Address correspondence and reprint requests to Ke Zen, Chenyu Zhang and Qipeng Zhang, Jiangsu Engineering Research Center for Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, 22 Hankou Road, Nanjing, Jiangsu 210093, China. E-mails: email@example.com; firstname.lastname@example.org; email@example.com
In this study, we first characterized synaptosome microRNA (miRNA) profiles using microarray and qRT-PCR. MicroRNAs were detected in isolated synaptic vesicles, and Ago2 immunoprecipitation studies revealed an association between miRNAs and Ago2. Second, we found that miR-29a, miR-99a, and miR-125a were significantly elevated in synaptosome supernatants after depolarization. MiRNA secretion by the synaptosome was Ca2+-dependent and was inhibited by the exocytosis inhibitor, okadaic acid. Furthermore, application of nerve growth factor increased miRNA secretion without altering the spontaneous release of miRNAs. Conversely, kainic acid decreased miRNA secretion and enhanced the spontaneous release of miRNAs. These results indicate that synaptosomes could secrete miRNAs. Finally, synthesized miRNAs were taken up by synaptosomes, and the endocytosis inhibitor Dynasore blocked this process. After incubation with miR-125a, additional miR-125a was bound to Ago2 in the synaptosome, and expression of the miR-125a target gene (PSD95 mRNA) was decreased; these findings suggest that the ingested miRNAs were assembled in the RNA-induced silencing complex, resulting in the degradation of target mRNAs. To our knowledge, this is the first study that demonstrates the secretion of miRNAs by synaptosomes under physiological stimulation and demonstrates that secreted miRNAs might be functionally active after being taken up by the synaptic fraction via the endocytic pathway.
MicroRNAs (miRNAs) are a group of endogenous non-coding RNAs that consist of 18 to 25 nucleotides. They play an important role in regulating gene expression by priming to complementary sites on target mRNAs and function to either block mRNA translation or trigger their degradation (He and Hannon 2004). In mammalian cells, mature miRNAs are assembled into a multi-protein RNA complex known as the RNA-induced silencing complex (RISC), and Ago2 is a core component of the RISC, which directly binds to miRNAs (Liu et al. 2004; Meister et al. 2004; Diederichs and Haber 2007). As miRNAs target approximately 30–60% of mammalian proteins, they function in a variety of biological processes including cell proliferation, differentiation, migration, and apoptosis (Bartel 2004; Lewis et al. 2005; Esquela-Kerscher and Slack 2006; Friedman et al. 2009; Huang and He 2010). Hundreds of miRNAs are expressed in the brain, which suggests that they may serve an important function in the nervous system (Lagos-Quintana et al. 2002; Krichevsky et al. 2003). Impairment of miRNA biogenesis processing is known to disrupt the development of the nervous system including differentiation, learning, memory, and the survival of individual neurons (Davis et al. 2008; Kawase-Koga et al. 2009, 2010; Hebert et al. 2010; Huang et al. 2010; Konopka et al. 2010; Zehir et al. 2010).
Furthermore, increasing evidence suggests that a large body of miRNAs and their precursors exist in synaptic fractions (Lugli et al. 2008). Moreover, Ago2 protein is also located in synaptic fractions (Lugli et al. 2005). Recent studies show that synaptic miRNAs function in an activity-dependent regulatory manner and control local mRNA translation (Wayman et al. 2008; Siegel et al. 2011). Similar to circulating miRNAs, miRNAs are also present in the CSF and are potentially useful as novel non-invasive biomarkers for the diagnosis of CNS disorders (Cogswell et al. 2008; Baraniskin et al. 2011). It has been demonstrated that the majority of circulating miRNAs consist of miRNAs that have been released via active exocytosis, which indicates that extracellular miRNAs have specific functions (Chen et al. 2012). However, the function of CSF miRNAs remains unknown.
