Rsk2 Knockdown in PC12 Cells Results in Sp1 Dependent Increased Expression of the Gria2 Gene, Encoding the AMPA Receptor Subunit GluR2

The RSK2 protein is a member of the RSK serine-threonine protein kinase family and is encoded by the X-linked rps6ka3 gene in human. Highly heterogeneous loss-of-function mutations affecting this gene are responsible for a severe syndromic form of cognitive impairment, Coffin-Lowry syndrome. RSK2, which is highly conserved in mammals, acts at the distal end of the Ras-ERK signaling pathway and is activated in response to growth factors and neurotransmitters. RSK2 is highly expressed in the hippocampus, and Rsk2-KO mice display spatial learning and memory impairment. We recently showed that ERK1/2 activity is abnormally increased in the hippocampus of Rsk2-KO mice as well as the expression of the AMPA receptor subunit GluR2. The mechanism via which RSK2 deficiency affects the expression of GluR2 in neural cells was unknown. To address this issue we constitutively suppressed the expression of RSK2 in PC12 cells via vector-based shRNA in the present study. We show that Rsk2 silencing leads also to an elevation of ERK1/2 phosphorylation as well as of GluR2 expression and that the increased level of GluR2 expression results from the increased ERK1/2 activity on the transcription factor Sp1. Our results provide evidence that RSK2 modulates ERK1/2 activity on Sp1, which regulates GluR2 expression through transcriptional activation.


Introduction
The 90 kDa ribosomal S6 kinases (RSKs) constitute a family of four homologous Ser/Thr kinases (RSK1-4) that are activated by Mitogen activated protein kinase/extracellular-regulated kinases 1 and 2 (MAPK/ERK1/2) in response to growth factors, hormones, chemokines and neurotransmitters. When activated by Extracellular signal-regulated kinase (ERK), a fraction of RSK2 in turn phosphorylates a number of cytoplasmic substrates such as BCL2-associated agonist of cell death (Bad), p53, L1 cell adhesion molecule (L1CAM) and Glycogen synthase kinase 3 (GSK3) [1]. Another fraction of RSK2 translocates into the nucleus where it is thought to regulate gene expression through phosphorylation of transcription factors such as cAMP-response element binding protein (CREB), the cellular FBJ murine osteosarcoma viral oncogene homolog (c-fos), the oestrogen nuclear receptor, the nuclear receptor subfamily 77 (Nurr 77) and the Activating transcription factor 4 (ATF4) [1,2]. The roles of RSKs in cellular signaling, cell survival, growth and differentiation has been well established; however, their implication in biological processes in vivo is less well known. An important physiological role of one RSK member, RSK2, was uncovered by the discovery of Rsk2 gene defects in the Coffin-Lowry Syndrome (CLS) [3]. CLS is an X-linked disorder with progressive skeletal abnormalities, facial dimorphism, and strong psychomotor impairment [4]. The vast majority of male CLS patients show severe cognitive deficiency with intelligence quotient (IQ) ranging from 15 to 60. Delay in speech acquisition is particularly common with most affected males having a very limited vocabulary. The mouse knockout (KO) model (Rsk2-KO) for CLS exhibits a severe impairment in spatial learning and a deficit in long-term spatial memory [5]. Moreover, in human and mouse brain, RSK2 is highly expressed in the hippocampus, an essential brain structure in cognitive function and learning [6,7]. Together the data suggest that RSK2 plays an important role in cognitive function in human and in mice.
To gain greater insight into the molecular mechanisms leading to learning and memory impairments in the Rsk2-KO mice and to intellectual disability in CLS, we examined recently global gene expression profiles in hippocampus from KO mice, It revealed altered expression of 100 genes encoding proteins acting in various biological pathways [8]. We investigated in detail the consequences of the two-fold up-regulation of one of these genes, Glutamate receptor, ionotropic 2 (Gria2), which encodes the subunit GluR2 of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor. Abnormally increased expression of GluR2 at the protein level was confirmed in the hippocampus of Rsk2-KO mice and at synapses from Rsk2-KO hippocampus primary neuron cultures. Basal AMPA receptor-mediated activity was found to be significantly decreased [8]. RSK2 deficient mice have also an increased baseline level of phospho-ERK1/2 as well as enhanced activation of ERK1/2 after glutamate stimulation of hippocampal neurons [9]. These latter findings definitively demonstrated that RSK2 exerts a feedback inhibitory effect on the ERK1/2 pathway in physiological conditions.
To define the molecular events leading to GluR2 over expression in RSK2 deficient neurons we used in the present study the RNA interference technology to knockdown the Rsk2 gene in PC12 cells. We show that Rsk2 silencing in PC12 cells leads also to an increase of ERK1/2 activity and of GluR2 expression at the mRNA and protein levels. Our data reveal that the increased GluR2 expression results from the increased ERK1/2 activity on the transcription factor Specificity Protein 1 (Sp1).

