Considerable evidence has shown that nitric oxide (NO) functions in a wide variety of physiologic and pathophysiologic processes such as neurotransmission, synaptic plasticity, neuroprotection, neurotoxicity, and pathologic pain (Snyder,1992; Prast and Philippu,2001). Previously, a great deal of information about the role of NO has been obtained from studies with experimental settings where exogenous NO was given (mostly by NO donors) to cultured CGNs and CGN apoptosis was subsequently induced. However, the precise role of endogenous NO in developing CGNs in vivo remains unclear (Bobba et al.,2007). We therefore asked the question whether NO is necessary in sustaining the survival of CGNs during their development under physiological conditions. The cerebellar granule neurons (CGNs), where NO primarily produced by the neuronal nitric oxide synthase (nNOS) isoform (Oldreive et al.,2008), have long been used to study the molecular roles of NO during neuronal development. Endogenous NO from nNOS has been found to protect developing CGNs against the toxic effects of alcohol (Pantazis et al.,1998; Karacay et al.,2007; Bonthius et al.,2009). Recently, several studies have argued that NO does not always act as a pro-survival factor throughout CGNs development because NO supported the survival of CGNs in primary only within 3 hr of the induction of cell death by serum/potassium withdrawal (Bobba et al.,2007; Oldreive et al.,2008). This study argued that NO is not necessary in the early development of cerebellar granule neurons. So, the results of endogenous NO effects on developing CGNs are conflicting. We hypothesized that manipulating nNOS gene in CGNs in primary culture could be a helpful strategy to study the precise role of endogenous nNOS/NO in CGNs development.
Antisense oligodeoxynucleotides (AS-ODN) have been widely used in selective gene suppression in the nervous system. AS-ODN–induced regulation of many different types of proteins, including receptors (Wahlestedt et al.,1993; Pizzi et al.,1999; Kumar et al.,2001; Alessandri-Haber et al.,2003), transcription factors (Sugiyo et al.,2001), neuropeptides and neurotransmitters (Spampinato et al.,1994), and growth factors (Kitajima et al.,1999), has helped neurobiologists to understand certain physiological or pathological processes in a specific neural functions. Recently, AS-ODN has been used to explore the molecular events or signal pathways induced by functional gene expression in neuronal development (Formisano et al.,2007; Huh et al.,2008; Li et al.,2008). We also used AS-ODN in vivo to explore the molecular mechanisms of nNOS in motoneuron death (Zhou and Wu,2006). In the present study, the same AS-ODN strategy was used to study the role of endogenous nNOS in CGNs development. We reported here that down-regulation of nNOS protein decreased viability and survival of cultured CGNs in maturity.
The nNOS Gene Expressed in Developing CGNs In Vitro
Primary dissociated cerebellar cells from postnatal day 7 rat pups were cultured as described in the Experimental Procedures section and essentially contained granule neurons. These CGNs were characterized by a small soma (<10 μm in diameter), scant cytoplasm and two to six rather short, unbranched processes before 5 days in culture (DIC; Fig. 1A). After 7 DIC, the neurons were viable and fully differentiated and had integrated within a dense network (Fig. 1B). The normal mature neurons showed cell clumping with shrinkage of the cell body (Fig. 1C) and damaged neuritis across 10 to 14 DIC (Fig. 1C,D). The nNOS-positive CGNs were detectable after 7 DIC, and they were characterized by round or multipolar soma and two to six long, branched processes (Fig. 1E). The nNOS immunofluorescence was colocalized with the Neurofilament (a neuronal marker), distributed in the cytoplasma surrounding the nucleus (Fig. 1E,F). The nNOS-positive reaction was not detected in CGNs controls, which were incubated without primary or secondary antibody, or when nNOS antibody was preincubated with nNOS blocking peptide. The endogenous nNOS gene in cultured CGNs was further confirmed by both Western blot and reverse transcriptase-polymerase chain reaction (RT-PCR) analysis. The nNOS protein was detectable from 7 to 14 DIC (Fig. 1G, above). The nNOS mRNA was detected from 4 to 14 DIC, earlier than the nNOS protein (Fig. 1H, above). Both nNOS mRNA (Fig. 1H) and protein (Fig. 1G) increased until the end of the study, that was at 14 DIC.