In the nervous system, it remains unknown which neural cell types and which parts of the cell are sources of CSF miRNAs. In this study, we explored the possibility that excitable synaptic fractions might secrete miRNAs under physiological stimulation. Synaptosomes are sealed presynaptic nerve terminals that can respire, take up oxygen and glucose, extrude Na+, accumulate K+, maintain a normal membrane potential and release neurotransmitters upon depolarization in a Ca2+-dependent manner. We found that potassium chloride-induced depolarization results in a clear secretion of miRNAs (miR-29a, miR-99a, and miR-125a). This secretion was inhibited by the exocytosis inhibitor, okadaic acid, and miRNA secretion was completely abolished in a Ca2+-free solution, which suggests that the secretion of these miRNAs was Ca2+-dependent. Incubation with nerve growth factor (NGF) enhanced the secretion of miRNAs without altering the spontaneous release of miRNAs. Conversely, incubation with kainic acid (KA) decreased miRNA secretion and elevated the spontaneous release of miRNAs. Interestingly, we also found that synthesized miRNAs were taken up by synaptosomes and that the endocytosis inhibitor, Dynasore, partially inhibited the miRNAs uptake, suggesting that miRNAs were taken up by synaptosomes via an endocytic pathway. After incubation with synthesized miR-125a, Ago2-associated miR-125a was increased and postsynaptic density protein 95 (PSD95) mRNA was significantly degraded in the synaptosomes. Taken together, our results suggest that physiological stimulation modulates the release of miRNAs from the synaptic fraction and that secreted miRNAs are functional after being taken up by the synaptic fraction.
Animals and reagents
Adult male C57BL/6 mice (20–30 g) were employed in this study. All of the animals were handled in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals and approved by the Animal Care Committee of Nanjing University (Nanjing, China). Sucrose (DNase- and RNase-free), potassium chloride (KCl), glutamate, GABA, okadaic acid (OA), tetanus toxin (TeXN), NGF, KA, and Dynasore were purchased from Sigma (Sigma-Aldrich, St. Louis, MO, USA). Pitstop2 was purchased from Abcam (Abcam, Cambridge, MA, USA). Ultrapure water, HEPES, and glycogen were purchased from Invitrogen (Invitrogen, San Diego, CA, USA). Synthesized mature miRNAs (mmu-miR-29a, mmu-miR-99a, mmu-miR-125a, mmu-miR-138, and negative control miRNA) were purchased from Invitrogen (Invitrogen, Shanghai, China).
Synaptic fraction preparation
Synaptosomes were prepared using two forebrains that included the cortex and hippocampus. The synaptosomes were prepared using a combination of differential centrifugation and sucrose density gradient centrifugation techniques according to a previously described method by Gray and Whittaker (Gray and Whittaker 1962). Some modifications were made to preserve the integrity of the RNA, such as using DNase- and RNase-free ultrapure water and sucrose. Briefly, forebrains were gently homogenized using a Potter Teflon-glass homogenizer in ice-cold gradient buffer (320.0 mM sucrose, 5.0 mM HEPES, 0.1 mM EDTA and 1.0 mM dithiothreitol, pH 7.5) containing a cocktail of protease (Sigma) and recombinant RNase inhibitors (Takara, Dalian, China). The mixture was centrifuged at 1000 g for 20 min, and the supernatant was centrifuged again to exclude cell nuclei and debris. The supernatant (S1) was centrifuged at 10 000 g to collect the crude synaptosome-enriched pellet (P2). This pellet was then washed and resuspended in ice-cold gradient buffer and loaded onto a sucrose gradient (0.8 M : 1.2 M = 1 : 1). After centrifugation at 100 000 g for 1 h, the synaptosomes (a thin layer between 0.8 M sucrose and 1.2 M sucrose) were collected and diluted with an equal volume of ice-cold ultrapure water and then centrifuged again to collect the purified synaptosome pellet (Syn). The synaptosome pellet was resuspended in ice-cold artificial CSF (aCSF, 2.0 mM calcium chloride, 132.0 mM sodium chloride, 3.0 mM potassium chloride, 2.0 mM magnesium sulfate, 1.2 mM sodium dihydrogen phosphate, 10.0 mM HEPES, and 10.0 mM glucose, pH 7.4) containing a cocktail of protease inhibitors (Sigma) and RNase inhibitors (Takara) and kept on ice for 30 min to recover the metabolic activities prior to use in further experiments.
The PSD was prepared according to the method described by Villasana et al. (2006) with some modifications. The initial starting material was the synaptosome suspension, which was added to ultrapure water and sucrose (DNase- and RNase-free) to preserve the integrity of the RNA. The purified PSD fraction was dissolved in a buffered solution by sonication for 240 s, and the synaptosomes and PSD samples were collected for western blotting analysis.
