Abstract
Neuronal ELAV/Hu (nELAV) are RNA-binding proteins that mainly regulate gene expression by increasing the stability and/or translation rate of target mRNAs bearing ARE (adenine and uracil-rich elements) sequences. Among nELAV target transcripts there is ADAM10, an α-secretase involved in the non-amyloidogenic processing of the amyloid-β protein precursor (AβPP) which leads to the production of the neuroprotective sAβPPα peptide. The aim of this study was to evaluate if nELAV depletion affects ADAM10 expression in human SH-SY5Y neuroblastoma cells. We also studied the effects of Bryostatin-1, a molecule able to activate nELAV protein cascade. The specific HuD/nELAV gene silencing decreased both nELAV and ADAM10 protein contents; similar results were obtained by Aβ40 treatment in wild-type SH-SY5Y cells. In HuD-silenced cells, the exposure to Bryostatin-1 counteracted both nELAV and ADAM10 proteins downregulation, by restoring nELAV/ADAM10 basal levels. We also found that sAβPPα release, which seemed not to be compromised by Aβ40 challenge or HuD-silencing, was favored by Bryostatin-1. Overall, these findings strongly suggest that a deficiency in nELAV content negatively affects ADAM10 expression and may play a role in neurodegenerative diseases, which may benefit by molecules activating ELAV cascade.
INTRODUCTION
RNA-binding proteins (RBP) are emerging as key factors in the regulation of gene expression at post-transcriptional level not only for the proper development of the nervous system, but also for the maintenance of neural functions and activities [1]. RBP have been described to be involved in neuronal differentiation, in the formation of the extensive network of projections and connections among the different neurons, and in the maintenance of key neuronal functions, brain development and neuronal plasticity [2]. They act as trans-acting factors on specific target transcripts and control crucial steps of the mRNA life, from splicing, polyadenylation, and nuclear export to cytoplasmic localization, stability, and translation [3]. Consequently, it is not surprising that a derangement of RBP, and hence of RNA metabolism, is involved in the pathogenesis of neurodegenerative diseases [4, 5].
The ELAV (Embryonic Lethal Abnormal Vision) RBP family in vertebrates is composed of three neuron-specific members, HuB, HuC, and HuD (nELAV), whereas HuR is ubiquitously expressed [6, 7]. All four proteins are highly homologous in sequence, and contain three ≈ 90-aminoacid-long RNA recognition motifs (RRM) [8]. ELAV proteins act post-transcriptionally by binding to adenine and uracil-rich elements (ARE) preferentially found in the 3’-untranslated region (3’-UTR) of target mRNAs, and mainly enhance gene expression by increasing the cytoplasmic stability and/or rate of translation of these ARE-containing mRNAs [9]. Focusing on the HuD/nELAV protein, literature data demonstrate its physiological role in controlling gene expression in spatial memory, learning, synaptogenesis, and neuronal differentiation [10–12]. Thus far, a variety of HuD target mRNAs have been identified, including neuronal structural proteins such as the growth-associated protein-43 (GAP-43) [13, 14], tau [15], and the myristoylated alanine-rich C-kinase substrate (MARCKS) [16]; transcription factors such as N-myc [17] and c-myc [18]; cellular signaling regulators such as neuroserpin [19], p21waf1 [20], vascular endothelial growth factor (VEGF) [21], acetylcholinesterase (AChE) [22], and other neuronal RBP such as Msi-1 [23] and Nova1 [24].
Since the HuD RBP acts as an important regulator of neuronal differentiation and function, a derangement in its activity may have implications in neurodegenerative pathologies characterized by loss of memory and neuronal dysfunctions such as Alzheimer’s disease (AD). Within this context, we previously documented that nELAV protein levels decrease along with clinical dementia progression in regions involved in the memory circuitry, such as the hippocampus, and that amyloid-β (Aβ) itself may have direct action on nELAV pathway [25]. In particular, among the genes involved in AD pathogenesis, we found that ADAM10, a member of the ADAM (A Disintegrin And Metalloproteinase) family of integral membrane proteins acting as constitutive α-secretases on the amyloid-β protein precursor (AβPP) [26, 27], represents a target of nELAV RBP, whose binding is disrupted by Aβ42 in vitro [25]. These observations prompted us: 1) to investigate if also the more physiological and abundant Aβ40 is able to affect the HuD/nELAV cascade; 2) to study possible downstream changes in ADAM10 protein content and sAβPPα release in an ad hoc established cellular model characterized by a consistent HuD/nELAV downregulation. Moreover, since we previously proved the existence of a pathway involving protein kinase C α (PKCα) as a regulator of nELAV protein recruitment and activity [10], we also explored the effect of PKCα activation, mediated by Bryostatin-1, on the HuD/ADAM10 cascade in HuD-silenced neuronal cells.