Antisense ODN Down-regulated nNOS Gene in Cultured CGNs in Maturity
Because down-regulation of nNOS by AS-ODN required efficient ODN delivery into cultured CGNs, we first determined whether FITC labeled R-ODN, could be taken up by CGNs in culture. According to the protocol from the Biognostik Company and our previous study in vivo, we added 8 μM FITC-R-ODN to CGNs at 7 DIC. The cell uptake of R-ODN was assessed by the distribution of fluorescein isothiocyanate (FITC) inside the CGNs (Fig. 2A–F). The results showed that the labeled CGNs were first detected at 2 hr (Fig. 2B), increased from 4 to 8 hr (Fig. 2C), accumulated in more than 90% of the cultured CGNs within 24 to 48 hr (Fig. 2D,E), and then gradually attenuated from 72 hr (Fig. 2F). The labeled cells showed normal morphology and cell density in culture. The effective time of AS-ODNs on cultured CGNs was set as 72-hr therefore. To minimize the cytotoxicity of the phosphorothioate modification of the ODNs, the R-ODNs at 2, 4, 8, 10, or 20 μM was added to the culture medium of CGNs at 7 DIC using the same volume of TE-buffer as negative controls. Then cell viability of CGNs at 10 DIC was detected by MTT assay (Fig. 2G). Statistical analysis showed that R-ODN at concentrations above 10 μM significantly impaired the viability (Fig. 2G). So the concentration of 8 μM was used to test the activity of AS-ODN on nNOS gene in cultured CGNs.
Because the specificity of the present nNOS AS-ODN was shown to be effective in inhibiting nNOS enzyme activity in spinal motoneurons in vivo in our previous study (Zhou and Wu,2006), we tested AS-ODN effects on both nNOS protein and nNOS mRNA levels in cultured CGNs. AS-ODN at 8 μM was added to CGNs at 7 DIC for 72 hr, with 8 μM R-ODN and the same volume of TE-buffer as controls. The Western blotting result showed that the level of nNOS protein in AS-ODN treatment was the lowest, decreased by approximately 59% compared with that in R-ODN, and by 60% compared with that in TE-buffer controls (Fig. 2H, above). Statistical analyses showed that the optical density ratio of nNOS/β-actin proteins was not different between R-ODN and TE-buffer controls, but it was significantly lower in AS-ODN –treated CGNs at 10 DIC compared either with R-ODN or TE-buffer control (Fig. 2H, below). This result confirmed that AS-ODN effect on knockdown nNOS gene inside CGNs was sequence-specific because R-ODN had no effect on nNOS protein expression. The present study, using a sensitive and specific RNase protection assay, revealed the nNOS AS-ODN exposure for up to 72 hr had no effect on nNOS mRNA levels in cultured CGNs as compared to that of R-ODN or TE-buffer control (Fig. 2I).
nNOS AS-ODN Decreased the Viability and Survival of Cultured CGNs in Maturity
Because R-ODN at concentrations over 10 μM was shown to be cytotoxic (Fig. 2G) and AS-ODN at 8 μM was demonstrated to knockdown the nNOS protein levels in CGNs across 7 to 10 DIC, we chose 8 μM as an optimal concentration to test the activity of nNOS AS-ODN on the development of CGNs. AS-ODN at a concentration of 8 μM was added to CGNs at 7 DIC, with 8 μM R-ODN and the same volume of TE-buffer as controls. The MTT results showed that the mitochondrial activity was 0.36 ± 0.01 in AS-ODN-treated CGNs, 0.52 ± 0.05 in R-ODN, 0.54 ± 0.05 in TE-buffer and 0.53 ± 0.04 in blank controls (Fig. 2J). Statistic analysis showed the cell viability in AS-ODN treatment was significantly lower than any of the controls. The difference of cell viability in any other two controls was not significant. Considering that the cell uptake of ODNs in CGNs attenuated from 72 hr, we treated the CGNs with AS-ODN, R-ODN, or TE-buffer controls for every 72 hr. At the end of the study, the survival of CGNs at 13 DIC was quantified by Neutral red staining (Fig. 2L–O). Quantitative analysis showed the number of survival CGNs was 18.25 ± 1.38 in AS-ODN treatment, which was significantly lower than 64.50 ± 1.55 in R-ODN, 72.00 ± 4.02 in TE-buffer, and 74.75 ± 2.06 in blank controls. The number of surviving CGNs was not significantly different among the three controls (Fig. 2K).