The preparation protocol for the crude synaptic vesicles (SV) was adapted from a method previously described by Huttner et al. (1983), Each synaptosome and synaptic vesicle preparation consisted of 10 forebrains; one half of the synaptosome-enriched pellet (P2) was used to prepare the synaptosome (Syn) fraction as described above, and the remaining half of the P2 pellet was washed and centrifuged to obtain the P2' pellet (P2'). The P2' pellets were resuspended in buffered sucrose and lysed by hypoosmotic shock (the samples were added to 9 volumes of ultrapure water and homogenized). The resulting lysates were poured rapidly into a tube containing 1 M HEPES-NaOH buffer (pH 7.4), and the suspension was kept on ice for 30 min. The samples were then centrifuged for 20 min at 25 000 g to yield the lysate pellets (LP1) and lysate supernatants. The supernatants (LS1) were then transferred into new tubes and centrifuged for an additional 2 h at 167 000 g. The supernatants (LS2) were then removed, and the pellets (LP2 or SV) were washed and resuspended to a total volume of 300 μL in buffered solution. The P2', Syn, LP1, and SV fractions were collected for western blotting analysis.
Transmission electron microscopy
Fresh synaptosomes were fixed with EM fixative solution (4% paraformaldehyde, 5% glutaraldehyde in 0.1 M sodium cacodylate and 3.4 mM calcium chloride, pH 7.2) and sectioned into 80-nm thick samples. These samples were examined by transmission electron microscopy (Philips 201C, Philips, Eindhoven, Netherlands) to characterize the morphology of the synaptosomes. Standard images were obtained using a CCD camera attached to the TEM.
After the pre-column derivation, the synaptosome supernatants were injected into an HPLC apparatus using a Gemini C18 column (250.0 × 4.6 mm, 5 μm, Phenomenex, Torrance, CA, USA) and measured using a fluorescence detector (Ex: 338 nm, Em: 450 nm). The injected volume was 10 μL, and the flow rate was 1 mL/min. Glutamate and GABA standards were used to quantify the amounts of glutamate and GABA.
LDH activity assay
We monitored the activity of the cytosolic enzyme, lactate dehydrogenase (LDH, EC 188.8.131.52), to reveal any membrane leakage in the synaptosomes. LDH, which converts exogenously applied pyruvate and NADH into lactate and NAD+, can be used as an indicator of cell injury (Koh and Choi 1987). The LDH activity assay kit was purchased from Promega (Madison, WI, USA) and the assay was performed according to the manufacturer's instructions.
Western blotting analysis
Protein samples were homogenized in 10% (w/v) extraction buffer (10.0 mM Tris PH 7.4, 150.0 mM sodium chloride, 0.4% TritonX-100 and complete protease inhibitors) and then centrifuged at 17 000 g for 20 min. The supernatants were collected, and the protein concentration was quantified using a BCA kit (Thermo Scientific, Rockford, IL, USA). The prepared protein samples were separated using either 12% or 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (0.45 μm Whatman, Dassel, Germany). The membranes were blocked with 5% non-fat dry milk in 0.01 M phosphate-buffered saline containing 0.05% Tween-20 (PBST, pH 7.4) at 25°C for 1 h. Next, the membranes were incubated with primary antibodies at 4°C for 12 h. The final dilutions for the primary antibodies were as follows: anti-synaptotagmin1 (Cell Signaling, 1 : 500, Cell Signaling Technology, Danvers, MA, USA), anti-PSD95 (Abcam, 1 : 2000), anti-Ago2 (Cell Signaling, 1 : 1000) and anti-GAPDH (Abcam, 1 : 2000). After three washes in PBST, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (Amersham, Arlington Heights, IL, USA) diluted at 1 : 5 000 in PBST for 1 h. After three washes, the blots were visualized using an enhanced chemiluminescence kit (Amersham).
Five independently prepared synaptosome pellets were pooled for RNA isolation, and the total RNA was analyzed using TaqMan® Rodent MicroRNA Arrays (v2.0 Invitrogen Inc.). This array detected 585 mmu- or rno-miRNAs according to the CT value, and the average ΔCT was compared with mammalian sno RNA U6.
