York, New York, USA); fetal bovine serum (Corning, New York, New York, USA); Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies Inc., New York, New York,
USA); plasmid prep Maxi kit (Qiagen, Hilden, Germany); anti-Ser3 phosphorylated cofilin (p-cofilin) (AB3831; Merck Millipore, Darmstadt, Germany); polyclonal anti- phosphorylation level of LIM kinase (p-LIMK) (07-850; Merck Millipore). Western blotting blocking solution, washing solution, and all secondary antibodies were pur- chased from Beyotime Institute of Biotechnology. S3 peptide, which contains the unique Ser3 phosphorylation site of ADF/cofilin, was used as a specific competitor substrate for active LIMK1. The structure of S3 peptide was MASGVAVSDGVIKVFN RQIKWFQNRRMKWKK (Shanghai Biopeptide, Shanghai, China). The S3 peptide was dissolved in sterile double-distilled water at a con- centration of 2.5 mM and added to the cultures at indicated final concentrations.
Experimental methods
Neuronal cultures
All animal studies were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Yanbin University Institutional Animal Care and Use Committee. Primary cortical neuronal cultures were established from embryonic fetuses (day 18–19) as previously described [10]. High-density cultures (2500 cells/ mm2) were used for biochemical analysis; low-density cultures (100 cells/mm2) were used for immunocy- tochemistry. The cells were plated in DMEM (Invitrogen, Gaithersburg, Maryland, USA) plus 10% horse serum (HyClone, Logan, Utah, USA) on poly- L-lysine (1 mg/ml)-coated dishes or coverslips; after 2 h, the medium was replaced with glia-conditioned DMEM supplemented with N2 and B27 (Invitrogen). The cultures were maintained at 37°C in a 5% CO2 humidified atmosphere.
Plasmid transfection and expression of primary cortical neurons
The transfection procedure was completed according to the specifications of the Lipofecter Plasmid 2000 (Baiaocaibo Inc., Beijing, China) transfection reagent with pEGFP-N1 empty vector, APP-CT99, or APP- CT57 construct. These transfected cells were observed under a fluorescence microscope.
Western blot analysis
Cells were washed with PBS and lysed in radio- immunoprecipitation assay buffer on ice. Proteins were separated by SDS polyacrylamide gel electrophoresis and transferred onto a polyvinylidene fluoride membrane. The membrane was blocked with Tris-buffered saline for 1 h at room temperature, and the blot was treated with anti-Ser3 p-cofilin (1 : 1000; AB3831; Merck Millipore) and polyclonal anti-p-LIMK (1 : 200; 07-850; Merck Millipore). Detection was performed using a horseradish peroxidase-conjugated secondary antibody. The expression level of the protein was normalized to β-tubulin expression and the ratio versus mock control was calculated.
Immunocytochemistry
Cells were fixed in 4% paraformaldehyde for 1 h at 4°C and then incubated for 30 min at room temperature. After treating the cells with 50 mM NH4Cl-PBS for 10 min, cells were washed in PBS and permeabilized for 30 min at room temperature in PBS containing 0.1% Triton X-100 and 1 mg/ml of bovine serum albumin (permea- bilization buffer). The following steps were performed at room temperature in permeabilization buffer. Cells were incubated with primary antibodies for 1 h. After three washes, primary antibodies were detected by incubating the cells for 1 h with Alexa Fluor 555-conjugated sec- ondary antibodies (Thermo Fisher Scientific Inc., Massachusetts, USA). After three washes in permeabili- zation buffer and one wash in PBS, cells were subjected to confocal microscopy (Leica 8751, Leica Biosystems, Shanghai Agency, Shanghai, China).
Immunohistochemistry
For immunohistochemical staining, brain tissue slides were deparaffinized and hydrated. After washing twice with distilled water for 5 min each, the slides were washed with TBS (0.05 M, pH 7.6) for 10 min and sub- jected to antigen retrieval with 10 mM citrate buffer (pH 6.0) for 1 h. Nonspecific binding of antibodies was eliminated using a blocking buffer (0.5% Triton-X with 20 μl/ml goat serum in PBS) for 5 min (thrice), followed by overnight incubation at 4°C with anti-Ser3 p-cofilin (1 : 500 dilution). Detection was performed using horse-
radish peroxidase -conjugated goat anti-mouse IgG antibody (1 : 500; Vector Laboratories, Burlingame, California, USA). Antibody reactivity was visualized using the VECTASTAIN ABC Elite Kit (Vector Laboratories).