Evaluation of Potential Pathways of nNOS on Cultured CGNs in Maturity
To demonstrate nNOS activity on the cell viability of developing CGNs in vitro, addition of the following NO donor or antagonists was performed at 7 DIC: SNP (Sodium Nitroprusside, an NO donor, 10 μM), nNOSi (nNOS inhibitor I, 100 μM), MK-801 (block of NMDA receptor, 30 μM), and ODQ (antagonist of sGC, 50 μM). After 24 hr treatment, the cell viability was assessed by MTT assay. The results showed SNP itself induced an increase of cell viability beyond control (0.398 ± 0.056 in SNP vs. 0.341 ± 0.03 in control). All antagonists decreased the cell viability of cultured CGNs (0.257 ± 0.04 in nNOSi, 0.277 ± 0.02 in MK-801, or 0.265 ± 0.04 in ODQ vs. 0.341 ± 0.03 in control). The antagonist-induced attenuated cell viability could be reversed by provision of SNP, respectively (0.364 ± 0.03 in SNP+nNOSi, 0.337 ± 0.02 in SNP+MK-801, or 0.341 ± 0.01 in SNP+ODQ vs. 0.341 ± 0.03 in control; Fig. 3).
In the present study, we applied RT-PCR, Western blot and immunofluorescence to confirm the expression pattern for nNOS gene in developing CGNs in vitro. The results correlate well with those reported by other authors (Giuili et al.,1994; Viani et al.,1997; Jurado et al.,2004). Furthermore, our study showed the increased nNOS expression, both in terms of mRNA and protein, during the maturity of the developing CGNs (Viani et al.,1997). Because we were just able to detect nNOS mRNA from 4 DIC but nNOS protein from 7 DIC, we considered that cerebellar granule cells lack nNOS activity during early differentiation and then acquire the enzyme as they mature.
We have developed a very useful approach, the antisense strategy, to manipulate the endogenous nNOS gene in CGNs in vitro. The following evidences confirmed that our AS-ODN effects were due to the antisense character of the sequence to the nNOS gene. First, the present AS-ODN may directly target nNOS mRNA inside CGNs because FITC-R-ODN was distributed in intracellular organelles in the cytoplasm (Iversen et al.,1992) of cultured CGNs. Second, the down-regulation of the nNOS protein was not caused by an artifact related to CGNs detachment from the culture dish or other nonsequence specific effect because the levels of β-actin were not changed and R-ODN did not affect any of the nNOS gene expressions. Third, the AS-ODN –induced decreases in CGN viability was due to functional inhibition of the nNOS gene, not caused by the cytotoxicity of phosphorothioate modification because R-ODN did not cause any significant change in CGN viability. Indeed, both gene knockdown methods and pharmacological inhibitors have the potential to cause nonspecific effects, and we believe that employment of both inhibitory strategies can help rule out such nonspecific effects because they usually involve different off-target mechanisms.