RNA isolation and qRT-PCR
Total RNA was isolated using Trizol reagent (Invitrogen) according to the manufacturer's instructions. Glycogen (Invitrogen) was added to the RNA precipitation step. Quantitative RT-PCR of mature miRNAs was performed using TaqMan microRNA probes (Applied Biosystems, Foster, CA, USA) according to the manufacturer's instructions. Briefly, the RNA from each sample was reverse-transcribed into cDNA using AMV stem-loop RT primers (Ambion, Foster, CA, USA) and reverse transcriptase (Takara, Dalian, China). Quantitative real-time PCR was performed using a TaqMan PCR kit on an Applied Biosystems 7300 Sequence Detection System. All of the reactions, including the no-template controls, were run in triplicate. The threshold cycle for each sample was chosen within the linear range and converted into a starting quantity by normalizing to the internal standard U6 run on the same plate.
MicroRNA immunopurification by anti-Ago2 antibody
The Ago2-associated miRNAs were purified according to the protocol described by Keene et al. (2006). Briefly, a monoclonal anti-Ago2 antibody (Abcam) was used for co-immunoprecipitation studies, and an identical IgG-type antibody was used as the negative control antibody. Protease inhibitor cocktail (Sigma), RNase inhibitors (Takara), and vanadyl ribonucleoside complexes (New England Biolabs, Ipswich, MA, USA) were added into the polysome lysis and NT2 buffers. For the synaptosome pellets, a mild lysis was performed by adding polysome lysis buffer, and the lysed solution was kept on ice. After immunoprecipitation, protein A/G beads (Santa Cruz, CA, USA) were washed four times, and total RNA was isolated from the immunoprecipitated pellets by adding Trizol reagent (Invitrogen) to the beads. Glycogen (20 μg/mL) was added as a carrier to the RNA precipitation reaction, and the total RNA was analyzed using qRT-PCR. A portion of the protein A/G beads and synaptosome lysates were collected for western blotting analysis.
Depolarization stimulation and drug treatment
After resting in an ice-bath, the resuspended synaptosomes were added to three volumes of pre-warmed (37°C) aCSF, mixed gently with a pipette and pre-incubated in a 37°C water bath for 3 min. A depolarizing agent (potassium chloride) was then added to each tube to a final concentration of 30 mM, and the tubes were then mixed gently and placed in the 37°C water bath for an additional 5 min. At the end of the incubation, the contents of the assay tubes were quickly pipetted into Eppendorf tubes containing 50 μL of an oil mixture [silicone oil and dinonylphthalate mixed at 45 : 55% (v/v)]. The samples were then centrifuged at 15 000 g for 1 min to terminate the depolarization reaction. Aliquots were removed from the supernatants for neurotransmitter analysis and LDH assay, and the remaining supernatants were collected for miRNA quantitative analysis. The synaptosome pellets were also collected and frozen at −70°C for further analysis.
For the OA treatment, the synaptosomes were incubated with 100 nM OA or dimethylsulfoxide for 30 min. The synaptosomes were then collected by centrifugation, and the synaptosome pellets were resuspended in aCSF and depolarized as previously described.
For the TeXN treatment, the synaptosomes were incubated with 10 ng/mL TeXN or saline for 60 min. The synaptosomes were then collected by centrifugation, and the synaptosome pellets were resuspended in aCSF and depolarized as previously described.
For the NGF treatment, the synaptosomes were incubated with 100 ng/mL NGF or saline for 30 min. The synaptosomes and supernatants were then collected by centrifugation, and the synaptosome pellets were resuspended in aCSF and depolarized as previously described.
For the KA treatment, the synaptosomes were incubated with 100 μM KA or saline for 10 min. The synaptosomes and supernatants were collected via centrifugation, and the synaptosome pellets were resuspended in aCSF and depolarized as previously described.
In these experiments, both the supernatants and synaptosome pellets (with or without having experienced specific treatments) were collected for miRNA analysis.
MiRNA incubation and RNase digestion
Synaptosomes were resuspended in aCSF at 3 mg/mL, and various concentrations of synthesized miRNAs were added to the solution and incubated at room temperature for 1 h. After extensive washing (five washes), a final concentration of 10 U/mL RNaseA (Promega) was added to the samples, and the synaptosomes were incubated at 37°C for 1 h. After extensive washing, Trizol was added to the synaptosome pellets. In parallel experiments, endocytosis inhibitors such as Dynasore (100 nM) or Pitstop2 (30 μM) and vehicle controls were added to the synaptosomes prior to miRNA incubation. After 20 min of pre-treatment, the synaptosomes were incubated with miRNAs (0.5 pmol) for 1 h and digested with RNaseA as previously described.