Rsk2 Knockdown of PC12 Cells
For this study we used the PC12 cell line, an immortalized cell line derived from a pheochromocytoma tumor of the rat adrenal gland. NGF-treated PC12 cells cease proliferation, grow long neurites, and show changes in cellular composition associated with neuronal differentiation [10], and can therefore serve as a model system for primary neuronal cells. PC12 cells express of RSK2 as shown in (Figure 1). We tested the efficacy to block RSK2 expression in PC12 cells of three plasmids producing different RSK2-specific small hairpin RNAs (Rsk2-shRNA) targeted to different coding regions of the mRNA of RSK2. The recombinant plasmids were transfected into the PC12 cells by lipofectamine 2000, recombinant cells selected 72 h on G418, and mRNA and protein expression levels of the Rsk2 gene were assayed by QRT-PCR and Western blot. One shRNAs (shRNA clone ID7) decreased Rsk2 gene expression dramatically (Figure 1a). Quantification of the immunoblots revealed a 90% reduction in the expression of RSK2 protein, compared to untransfected PC12 cells. No significant change was found in the level of RSK2 mRNA and protein expression in cells transfected with a shRNA containing a scrambled sequence (negative control) ( Figure 1a). (Expression of RSK2 protein in Rsk2-shRNA transfected cells: 0.07 ± 0.03; untransfected cells: 1.0 ± 0.09; scrambled-shRNA transfected cells: 0.917 ± 0.1; * p < 0.05 by Student's t-test) (Figure 1a). QRT-PCR revealed also a dramatic decrease of RSK2 mRNA expression (Expression of RSK2 mRNA in Rsk2-shRNA transfected cells: 0.316 ± 0.05; untransfected cells: 1.0 ± 0.16; scrambled-shRNA transfected cells; 0.865 ± 0.15; * p < 0.05 by Student's t-test) ( Figure 1c). Thus, the shRNA clone ID7 was used in all subsequent studies. We then asked whether the severe down regulation of RSK2 would also affect the expression of other RSK proteins in PC12 cells. The result for RSK1 is shown in Figure 1b. (Expression of RSK1 protein in Rsk2-shRNA transfected cells: 0.895 ± 0.1; untransfected cells: 1.0 ± 0.28; scrambled-shRNA transfected cells: 0.894 ± 0.09; * p < 0.05 by Student's t-test). RSK3 was very weakly expressed in untransfected PC12 cells and was unchanged in transfected PC12 cells (result not shown). Thus, none of these two latter RSK family members apparently exhibited any significant change in its expression levels, documenting that the Rsk2-shRNA approach was highly specific for RSK2 and reducing the possibility of compensation for the RSK2 loss via over expression of other RSK family members.

Immunohistochemical Analysis of RSK2
We also examined the expression of RSK2 in PC12 cells by immunohistochemical staining. Most of untransfected, or scrambled-shRNA transfected PC12 cells showed strong RSK2 staining, whereas Rsk2-shRNA transfected cells expressed RSK2 very weakly ( Figure 2). It was concluded that the Rsk2-shRNA recombinant plasmid can effectively suppress the expression of the Rsk2 gene in PC12 cells.

Rsk2 Knockdown Results in up Regulation of Gria2 Gene (Encoding GluR2) Expression
We next examined the expression of GluR2 in Rsk2 knockdown PC12 cells by Western blot analysis and QRT-PCR. Detection of GluR2 protein levels, revealed a very strong increase of expression in Rsk2 knockdown PC12 cells (Expression of GluR2 protein in Rsk2-shRNA transfected cells: 1.0 ± 0.02; untransfected cells: 0.03 ± 0.008; scrambled-shRNA transfected cells: 0.382 ± 0.04; * p < 0.05 by Student's t-test) (Figure 4a). QRT-PCR revealed also an approximately four-fold higher Gria2 mRNA expression in Rsk2-silenced PC12 cells compared to untransfected cells (Expression of Gria2 mRNA in Rsk2-shRNA transfected cells: 1.0 ± 0.04; untransfected cells: 0.225 ± 0.007; scrambled-shRNA transfected cells: 0.476 ± 0.06; * p < 0.05 by Student's t-test) (Figure 4b). Cells transfected with the scrambled-shRNA showed a small increase of both GluR2 protein and mRNA but much lower than Rsk2-shRNA transfected cells, and which may be related to the small increase of ERK1/2 phosphorylation observed in these cells (Figure 4a,b). We assayed also GluR1 and GluR3 levels of expression. Detection of GluR1 protein levels revealed a small decrease of expression in Rsk2-shRNA transfected PC12 cells compared to untransfected PC12 cells or cells transfected with the scrambled-shRNA (Expression of GluR1 protein in Rsk2-shRNA transfected cells: 0.803 ± 0.11; untransfected cells: 1.0 ± 0.14; scrambled-shRNA transfected cells: 0.951 ± 0.09; * p < 0.05 by Student's t-test) (Figure 4c). GluR3 was hardly detectable in untransfected PC12 cells as well as in Rsk2-and scrambled-shRNA transfected cells (not shown), suggesting that expression of GluR3 was not altered in RSK2 deficient cells. These results showed that only the expression of GluR2 is up regulated in response to RSK2 deficiency and suggested also that the regulatory mechanisms of expression are different for each GluR gene.