MATERIALS AND METHODS
Cell cultures and in vitro treatments
Human neuroblastoma SH-SY5Y cells were purchased by the American Type Culture Collection (ATCC, Manassas, VA, USA) and grown in Eagle’s minimum essential medium (MEM) supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin and 100μg/ml streptomycin, 1% non-essential amino acids and sodium pyruvate (1 mM) at 37°C in an atmosphere of 5% CO2 and 95% humidity. SH-SY5Y cells (falling within 23 - 28 passages) were exposed to the solvent (DMSO) or to the following reagents for the indicated times: freshly dissolved 1μM Aβ40 (as monomer, without pre-formation of oligomers or fibrils) or the reverse Aβ40 - 1 peptide (Sigma, Milan, Italy) for 24 h; 100 nM Bryostatin-1 (Sigma) for 15 min; 2μM Gö6976 (Calbiochem, EMD Millipore Corporation, Billerica, MA, USA), according to our previous publication [10]. For the measurement of sAβPPα release, Bryostatin-1 (or solvent) was incubated for 15 min, and after this time the cell culture medium was replaced with fresh medium, which after 24 h was analyzed for sAβPPα content. In the case of concomitant treatment, Gö6976 was added to the cells 20 min before Bryostatin-1.
MTT assay
Mitochondrial enzymatic activity was estimated by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay from Sigma. A cell suspension of 2.5×103 cells/well in 100μl was seeded into 96-well plates. After each treatment, 10μl of MTT (concentration equal to 1 mg/ml) was added to each well. After incubation at 37°C for 4 h, the formed purple formazan crystals were solubilized in 100μl of lysis buffer [20% sodium dodecyl sulfate (SDS) in 50% dimethylformamide] overnight. Absorbance values were measured at 595 nm in a microplate reader from Bio-Rad Laboratories (Segrate, Italy) and the results expressed as % with respect to control.
LDH assay and trypan blue cell count
Cell viability was estimated by measuring LDH (Lactate Dehydrogenase) activity in the culture supernatant by a Cytotoxicity Detection Kit (Roche). Briefly, a suspension of 2×104 cells/200μl per well was seeded into 96-well plates. After each treatment, 100μl of culture supernatant were transferred into a new 96-well plate, and 100μl of Reaction Mixture provided by the kit were added. After 30 min incubation, the reaction was stopped by adding 50μl of HCl 1N. Absorbance was measured at 490/600 nm in Synergy HT microplate reader, BioTek (Winooski, VT, USA) and the results expressed as % with respect to control.
The adherent cells were detached from the 96-well plate and counted by the Trypan blue exclusion test. Cells from four different wells were counted and the results expressed as % with respect to control.
Immunocytochemistry
2×105 cells/500μl cells plated on coverslips were washed with phosphate buffered saline (PBS), fixed in 70% ethanol at –20°C, washed again with PBS and permeabilized for 15 min at room temperature (RT) with 0.01% TritonX-100 in PBS. Nonspecific binding sites were blocked by incubation for 30 min with PBS containing 1% bovine serum albumin (BSA). Cells were then incubated for 1 h at RT with the primary anti-Aβ mouse monoclonal 6E10 antibody (from Merck KGaA) diluted 1:75 in PBS containing 1% BSA. As negative control, the primary antibody was replaced by a PBS/1% BSA solution. Subsequently, cells were incubated with the R-phycoerythrin-conjugated anti-mouse rabbit antibody from Calbiochem (Darmstadt, Germany) diluted in PBS/1% BSA solution at 1:200. After the labeling procedures, cells were mounted on glass slides and counterstained for DNA with a 0.1μg/ml Hoechst solution.