PSD95 mRNA quantitation
Total RNA was extracted from the synaptosome pellets with and without miR-125a incubation, and real-time PCR was employed to quantify the level of PSD95 mRNA according to previously described methods (Muddashetty et al. 2011). The forward PSD95 primer used was 5′-GCC GTT TGA GTT CTC CTT T -3′, and the reverse primer was 5′-AGC ATT TCC TGT CCT CCC-3′. The reverse BC1 primer used for the reverse transcription was 5′-AAA GGT TGT GTG TG-3′, and the PCR primers were 5′-CTC AGT GGT AGA GCG CTT G-3′ (forward) and 5′-GGT TGT GTG TGC CAG TTA CCT TG-3′ (reverse). BC1 mRNA was used as an internal control for the PSD95 mRNA.
All of the quantitative RT-PCR data were representative of at least three independent experiments and are presented as the mean ± SEM. The differences were considered statistically significant at p < 0.05 using Student's t-test. Prism 5.0 software (GraphPad, Inc., La Jolla, CA, USA) was used for the data analysis.
Characterization of the miRNA profiling in synaptic fractions
In this study, we isolated synaptosomes from mouse forebrains using a sucrose density gradient centrifugation procedure. Prior to characterizing the expression profile of the miRNAs, we identified the synaptosomes using several methods. Fig. 1a illustrates a representative image of synaptosomes: synaptosomes are 0.5 μm-diameter, sealed membrane vesicles that contain synaptic vesicles and occasional mitochondria (arrows). We also determined the neurotransmitter levels in the synaptosome supernatants after the depolarizing stimulus and observed expected elevations of glutamate and GABA after the stimulation (Fig. 1b). To exclude the possibility of non-specific synaptosome leakage, a LDH activity assay was employed to monitor the leakage in synaptosomes. As shown in Fig. 1c, the LDH activity of the supernatants was less than 3% that of the LDH activity in the pellets, and there was no significant difference between the control and KCl-stimulated groups. Western blotting analysis showed that the synaptic vesicular protein, synaptotagamin1 (Syt1), was enriched in the synaptosome fraction (Syn). Furthermore, the post-synaptic protein PSD-95 was enriched in the insoluble fraction (Sp) after TritonX-100 extraction (Fig. 1d). These experiments confirmed that the synaptosomes were relatively intact and enriched in our preparation and that the synaptosomes displayed active metabolic function.
Next, the miRNA expression profile of the synaptosomes was analyzed using a qRT-PCR -based microRNA microarray. The results showed that approximately 33% of the miRNAs examined (192 of 585) were highly expressed in the synaptosomes (with an average threshold criterion of CT ≤ 25.0; Table S1). Among these highly expressed miRNAs, twenty-seven were further verified by qRT-PCR using TaqMan microRNA probes (Fig. 2a). The miRNAs were normalized against the U6 control, we found 16 miRNAs (miR-9, miR-16, miR-26a, miR-26b, miR-29a, miR-29c, miR-30a, miR-99b, miR-124, miR-125a, miR-125b, miR-128, miR-134, miR-138, miR-146a, and miR-146b) were highly expressed. Four miRNAs (miR-20a, miR-27b, miR-99a, and miR-130a) were moderately expressed, and 7 miRNAs (miR-15a, mir-21, miR-25, miR-27a, miR-34a, miR-148a, and miR-154) were poorly expressed in the synaptosomes. From these results, we found that brain tissue-enriched miRNAs (miR-9, miR-124, miR-128 and miR-138) were highly expressed in the synaptosomes. The regression coefficient between the expression profiles obtained from the microarray and the qRT-PCR assay was approximately 0.7 (Fig. 2b), indicating that the profiles obtained using the microRNA microarray and qRT-PCR was quite consistent. Among the synaptosome miRNAs studied here, miR-146b exhibited the highest expression levels. This result is consistent with a previous report that found that the synaptic intensity of miR-146b was among the highest expressed miRNA group (Lugli et al. 2008).
To evaluate whether miRNAs in synaptosomes were ‘naked’ or were associated with the protein complex required for miRNA function, we performed an immunoprecipitation assay using anti-Ago2 antibodies and determined whether miRNAs were co-immunoprecipitated with Ago2. Western blotting showed a relatively high efficiency of Ago2 immunoprecipitation and a high specificity of the Ago2 antibodies (Fig. 2c, upper panel and lower panel). We checked 21 miRNAs in the synaptosome lysates (open bar, Fig. 2d), and the anti-Ago2 antibody pulled down all 21 miRNAs from the lysates in the synaptosomes (shaded bar, Fig. 2d); however, different miRNAs had a different ratio of Ago2-associated/total. These results suggest that a portion of the miRNAs in the synaptosomes were associated with Ago2 and may exhibit biologically active functions.