Our present data first showed AS-ODN did down-regulate nNOS gene expression but in a posttranscriptional manner. Our AS-ODN was designed to hybridize to its specific mRNA target, which has been shown to create a conformational obstacle for protein translation (Kretschmer-Kazemi Far and Sczakiel,2003). Previous studies have demonstrated that AS-ODN could enter cells and stop protein translation by either blocking the translocation of ribosomes (Boiziau et al.,1992) or destroying the target mRNA through RNase H-mediated degradation (Sommer et al.,1998). Intracerebroventricular (i.c.v) infusion of AS-ODN caused a decrease in nNOS mRNA but not NOS enzymatic activity in rat hippocampus (Kolesnikov et al.,1997). Most studies, however, found that nNOS AS-ODN could down-regulate nNOS enzymatic activity and regulate nNOS-related neuronal behavior whether or not it suppressed nNOS mRNA (Parmentier-Batteur et al.,2001; Li et al.,2003).
The major finding of this study was that knockdown of the nNOS gene impaired the viability and increased the death of developing CGNs across 7 to 14 DIC. Because previous studies showed that purified CGNs were immature during the first 7 DIC and ageing after 11 to 14 DIC (Diaz et al.,2002; Manzini et al.,2006), we considered that the expression of nNOS protein might play an important role in the transition of CGNs from immature to maturity across 7 to 14 DIC. Previous studies have suggested that the mechanism of nNOS in CGNs development likely involves its principal product, NO (Ciani et al.,2002; Jurado et al.,2004; Bonthius et al.,2009). Our present results further support this hypothesis, because inhibition of the nNOS enzyme induced a decrease of cell viability of CGNs and this change could be returned by provision of exogenous NO donor, SNP. Moreover, we found that NMDA receptor antagonist (MK-801), which has proven to reduce the production of endogenous NO (Baader and Schilling,1996), induced decreases in cell viability and this change could also be reversed by provision of SNP.
Physiologically, NO activated soluble guanylate cyclase (sGC), leading to increased levels of the second messenger cyclic GMP (Moncada and Higgs,2006). So a block of the NO-cGMP pathway by the antagonist of sGC, ODQ, decreased the cell viability of CGNs in vitro. But the present data also showed ODQ did not completely block SNP-mediated increase of cell viability of cultured CGNs. It suggests that NO-cGMP may not be the sole pathway for NO-mediated active role in cell viability of developing CGNs in maturity. To date, two major signaling mechanisms were considered to mediate the cellular effects of NO, namely the cGMP pathway and S-nitrosylation. The latter has been found to mediate NO-induced anti-apoptosis function and neuronal protection recently (Tenneti et al.,1997; Lipton,1999; Kawano et al.,2009).
In summary, the present study demonstrated knockdown of endogenous nNOS protein decreased the viability of the cerebellar granule cells in transition from differentiation to maturity in vitro. The mechanism of AS-ODN–induced death of CGNs might be attributed to the reduced level of NO in CGNs cultures. Our present data suggest that a baseline of nNOS protein in developing CGNs is essential for the survival of the cerebellar granule neurons.
Cerebellar Granular Cell Culture
Cerebellar granule cell cultures were established from 7-day-old Sprague Dawley rats as described previously (Foister et al.,2005). In brief, the whole cerebellum was removed from individual rat, and all cerebella from a single litter were pooled for each replicate of the experiment. The pooled cerebella were dissociated, and the cells were plated at a density of 1 × 106 per dish precoated with 10 mg/L poly-L-lysine (Sigma, UK). Cells were maintained in Basal-Eagle's culture medium containing 10% fetal bovine serum (Globepharm, UK), 100 μg/ml gentamicin and 25 mM KCl. Cultures were treated with medium containing 10 μM cytosine arabinoside (Sigma, UK) after 24 hr to inhibit proliferation of non-neuronal cells. According to previous work (Foister et al.,2005), these cultures can be assumed to be at least 98% pure granule neurons.