Gene silencing
For HuD gene silencing, a short hairpin RNA (shRNA) plasmid-based technology was used. A double stranded oligonucleotide complementary and specific for human HuD mRNA sequence (5′-GGATTCATCCGCTTTGATA-3′) was BglII/HindIII cloned into a modified pSUPER.retro vector (a kind gift of G. Pelicci, IEO, Milan, Italy). Transfections were performed using Cellfectin Reagent from Life technologies (Milan, Italy) following the manufacturer’s instructions. Puromycin (0.7μg/ml) was used for selection for 2-3 weeks and, after isolation of the single stable shRNA transfectants, colonies were maintained in 0.4μg/ml puromycin.
Western blotting
Cells were harvested and homogenized using a teflon/glass homogenizer as previously described [28]. The protein content was measured via the Bradford’s method using BSA as a standard. Protein lysates were resolved on 12% SDS-polyacrylamide gel electrophoresis, and then processed following standard procedures. The antibodies anti-HuD and anti-ADAM10 (AbCam, Cambridge, UK), anti-panELAV, anti-PKCα (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Aβ and anti-AβPP (Merck KGaA, Darmstadt, Germany), and anti-α-tubulin (Sigma) were diluted at 1:1000 in TBST buffer [10 mM Tris-HCl, 100 mM NaCl, 0.1% (v/v) Tween 20, pH 7.5] containing 6% milk. The nitrocellulose membranes were processed using Pierce ECL Plus from Thermo Scientific (Rockford, IL, USA). Western blotting assays were performed twice on six different cell preparations using α-tubulin to normalize data. Densitometric analysis was performed using the NIH Image software (http://rsb.info.nih.gov/nih-image).
Quantitative real-time PCR
RNA was extracted from total cell homogenates by using RNeasy Micro Plus Kit from Qiagen (Valencia, CA, USA). The reverse transcription was performed with QuantiTect Reverse Transcription Kit from Qiagen following standard procedures. PCR amplifications were carried out in triplicate using the Lightcycler instrument from Roche Molecular Biochemicals (Mannheim, Germany) in the presence of QuantiTect SYBR Green PCR mix from Qiagen with primers designed by using the PRIMER3 software (http://bioinfo.ut.ee/primer3/). Primer sequences are reported in Supplementary Table 1.
RPL10a and GAPDH were chosen as the reference transcripts for content normalization of nELAV and ADAM10, respectively, since they remained substantially stable during all the treatments (not shown). Values were expressed as cycle threshold (Ct), ΔCt was calculated versus RPL10a or GAPDH values and expressed as mean percentages±S.E.M. with respect to control.
To study the effect of 24 h Aβ40 treatment on the binding between nELAV proteins and ADAM10 mRNA in SH-SY5Y cells, nELAV proteins immunoprecipitation followed by RNA extraction and real-time quantitative PCR for ADAM10 transcript was performed according to our previous publication [25]. The GAPDH mRNA was chosen to check the specificity of nELAV binding to ADAM10.
Detection of the extracellular protein sAβPPα by ELISA assay
The sAβPPα released in the medium was measured using a commercial ELISA kit from IBL International (Hamburg, Germany). The cells were treated as described previously and after 24 h 100μl of diluted conditioned medium were transferred into wells pre-coated with a monoclonal anti-human sAβPPα mouse IgG (clone 2B3). Samples were incubated overnight at 4°C, washed, and then treated with HRP-conjugated anti-human AβPP mouse IgG (clone R101A4). The color reaction was performed using 3,3′,5,5′-tetramethylbenzidine substrate and the absorbance was measured at 450 nm using a microplate reader from Bio-Rad Laboratories. All the detections were performed in duplicate and the experiments were repeated at least three times. The results are expressed as % with respect to control.
Statistics
For the statistical analysis the GraphPad InStat program (GraphPad software, San Diego, CA, USA) was used. Data were analyzed by the analysis of variance (ANOVA) followed, when significant, by an appropriate post hoc comparison test, as specifically indicated. Differences were considered statistically significant when p≤0.05.