We further prepared crude synaptic vesicles and synaptosomes from equal-volume P2 fractions. Western blotting revealed a relatively high enrichment of Syt1 protein and poor expression of PSD-95 protein in the SV fraction. Furthermore, we examined the protein levels of Ago2 in these fractions and found that Ago2 was relatively enriched in the SV fraction compared to the P2' or Syn fractions. The LP1 fraction also exhibited high expression of Ago2 protein (slightly higher than in the SV fraction) (Fig. 2e). The LP1 pellets were PSD-enriched fraction, and the high expression level of Ago2 was consistent with a previous report (Lugli et al. 2008). Next, we quantified 8 miRNAs from synaptic vesicles or the synaptosome fraction and found that, on average, approximately 31 ± 6% miRNAs from the synaptosomes were present in synaptic vesicles (Fig. 2f, open and shaded bars). These results suggest that a group of miRNAs is associated with the Ago2 protein in the synaptic fraction, and furthermore, several had been engulfed by synaptic vesicles.
Three miRNAs (miR-29a, miR-99a, and miR-125a) were increased in the synaptosome supernatants after depolarization in a Ca2+-dependent manner
In a chemical synapse, neurotransmitters are released into the synaptic cleft after the arrival of an action potential at the axon terminal. Because we found miRNAs were also present at the synaptic vesicles, we hypothesized that miRNAs might be actively secreted from the pre-synaptic region of the synaptosomes. To test whether a depolarizing stimulus results in miRNA secretion from synaptosomes, we quantified 20 miRNAs in the synaptosome supernatants in the presence or absence of a depolarizing stimulus. We found that the levels of miR-29a, miR-99a, and miR-125a were significantly elevated in the synaptosome supernatants after a depolarizing stimulus (Fig. 3a). To test whether these secreted miRNAs were potentially active, we performed an Ago2 immunoprecipitation assay. Similar to the miRNAs found in the synaptosome, a large proportion of the secreted miRNAs (miRNAs in the supernatant) was associated with Ago2 (Fig. 3b).
As synaptosome neurotransmitter release is a Ca2+-dependent exocytosis process, we investigated whether miRNA secretion from synaptosomes is also Ca2+ dependent. We applied depolarizing stimuli to synaptosomes in a Ca2+-free solution (Ca2+-free aCSF) and found no obvious elevation in the expression levels of the three miRNAs (Fig. 3c). Furthermore, OA, an exocytosis inhibitor, blocked the secretion of miRNA in Ca2+-containing aCSF (Fig. 3d). A second exocytosis inhibitor, TeXN, exhibited a moderate, but consistent inhibitory effect on miRNA secretion (Fig. 3e). These results indicated that physiological stimulation events such as depolarization induce synaptosomes to secrete miRNAs and that this process is dependent on calcium influx and may be mediated via the exocytosis pathway.
NGF enhances the stimulated secretion of miRNAs without altering the spontaneous release of miRNAs during incubation, whereas KA decreased the stimulated secretion of miRNAs and elevated the spontaneous release of miRNAs during incubation
We pre-treated the synaptosomes with several neuroactive molecules and then examined whether these molecules could modulate miRNA secretion. NGF is a neurotrophic factor, which promotes neuronal growth, maintains the survival of differentiated neurons and protects neurons from neurotoxins. We pre-treated the synaptosomes with NGF (100 ng/mL) and quantified the spontaneous release of miRNAs during NGF treatment. There was no significant difference in the miRNA content of the supernatants (Fig. 4a); however, NGF pre-treatment enhanced miRNA secretion after a depolarizing stimulus (Fig. 4b). Moreover, KA, a neurotoxin that binds to ionotropic glutamate receptors and transduces a secondary signal that mimics the toxic effects of a high dose of glutamate, was applied to the synaptosome supernatants. Compared with the control group, the spontaneous release of all three miRNAs from the synaptosomes was significantly enhanced with KA treatment (Fig. 4c), but the expression of miRNAs in the supernatants after the depolarizing stimulus decreased (Fig. 4d). We also examined miRNAs in the synaptosome pellets and found no significant differences (data not shown). These results indicate that specific neuroactive molecules may regulate the spontaneous or stimulated secretion of miRNAs.