Culture Treatments With Antisense ODNs
All oligos used in this study were purchased from Biognostik (GmbH, Goettingen, Germany). The nNOS AS-ODN consisted of a 14-bp sequence (GGA GAC GCA CGA AG) complementary to the translation start region of the rat nNOS mRNA (GenBank accession no.U24703). Random sequence oligos (R-ODNs) (ACC GACCGACGTGT), which had the same sequence length, G/C content, and phosphorothioate modifications as AS-ODN were set as sequence controls. TE buffer served as the buffer controls. FITC-labeled R-ODNs were used to check the cellular uptake of oligos. All oligos used in this study were subjected to a BLAST search, and no cross-homologies were found in the GenBank database. The nNOS AS-ODN had a positive match only for the targeted nNOS mRNA sequence. No positive matches were produced for the R-ODN. According to our previous study (Zhou and Wu,2006), a concentration of 8 μM FITC-R-ODN was applied to CGNs at 7 DIC and imaged under a fluorescence microscope to detect the uptake of ODNs by cultured CGNs. The 72 hr was set as the optimal transfaction time of antisense oligos. To minimize the cytotoxicity of the phosphorothioate modification of the ODNs, we tried concentrations of 2, 4, 8, and 10 μM nNOS AS-ODN applied to culture CGNs and compared with the effects of the R-ODN controls. A concentration of 8 μM of the nNOS AS-ODN was found to be optimal, and this dose was used in the present knockdown study. The nNOS AS-ODN or R-ODN was added to 7 DIC CGNs at 8 μM for 72 hr, and nNOS expression was examined 72 hr later. To assess the knockdown nNOS activity on the survival of cultured CGNs, the nNOS AS-ODN or R-ODN was added to CGNs from 7 DIC then renewed for every 72 hr.
To determine the AS-ODN effect on nNOS gene expression, total RNA samples were isolated from 10 DIC CGNs using the TRIzol (TIANGEN BIOTECH, CHINA) method. RT-PCR and complementary strand DNA were synthesized according to the manufacturer's instructions (AMV Ver 3.0 RNA PCR kit, Takara Biotechnology, China). After concentration determination on a ultraviolet spectrophotometer (Eppendorf Biophotometer), 1 μg of cDNA was used as the template for amplification of the nNOS message. The nNOS and GAPDH were amplified using the following primers: nNOS upstream, 5′-TGGCATAGGC TTGTGATTT-3′; nNOS downstream, 5′-GGCGTCCGTGACTA CTGT-3′, PCR product size, 532 bp; GAPDH upstream, 5′-GGCATCCTGACCCTGA AGTA-3′; GAPDH downstream, 5′GC CGATAG TGATG ACCTGACC-3′, PCR product size, 495 bp.
Western Blotting Analysis
The cell pellets were scraped down and the protein extract was isolated using a RIPA buffer (Santa Cruz Biotechnologies, Santa Cruz, CA). The protein concentration was measured using the BCA method with a Bio-Rad Protein Assay Reagent (Bio-Rad, Hercules, CA). For Western blotting, samples (20 μg protein/lane) were separated on 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Pall, USA). To detect nNOS protein, the membranes were blocked with 5% nonfat dry milk in Tris-buffered saline (TBS) containing 0.1% Tween-20 at room temperature for 2 hr and then incubated with the anti-nNOS antibody (diluted 1:500, Santa Cruz Biotechnologies, Santa Cruz, CA) or with anti–β-actin antibody (diluted 1:1,000, Boster, CHINA) overnight at 4°C. After washing and incubating for 2 hr at room temperature with a secondary antibody conjugated with horseradish peroxidase (diluted 1:4,000, Amersham Biosciences, Piscataway, NJ), the blots were washed and developed by chemiluminescence according to the manufacturer's protocol (SuperSignal West Pico Chemiluminescence Substrate, Pierce). Blots were exposed to X-ray film (FUJIFILM, Japan), and the intensity was quantified for both bands. The relative level of protein in the different lanes was compared by analyzing scanned images using the NIH IMAGE program. All studies were performed a minimum of three times using independent cultures.