RESULTS
Aβ40 exposure leads to a decrease of nELAV and ADAM10 protein levels in neuroblastoma SH-SY5Y cells
Since we previously demonstrated that 1μM Aβ42 for 24 h directly affects nELAV pathway and that ADAM10 mRNA is a target of nELAV proteins [25], in order to investigate if ADAM10 gene expression can be altered by Aβ40 peptide, we exposed SH-SY5Y cells to freshly prepared 1μM Aβ40 or to the solvent alone for 24 h. We first evaluated the ability of this peptide to penetrate into the cells. Immunocytochemistry analyses (Fig. 1A-B) show that Aβ40 enters the cells, where it has a cytoplasmic localization. As expected, since Aβ40 is physiologically produced by SH-SY5Y cells, also control cells present an Aβ40-positive staining, although much fainter than treated cells. We performed western blotting experiments on total cell lysates and cell culture medium using the 6E10 antibody, which recognizes the aminoacids 1–17 within Aβ fragment, to detect both Aβ monomer and its self-assembled oligomers. We found intracellularly a similar pattern between control and Aβ-treated cells, although the 4 kDa monomer is detectable only in Aβ-treated cells (Fig. 1C, left). In the culture medium, beside the Aβ monomer, Aβ-treated cells show the presence of various Aβ-positive bands of higher molecular weight, indicating that Aβ40 gives rise to aggregates (Fig. 1C, right).
Cell treatment with Aβ40 for 24 h induced no significant effects on mitochondrial function and cell viability as measured by the MTT, LDH, and Trypan blue exclusion assays (Fig. 1D–F), in accordance with previous observations on the Aβ40–induced effects on SH-SY5Y cells [29], and on primary rat hippocampal neurons [30].
We then evaluated the effect of Aβ40 exposure on nELAV protein levels in total homogenates, finding that they are significantly decreased (–21%) after 24 h treatment (Fig. 2A). In parallel, we observed that also ADAM10 protein is downregulated by Aβ40 (Fig. 2B). Importantly, the effect of Aβ40 on nELAV and ADAM10 protein levels is specific, since their content is not influenced by exposure to the reverse peptide Aβ40 - 1 (Fig. 2A-B). If we compare the Aβ40- and Aβ42-mediated effects on nELAV and ADAM10 protein contents, no significant differences are found [for nELAV: Aβ40: 470.0±31.8; Aβ42: 481.15±31.0. For ADAM10: Aβ40: 300.5±38.5; Aβ42 361.6±36.2. Values expressed as nELAV/α-tubulin or ADAM10/α-tubulin immunoreactivities×1000±S.E.M.; p > 0.05, Dunnett’s multiple comparisons test, n = 3 –6]. As we previously showed for Aβ42 [25], also the binding of nELAV proteins to ADAM10 mRNA is disrupted by Aβ40 exposure (Supplementary Figure 1); no variations in ADAM10 mRNA levels were observed in SH-SY5Y cells exposed to Aβ40 or Aβ40 - 1 for 24 h (Fig. 2C), as also documented for Aβ42 [25].
We also investigated the expression of another HuD-target, AβPP [31], finding that it is not significantly affected by Aβ40 in our model (Fig. 2D, insert). Since ADAM10 protein triggers AβPP proteolysis resulting in the release of the neuroprotective sAβPPα peptide, in order to determine whether the Aβ40–mediated decrease in ADAM10 protein levels affects sAβPPα production, we measured the release of this peptide into the cell culture medium. As evidenced by the ELISA assay, 24 h treatment with Aβ40 does not significantly alter the secretion of sAβPPα (Fig. 2D). This result was also confirmed by western blotting analysis on this secreted peptide in the cell culture medium (CTR: 922.5±71.9; Aβ40: 842.2±75.8; values expressed as sAβPPα/α-tubulin immunoreactivities×1000±S.E.M., n = 7).