Synthesized miRNAs in the supernatant are taken up by the synaptosomes
Synaptosomes are sealed presynaptic nerve terminals, and occasionally with the post-synaptic densities (Fig. 1). Synaptosomes also take up materials from the environment to maintain metabolism and neurotransmitter cycles. We explored whether miRNAs could be taken up by synaptosomes. Synthesized exogenous miRNAs (miR-29a, -99a, -125a, and -138) or a negative control RNA were added to the aCSF and subsequently incubated with synaptosomes at 25°C for 1 h. After extensive washing and RNaseA digestion, mature miRNAs were extracted from the synaptosome pellets and quantified using qRT-PCR. MiRNA was significantly elevated in the synaptosome pellets compared with the negative control, and this increase was dose dependent. Interestingly, we found that in addition to the three secreted miRNAs, miR-138 was also taken up by the synaptosomes (Fig. 5a). These results suggest that the synaptosomes ingested the miRNA without sequence specificity. To test whether the miRNA ingestion was an endocytic process, we applied an endocytosis inhibitor (Dynasore at 100 nM or Pitstop2 at 30 μM) prior to miRNA incubation. We found that Dynasore could inhibit the uptake of miRNAs; however, Pitstop2 showed no effect on miRNA entrance (Fig. 5b). These results indicated that the synaptosomes ingested the miRNAs via a dynamin-dependent pathway.
Recently, Muddashetty et al. reported that miR-125a targets PSD95 mRNA (Muddashetty et al. 2011). Therefore, we tested whether PSD95 mRNA in the synaptosome was regulated after incubation with miR-125a. We performed an Ago2 immunoprecipitation on these synaptosome pellets after 4-h incubation with miR-125a. Quantification of the miRNAs showed that the level of Ago2-associated miR-125a was much higher in the miR-125a-incubated group than in the control group (Fig. 5c). This result indicates that the synthesized miR-125a is recruited to the RISC. Furthermore, RT-PCR revealed that the levels of PSD95 mRNA in the miR-125a-incubated synaptosome pellets were significantly lower than in the control group pellets, which suggests that elevated levels of miR-125a result in the degradation of PSD95 mRNA after entering into the synaptosome (Fig. 5d). Taken together, these results indicate that miRNAs may be taken up by synaptosomes via an endocytic process and that the ingested miRNAs are recruited to the RISC and can inhibit target genes.
Local translation in the synaptic fraction has been shown to play an important role in synaptic plasticity and provides a flexible and dynamic ability to regulate protein levels within a specific subcellular region (the synapse) of neuronal cells. Growing evidence has demonstrated that microRNAs are expressed in the synapse and can modulate local translation in response to various extracellular and intracellular signals (Wayman et al. 2008; Smalheiser and Lugli 2009; Wang et al. 2010; Siegel et al. 2011). Our results further demonstrated that synaptic miRNAs are associated with the Ago2 protein, the core component of the RISC complex, which suggests that synaptosome miRNAs are primed to repress protein translation at axon terminals. Furthermore, we identified a group of miRNAs from synaptic vesicles that may be released after depolarization. We further tested whether miRNAs could be released after depolarization and found that synaptosomes secrete miRNAs into the supernatant after a depolarizing stimulus. The secretion of miRNAs was Ca2+-dependent and could be inhibited by an exocytosis inhibitor. Theoretically, all of the miRNAs in synaptic vesicles are secreted after depolarization. Of the 20 miRNAs examined in the 5-min KCl-stimulation experiments, 3 miRNAs were significantly elevated in the supernatant after KCl stimulation, 12 were slightly increased, and 5 showed no changes. These results support our hypothesis that miRNAs are secreted by synaptosomes.
In this study, NGF treatment promoted synaptosome-stimulated secretion, whereas KA treatment enhanced the spontaneous release and subsequent repression of the stimulated miRNA secretion. The effects of KA on synaptosome miRNA secretion had some confounders. For example, the decreased stimulated secretion may have been caused by a reduction in the miRNA pool following the enhanced spontaneous release. Thus, we must be cautious when concluding that KA may repress the stimulated release of miRNAs. However, miRNA secretion (either spontaneous or stimulated) was low in the synaptosome pellets. Moreover, 5 min of depolarization did not cause an obvious change in the synaptosome miRNAs pools. These results suggested that different molecules might have different effects on miRNA secretion. Furthermore, the molecular mechanism of miRNA secretion remains unknown and requires further study.