Statistical analysis
Statistical analysis was performed using SPSS statistical software 17.0 (SPSS Inc., Chigago, USA) to study the relationship between the different variables (analysis of variance). Values of P less than 0.05 were considered to indicate statistical significance. The quantitative data were indicated as mean ± SEM.
Results
Amyloid precursor protein-C-terminal fragment expression increases the phosphorylation of cofilin
To determine whether APP-CTF overexpression influences the phosphorylation of cofilin, APP-CT99 and APP-CT57 were overexpressed in rat primary cortical neurons. Immunocytochemical analysis and western blotting results showed that p-cofilin levels were higher in the cells transfected with APP-CT99 and APP-CT57 as compared with those in the mock group (Figs 2a and c and 3a and b). Representative blots along with quanti- tative analyses are presented in Fig. 3.
Amyloid precursor protein-C-terminal fragment-induced cofilin phosphorylation is dependent on the LIMI kinase 1-mediated pathway
LIM kinases may induce inactivation of ADF/cofilin through its phosphorylation at Ser3 [31]. LIMK1, a LIMK family member predominantly expressed in the nervous system, is activated by phosphorylation at Thr508 [32]. To investigate whether LIMK1 activity is altered by APP-CTFs, we used an antibody that selec- tively recognizes LIMK1 phosphorylation at Thr508. Transfection of rat cortical primary neurons with APP- CT99 and APP-CT57 resulted in an increase in the p-LIMK1, as evidenced by immunofluorescence and western blotting analyses (Figs 2b and d and 3a and c).
APP-CTFs induce phosphorylation of cofilin and LIMK1. Primary cortical neurons were transfected with pEGFP (mock) or APP-CTFs-pEGFP (C99 or C57) for 48 h. (a) C99 or C57 was visualized by GFP signal and p-cofilin expression was detected by fluorescence staining with p-cofilin antibody and Alexa Fluor 555-conjugated secondary antibody. (b) p-LIMK expression was detected with p-LIMK antibody and Alexa Fluor 555-conjugated secondary antibody. Neurons were observed under a confocal microscope. (c, d) Quantitative analysis of fluorescence images (c, for p-cofilin; d, for p-LIMK). APP, amyloid precursor protein; CTF, C-terminal fragment; p-cofilin, phosphorylated cofilin.*P < 0.05 as compared with the mock-transfected group. To determine whether the inactivation of ADF/cofilin by endogenous LIMK1 is essential for APP-CTF-mediated phosphorylation of cofilin, a synthetic peptide S3 was used as a specific competitor substrate of active LIMK1. The S3 peptide is a synthetic peptide (MASGVAVSDGVIKVFN RQIKIWFQNRRMKWKK) that contains a unique Ser3 phosphorylation site of ADF/ cofilin and a penetrating sequence that facilitates its internalization in neurons. This peptide has been suc- cessfully used to prevent LIMK1-mediated phosphor- ylation of ADF/cofilin after semaphorin 3 A treatment [33]. As expected, addition of S3 peptide to rat cortical neuron cultures significantly eliminated the increase in p-cofilin level induced by APP-CTFs (Fig. 4). These experiments indicate that APP-CTF-induced cofilin phosphorylation requires LIMIK1-mediated inactivation of ADF/cofilin. Downregulation of amyloid precursor protein-C-terminal fragments decreases cofilin phosphorylation To investigate whether downregulation of APP-CTFs affects the phosphorylation of cofilin, we tested the effects of DAPT, an inhibitor of γ-secretase, on p-cofilin levels. Transfection of cells with the full-length APP-695 resulted in a decrease in the level of p-cofilin in DAPT treatment group than in the vehicle-treated control group (Fig. 5). These results indicate that the generation of CTFs is critical for the phosphorylation of cofilin. Nuclear translocation of amyloid precursor protein-C- terminal fragments is essential for cofilin phosphorylation To investigate if APP-CTF-induced cofilin phosphorylation is related to nuclear translocation of APP-CTFs, we transfected the deletion mutant of APP-CTFs in APP-CTFs induce phosphorylation of cofilin and LIMK1. Primary cortical neurons were transfected with pEGFP (mock) or APP-CTFs-pEGFP (C99 or C57) for 48 h. (a) Protein expression levels of phosphorylated cofilin (p-cofilin) and cofilin as well as phosphorylated LIMK1 (p-LIMK1) and LIMK1 were detected by western blotting. (b, c). Data are represented as the ratio of p-cofilin versus cofilin and p-LIMK versus LIMK.