In HuD-silenced cells, nELAV protein downregulation is counteracted by Bryostatin-1 treatment
To further evaluate the effects of HuD/nELAV depletion on ADAM10 levels, we established a stable SH-SY5Y cell line where HuD gene expression was silenced by shRNA-based technology. The efficiency of HuD gene silencing in the selected shHuD clone was evaluated by western blotting analysis with a specific anti-HuD antibody. We found that, in the stably knocked-down shHuD cells, HuD protein level decreased by 76% and 68% in comparison to parental cells and mock-transfected cells carrying the empty shRNA plasmid, respectively (Fig. 3A). Importantly, we also found that shHuD-silenced cells show a significant downregulation of all nELAV mRNAs and proteins (Fig. 3A–C). Conversely, there is no difference in the protein content of HuR, the ubiquitous member of the ELAV family, in the total homogenates of the three cell lines considered (parental: 822.1±84.90; mock-transfected: 714.0±103.7; shHuD: 704.5±162.2; values expressed as HuR/α-tubulin immunoreactivities×1000±S.E.M.; p > 0.05, Tukey–Kramer multiple comparisons test, n = 6; see a representative immunoblot in Fig. 3A). Since the content of nELAV proteins was comparable in parental and mock-transfected cells, the following experiments were performed only on mock-transfected cells used as a control.
Based on our previous observation that Bryostatin-1-mediated PKCα activation is associated with an upregulation of nELAV proteins both in vivo and in vitro [10], we evaluated the effect of this DAG-mimicking compound on HuD-silenced cells. To this aim, we exposed mock-transfected and shHuD cells to 100 nM Bryostatin-1 for 15 min. The treatment led to a significant increase of nELAV protein levels in both stable shHuD and mock-transfected cell lines (Fig. 4A). No statistically significant difference in nELAV protein content was found between mock-transfected and Bryostatin-1-treated shHuD cells.
By western blotting analysis, we found that, as expected, PKCα protein basal levels were comparable between the two cell lines, indicating that the expression of this enzyme is not modified by HuD silencing (Fig. 4B). The treatment with PKC activators was described to upregulate the expression of the enzymes themselves [10, 32–34]; we thus determined whether Bryostatin-1, besides stimulating PKCα activity, was able to affect also the amount of this isoenzyme in both mock-transfected and shHuD cells. In accordance to previous evidence, Bryostatin-1 treatment significantly increased PKCα protein amount in both mock-transfected and shHuD cells by +55% and +81%, respectively (Fig. 4B).
Downstream the PKC/ELAV cascade, we found a consistent decrease in ADAM10 protein content in shHuD cells with respect to control (Fig. 4C). Bryostatin-1 exposure promoted ADAM10 upregulation in both cell lines; again, no statistically significant difference in ADAM10 protein expression was found between mock-transfected and Bryostatin-1-treated shHuD cells, suggesting that PKCα activation is able to counteract ADAM10 downregulation in HuD silenced cells (Fig. 4C). Bryostatin-1 upregulated ADAM10 protein expression acting on PKCα, indeed in presence of the PKCα inhibitor Gö6976 this effect was blunted (Supplementary Figure 2).
In parallel, quantitative real-time PCR analysis demonstrated that PKC activation leads to a statistically significant upregulation of ADAM10 also at mRNA level in shHuD cells (CTR: 100.0±12.6; BRYO: 164.1±29.8, where the ADAM10 values, corresponding to the ΔCt versus GAPDH mRNA, are expressed as mean percentages±S.E.M.; p < 0.05 Student’s t test, n = 6).
sAβPPα release is favored by Bryostatin-1 treatment
To deepen further the effect of Bryostatin-1 treatment, we investigated whether sAβPPα levels were altered in the culture medium following PKC activation.
We treated both mock-transfected and shHuD cells with 100 nM Bryostatin-1 for 15 min and, after 24 h, we measured sAβPPα release in the conditioned media by ELISA. As reported in Fig. 4D, after Bryostatin-1-mediated PKCα activation there was a significant increase of sAβPPα release in both mock and shHuD cells.
Bryostatin-1 increases nELAV levels in shHuD Aβ40-treated cells
Since both 24 h Aβ40 treatment and HuD silencing produced a significant decrease in nELAV protein levels in human SH-SY5Y cells (Figs. 2A and 3B), we checked whether Bryostatin-1 is able to counteract the combined effect induced by both Aβ40 treatment and HuD silencing on nELAV proteins amount. Therefore, we pre-exposed shHuD cells to Aβ40 for 24 h and subsequently treated them with Bryostatin-1 for 15 min. We found that Bryostatin-1 significantly upregulated nELAV protein content also in these “double insult” experimental conditions (Fig. 5).