It remains unknown whether the miRNA secretion profile varies with the stage or status of the synaptic fraction. Recently, Richardo-Casas et al. reported that the miRNA expression profile in the synaptoneurosome differed among specific brain regions (Pichardo-Casas et al. 2012), which raises the possibility that synaptic miRNA secretion may be regional or cell-type specific. It would be interesting to test whether different brain regions or cell types release different miRNAs. Ricardo-Casas and colleagues also found that KA injection alters miRNA levels different in brain regions. In contrast to their in vivo experiments, we pre-treated the synaptosomes ex vivo for 10 min immediately prior to KCl stimulation. This short period of KA treatment did not alter the level of miRNAs in the synaptosome pellets. Further study is required to examine synaptosome miRNA secretion after in vivo KA injection.
The destinations of the secreted miRNAs remain unclear. However, similar to neurotransmitter recycling, we hypothesized that the recipient cells reassemble the miRNAs and transport them to neurons or other glial cells. As synaptosomes can take up material from solutions, we tested whether secreted miRNAs could be taken up by synaptosomes in our study. All of the 4 synthesized miRNAs were taken up by synaptosomes in a dose-dependent manner. To exclude the possibility that these miRNAs did not enter but were merely attached to the synaptosomes, we treated the synaptosomes with RNaseA to specifically examine miRNA ingestion. We then compared the levels of miR-29a in the synaptosome pellets with or without RNaseA treatment. We found that RNaseA digestion did not further diminish the levels of miR-29a within the synaptosomes (data not shown), which indicates that the secreted miRNAs were taken up by the surrounding synaptic fraction rather than adhering to the cell surface. The synaptosomes exhibited no selectivity regarding the ingestion of synthesized miRNA, which indicates that the ingestion might be mediated via an endocytic process. We found that Dynasore, a dynamin inhibitor, showed a significant inhibitory effect on the miRNA ingestion. Taken together, these results indicate that secreted miRNAs may be taken up by synaptosomes via a non-selective endocytic pathway.
Evidence is needed to demonstrate that secreted miRNAs modulate mRNA translation in their recipient cells. Recent studies suggest that miRNAs regulate key protein levels in the synapse. For example, Lippi et al. reported that the miR-29a/b target of Arpc3 and fine-tunes structural synaptic plasticity by regulating the branching of the actin network in mature and developing spines (Lippi et al. 2011). In addition, Muddashetty et al. reported that miR-125a targets PSD95 mRNA, demonstrating that miRNA is involved in selectively regulating mRNA translation at the synapse in response to receptor activation (Muddashetty et al. 2011). In this study, we found that the level of PSD95 mRNA dramatically decreased after synaptosome incubation with miR-125a, which mimics conditions where secreted miRNAs enters the synapse and results in the degradation of target mRNAs. We consistently found enhanced levels of miR-125a associated with Ago2, which reflects the ingestion of miR-125a and its assembly with RISC as well as its function as an inhibitory factor that targets mRNA. Due to the low expression level of PSD95 mRNA in the synaptosomes and the further decrease caused by miR-125a, we failed to detect Ago2-associated PSD95 mRNA. Furthermore, the isolated synaptosomes lost their activity within several hours after preparation, which limited their use in further studies to examine the level of target proteins or the physiological function of the secreted miRNAs.
We postulated that miRNA secretion in nerve cell terminals might be a new type of signal transduction pathway. Future investigations are necessary to explore the detailed mechanisms regarding miRNA secretion in synaptic fractions and the potential functions of secreted miRNA in recipient cells. The validated and identified properties of the miRNA secretion pathway may provide a new understanding of its synaptic function and pattern of regulation.
This study was supported by grants obtained from the National Natural Science Foundation of China (30771036, 30771039, 30871019, 30890044, 30890032, 30988003, 31071232, 31000323, 31100777, and 90608010) and the National Basic Research Program of China (973 Program Nos. 2011CB504803). Jie Xu, Qun Chen, and Qipeng Zhang performed the experiments; Jie Xu and Qipeng Zhang analyzed the data and wrote the manuscript; and Chenyu Zhang, Ke Zen, and Qipeng Zhang designed the experiments and revised the manuscript. The authors have no conflict of interest to declare.