*P < 0.05 as compared with the mock-transfected group. Data are expressed as mean ± SD for triplicate observation (one-way analysis of variance and Dunnett test). APP,amyloid precursor protein; CTF, C-terminal fragment. Treatment of hippocampal neurons with Aβ results in an increase in the phosphorylation level of Ser3 of ADF/ cofilin and Thr508 of LIMK1 (p-LIMK1) accompanied with dramatic remodeling of actin filaments, neuritic dystrophy, and neuronal cell death [25]. S3 peptide, a synthetic peptide that acts as a specific competitor of LIMK1 for ADF/cofilin phosphorylation, inhibited Aβ- induced ADF/cofilin phosphorylation and prevented actin filament remodeling and neuronal degeneration, indicative of LIMK1 involvement in Aβ-induced neuro- nal degeneration in vitro [25]. Immunofluorescence ana- lysis of AD brain showed a significant increase in the number of p-LIMK1-positive neurons in areas affected with AD pathology [25]. The C99 domain was processed by endogenous γ-secretase to generate Aβ1–42 and C57 or Aβ1–40 and C59. Both Aβ and CTFs are toxic abnor- mal proteins that contribute to AD pathology [4,5,10,11]. We evaluated the changes in p-cofilin level in cultured rat cortical cells in the presence of full-length APP-695 with or without the γ-secretase inhibitor, DAPT, to determine the effects of the endogenously generated APP-CTFs on p-cofilin expression. We failed to observe any significant increase in p-cofilin level in the DAPT- treated group as compared with the control group, sug- gesting that APP-CTF generation is critical for induction of cofilin phosphorylation. APP-CTFs were found to translocate to the nucleus to form a ternary complex with Fe65 and CP2/LSF/LBP1. This phenomenon was followed by an increase in GSK- 3β mRNA and protein levels, inducing tau phosphoryla- tion and apoptosis [5,8,10]. No significant increase was observed in the level of p-cofilin upon transfection of cells with the deletion mutant of APP-CTFs lacking the YENPTY domain. These data indicate that translocation of APP-CTFs to the nucleus is very important for indu- cing cofilin phosphorylation. A recent study showed that cofilin protein is regulated by highly conserved nuclear import and export signals [17] and that these signals function for the formation of appropriate rods during stress [28]. These studies indi- cate that the normal response of the cofilin-actin rod is likely integral for the proper functioning of cells in higher order organisms. A recent study hypothesized the potential physiological function of nuclear cofilin-actin rods and the reason for its dysregulated response, which could lead to the selective vulnerability of the most susceptible cell populations in HD [18]. The presence of cofilin rods/aggregates correlates with the extent of tau pathology, independent of patient age [28]. We hypothesize a new toxic pathway related to the translocation of APP-CTFs (C57) to the nucleus that affects nuclear cofilin-actin rods and induces cytoskeletal cell stress, eventually contributing to neurodegenerative disorders. Further studies are warranted to evaluate the in-depth molecular and biological mechanisms. Results of the present study suggest the existence of a relationship between APP-CTFs and ADF/cofilin and highlight a newly revealed toxic pathway in the pathol- ogy of AD. This study may further elucidate the inter- relationships between different pathologies of AD. APP- CTFs co-contribute to the pathology of AD with APLP2 (amyloid precursor like protein2) -CTF and Aβ; the in- depth molecular and biological mechanism underlying CRT-0105446 this needs to be studied further.