DISCUSSION
Several studies on the ectopic expression of nELAV in vertebrates and experiments of gene overexpression/knock-down in various neural cell lines reveal the key role of nELAV proteins in neural development and maintenance [35–38]. In parallel, biochemical investigations established the molecular function of these mammalian RBP as positive regulators of gene expression at post-transcriptional level, by primarily increasing the stability and/or translation of a wide, still growing number of ARE-containing mRNAs [8, 39–41]. Among them, several nELAV target transcripts are involved in the homeostasis of neuronal metabolism, neurite outgrowth and synaptic plasticity [21, 42]. As a consequence, alterations in nELAV protein levels and/or function likely reverberate on changes in the expression of their target genes and may play a role in some pathological conditions affecting the nervous system [3, 44]. Accordingly, Scheckel et al. recently listed the huge number of transcripts bound by nELAV proteins in the human brain, including ADAM10 mRNA, and highlighted that many of them code for proteins that are important for neuronal functions and are linked to neurodegenerative disorders, such as AD [45]. To this regard, we previously demonstrated that nELAV protein levels decrease as a function of clinical dementia progression in the hippocampus from subjects with a different degree of AD, and that ADAM10 content is impaired as well in the same post mortem samples [25].
Interestingly, we also found that nELAV protein expression inversely correlates with the hippocampal content of Aβ, and that in vitro the exposure to Aβ42 specifically determines a decrease in nELAV proteins and the disruption of their binding to the target ADAM10 mRNA, finally resulting in the downregulation of ADAM10 protein expression [25]. We here investigated the effects of Aβ40 peptide on the same cascade in SH-SY5Y neuroblastoma cells, showing that the exposure to this fragment is responsible for a similar significant decrease of both nELAV and ADAM10 protein levels. This result suggests that, although Aβ40 represents a more physiological and abundant product of the amyloidogenic AβPP processing with respect to Aβ42, at high concentrations Aβ40 can produce analogous negative effects on the ELAV/ADAM10 pathway in these cells. The effect is specific for this peptide, since the reverse peptide Aβ40 - 1 does not cause any alteration. According to our previous data on Aβ42-treated cells, no changes in ADAM10 mRNA levels were detected following Aβ40 exposure, suggesting that the effect of Aβ peptides on ADAM10 is likely to be at a post-transcriptional level. In agreement with this concept, we found that the binding of nELAV proteins to ADAM10 mRNA is disrupted by Aβ40 exposure, similarly to what previously demonstrated for Aβ42 [25]. These observations are in line with the data reported in a recent paper by Scheckel et al., where they show that in conditions of short- or long-term stress, including AD, the association of nELAV proteins with their physiological mRNA targets is reduced, likely contributing to disease [45].
To investigate the consequences of HuD deficiency on ADAM10, we established a new in vitro cellular model in which, besides the specific HuD silencing, we found a decrease in the other neuronal members of the ELAV family at both transcript and protein level, indicating that our cell line reflects a condition of a general nELAV deficit. The same observation was previously reported by us in vivo [46] and by others in a different neuronal cell line, NSC-34, probably because of a cross-autoregulatory mechanism [24]. Analogously to Aβ40-treated cells, the HuD-silenced cells contain lower ADAM10 protein levels in comparison to controls, further validating nELAV proteins as key regulators of ADAM10 expression. In the future, it will be of interest to confirm these findings in primary neonatal neuronal cultures; however, notwithstanding their intrinsic limitations, SH-SY5Y human neuroblastoma cells exhibit many biochemical and functional properties of neuronal phenotype, and for this reason they are used in in vitro studies requiring neuronal-like cells [47].