Immunofluorescence was carried out according to standard protocols. Briefly, the cells were fixed with 4% paraformaldehyde and washed in phosphate buffered saline (PBS) three times for 30 min, then treated with 0.1% bovine serum albumin, 0.3% Triton for 30 min. Cells were then incubated overnight with mouse anti-nNOS (diluted in 1:2,000; Santa Cruz, CA) and rabbit anti-Neurofilament 200 (diluted in 1:400; Sigma, St. Louis, MO) at 4°C overnight. After washing in PBS, cells were incubated with Anti-mouse IgG tetrarhodamine isothiocyanate (TRITC) conjugate (diluted in 1:1000, Sigma) and anti-rabbit IgG FITC conjugate (diluted in 1:400, Sigma) in the dark for 1 hr. Then the cells were counterstained with Hoechst 33258 (2 g/ml, Sigma). Control experiments included omission of the primary and secondary antibodies, or preincubated the nNOS antibodies with the nNOS blocking peptide (amino acid 1422-1433; ESKKDTDEVFSS, Cayman Chemical, USA) for 2 hr.
Addition of the following drugs was performed at 7 DIC: SNP (sodium nitroprusside, an NO donor, 5–50 μM, Sigma), nNOSi (nNOS inhibitor I, 10–200 μM, Calbiochem, Los Angeles, CA), MK-801 (block of NMDA receptor, 10–100 μM, Calbiochem), and 1H-[1,2,4]Oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, antagonist of sGC, 10–100 μM, Calbiochem). Effects of the different drugs on cell viability of CGNs were first evaluated by testing different concentrations of each drug and verifying the cell viability of the cultured CGNs. Concentrations (SNP at 10 μM, nNOSi at 100 μM, MK-801 at 30 μM, and ODQ at 50 μM) chosen for use were the lowest that did not promote any visible toxic effect but significantly decreased the cell viability. Cultures were incubated with different drugs at the indicated concentrations for 24 hr. Control cultures were run in parallel without the presence of the same drugs. To evaluate the different pathways which mediated the NO effects on developing CGNs, the different antagonists were applied to cultured CGNs at 7 DIC 2 hr before the following 24 hr SNP treatment.
Assessment of Cell Viability
Mitochondrial activity, as a measure of cell death, was analyzed by measuring the dehydrogenase activity levels by means of the MTT assay as previously described (Gunn-Moore et al.,1997). Briefly, MTT (0.5 mg/ml, Sigma) was added to each well and incubated with cultured CGNs for a further 4 hr at 37°C, 5% CO2. At that point, all media were moved, and 150 μl dimethyl sulfoxide was added to each well to solubilize the formazan salt. In the present study, the resulting purple azo-dye was detected at 490 nm with a Biohit BP800 plate reader (in AS-ODN experiment) or with the Multiskan Ascent microplate reader (in pharmacological experiment). Data were expressed as a percentage mitochondrial activity relative to that detected in untreated wells to normalize for differences in plating density between individual experiments. The viability of CGNs was also assessed by neutral red staining (2 g/ml, Sigma), the stain for Nissl bodies in viable neurons. The number of neutral red–positive neurons was counted as survival CGNs in each group. Results were presented as mean ± standard error of the mean of triplicate samples from at least six separate experiments (n = 6).
To verify the survival of cultured CGNs, survival cells with Nissl bodies stained red dye. For the survival cell counting, six random microscopic fields per well were counted at ×400. The number of positive cells per well was expressed as the total number of cells indicated by Neutral red per field, the average value from the six random fields. Every treatment of the culture was repeated with three wells at each time and was repeated for four times. The average values (X) and standard deviations (SD) served as descriptive parameters. Data quantification and analysis, in quantifying the bands densities of RT-PCR and Western blotting results, in counting survival CGNs, and in measuring MTT of CGNs cultures, were performed by two independent persons, both of whom were blinded to the treatment of the culture. Statistical calculations and data handling were performed by one-way analysis of variance as well as a Bonferroni post hoc multiple comparison tests using SPSS version 11.0. The statistical significance was set at P < 0.05.