We then evaluated the effect of Bryostatin-1, a PKCα activator, in our cellular model characterized by nELAV deficit. First, we found that the protein levels of PKCα, an upstream regulator of nELAV activity in SH-SY5Y cells [10], are not affected by HuD downregulation. In this shHuD cell line, as well as in mock-transfected cells, Bryostatin-1 upregulates both nELAV and ADAM10 proteins if compared to their respective non-treated counterparts. Notably, no statistically significant difference in both nELAV and ADAM10 protein content was found between untreated mock- and Bryostatin-1-treated shHuD cells, suggesting that PKC activation may be useful to restore the basal levels of nELAV/ADAM10 in conditions characterized by a deficit of these proteins. Interestingly, besides nELAV and ADAM10, also PKCα protein levels were upregulated following Bryostatin-1 treatment. This last result is in line with our previous findings showing that PKC activators are able, besides inducing a change in the subcellular localization of the kinase, to increase the expression of the enzyme itself [10, 48], possibly through a sort of feedback mechanism acting at transcriptional level. Since Bryostatin-1 activates both -α and – ɛ PKC isoenzymes [49], in order to assess the specific involvement of PKCα in ADAM10 regulation, we carried out experiments using Gö6976, a PKC inhibitor acting specifically on the α isoform, but not on the ɛ isoform. Indeed, the Bryostatin-1-induced upregulation of ADAM10 protein is blunted by Gö6976, indicating that PKCα is the isoenzyme specifically involved here.
Bryostatin-1 treatment significantly increases nELAV protein levels also in Aβ40-treated shHuD cells, further suggesting its potential role in counteracting nELAV deficiency also in conditions where a “double insult” on the same target occurs.
With the aim to explore the biological consequences of the Aβ40 challenge or HuD silencing downstream the nELAV/ADAM10 pathway, we focused on the neurotrophic peptide sAβPPα released after the ADAM10-mediated cleavage of AβPP [27]. We found no changes in sAβPPα release neither in SH-SY5Y cells treated with Aβ40, nor in shHuD cells; this suggests that, although decreased in both conditions, the remaining amount of ADAM10 protein is sufficient to guarantee a basal release of this neurotrophic peptide. However, we cannot exclude that a compensation by other α-secretases may occur. Interestingly, a significant increase in sAβPPα secretion was observed in both mock cells and in shHuD cells treated with Bryostatin-1. This is particularly relevant, given that a recent clinical trial deems molecules favoring ADAM10 activity as potential therapeutic approaches in AD [50]. Our finding is also in line with literature data reporting that the direct activation of PKC by Bryostatin-1 or other diacylglycerol-analogues increases sAβPPα secretion to the detriment of the amyloidogenic peptides [34, 51–53], supporting the evidence of a neuroprotective role of Bryostatin-1. The relevance of Bryostatin-1 as a neuroprotective agent has been previously documented by the improvement of cognitive functions in AD transgenic mice exerted by this PKC agonist [54] and, more recently, by the evidence of its positive effects on synaptic plasticity and other neuronal processes and functions [49, 55]. Noteworthy, a recent case report evidenced clinical benefits after a compassionate use of Bryostatin-1 in a patient with advance symptoms of AD [56].
PKC is comprised in the so-called cognitive kinases and its activation is considered a potential therapeutic tool for the treatment of cognitive disorders [54–59]. For this reason, a better understanding of the molecular and cellular effects mediated by PKC activators, such as Bryostatin-1, is of great relevance.
In conclusion, on one side, we established a new cellular model of nELAV deficiency and validated ADAM10 as a target of nELAV; on the other, we demonstrated that the exposure to Aβ40 peptide decreases nELAV content and ADAM10 protein level. We also showed that Bryostatin-1 exerts a positive effect on nELAV and on their downstream target ADAM10, providing evidence that it is possible to modulate this cascade acting on upstream regulatory factors like PKC. Considering the critical involvement of nELAV proteins in memory processes [46, 61] and their dramatic reduction in AD brains [25], these findings suggest that Bryostatin-1 is a promising molecule to counteract physiological and pathological age-related cognitive decline. Of relevance, this cellular model may represent an innovative tool useful to explore the effect of novel molecules potentially acting on this molecular cascade, not only upstream but also directly on nELAV, such as ELAV-mimicking peptides [62].
DISCLOSURE STATEMENT
Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/16-0299r1).