Inhibition of GluR2 over Expression in Rsk2 Knockdown PC12 Cells by U0126
We then evaluated if the increased level of active ERK may be the cause of the strong increase of GluR2 expression in Rsk2 knockdown PC12 cells. To this end, Rsk2-and scrambled-shRNA transfected PC12 cells as well as untransfected cells were treated for 24 h, before being harvested, with U0126 (20 μM), a MEK inhibitor (MEK being the ERK1/2 upstream kinase in the ERK/MAPK pathway).

Involvement of Sp1 in the up Regulation of GluR2 Expression in Rsk2 Knockdown PC12 Cells
It was previously shown that GluR2 expression is strongly influenced at the transcriptional level by a positive Sp1 regulatory element in the 5' proximal region of the Gria2 gene promoter [11]. The transcription factor Sp1 recruits basal transcription factor TFIID to DNA and induces transcription. It was also reported that Sp1 activity increases when phosphorylated by ERK at two specific threonine residues (Thr453 and Thr739) [12]. Thus, we wondered whether the expression and/or the phosphorylation of Sp1 are also up regulated in Rsk2 knockdown PC12 cells. Total protein was extracted for analysis of Sp1 and Phospho-Sp1 (P-Sp1) expression by Western blotting. Total Sp1 levels, revealed no difference of Sp1 mRNA and protein expression between untransfected and Rsk2-shRNA transfected PC12 cells (mRNA expression of Sp1 in Rsk2-shRNA transfected cells: 1.0 ± 0.009; untransfected cells: 1.0 ± 0.006; scrambled-shRNA transfected cells: 0.921 ± 0.01; expression of Sp1 protein in Rsk2-shRNA transfected cells: 1.0 ± 0.12; untransfected cells: 0.89 ± 0.1; scrambled-shRNA transfected cells: 0.961 ± 0.06 * p < 0.05 by Student's t-test) (Figure 6a-c). However, immunoblot analysis of phosphorylated forms of Sp1 showed that the levels of phosphorylation at both phosphorylation sites, T453 and T739, were significantly higher in   To further determine whether treatment with the MEK inhibitor U0126 impacts Sp1 phosphorylation, Rsk2-shRNA transfected and untransfected PC12 cells were treated with this inhibitor (20 μM) for 24 h before being harvested. Total proteins were extracted for Western blot analysis and Sp1 protein levels, revealed no difference between Rsk2-transfected, untransfected and scrambled cells (expression of Sp1 in Rsk2-shRNA untransfected: 0.91 ± 0.04 transfected cells: 1.0 ± 0.05; scrambled shRNA transfected cells: 0.98 ± 0.03). (Figure 6a,b) Inhibition of ERK1/2 activity resulted in a dramatic decrease of Sp1 phosphorylation at both T739 and T453 sites (Expression of P-Sp1 (T453) after U0126 treatment in Rsk2-shRNA untransfected cells: 0.09 ± 0.002; transfected cells 0.13 ± 0.02; scrambled-shRNA transfected cells: 0.001 ± 0.005). Expression of P-Sp1 (T739) after U0126 treatment was undetectable (Figure 6a,e). These results demonstrated that ERK is the major kinase that phosphorylates Sp1 at both of these sites.

Confirmation of Significantly Increased Sp1 Phosphorylation in the Hippocampus of Rsk2-KO Mice
Western blot analysis of total protein extracts from hippocampi of Rsk2-KO adult mice also revealed significantly increased levels of P-Sp1 (T453) as compared to WT littermate.

Discussion
The (AMPA)-type glutamate receptor channels are expressed ubiquitously in brain neurons and mediate fast excitatory neurotransmission in most excitatory synapses. AMPA receptors are assembled from combinations of four subunits, GluR1, GluR2, GluR3 and/or GluR4. AMPA receptors play also a crucial role in synaptogenesis and formation of neuronal circuitry, as well as in synaptic plasticity [13]. There is a huge amount of evidence indicating that the GluR2 subunit plays a pivotal role in AMPA receptors function. The Ca ++ permeability, rectification, and single-channel conductance of AMPA receptors are all dominantly influenced by inclusion of an edited GluR2 subunit in the receptor complex [14][15][16][17][18][19][20]. Receptors that contain a single edited GluR2 subunit have maximally reduced Ca ++ permeability, whereas inward rectification is reduced in a graded manner as the number of GluR2 subunits in a receptor increase [20]. Thus, moderate changes in GluR2 expression may have functional consequences on receptor subunit rearrangement and lead to functional differences in the functioning of synaptic circuits. GluR2 expression is governed by neuronal activity [21], and its synaptic distribution and levels are modulated as part of the mechanism of synaptic plasticity [22,23]. GluR2 interacts also with various molecules implicated in receptor trafficking [24][25][26]. The crucial role played by GluR2 in neuronal function may therefore explain why it is the most tightly regulated of the AMPA receptor subunits [26]. The two-fold increased level of GluR2 expression in Rsk2-KO mice prompted us to elucidate the molecular mechanism leading to this alteration of expression by silencing the Rsk2 gene in PC12 cells. The PC12 cell line, cloned from rat pheochromocytoma cells, can be induced to differentiate and to acquire a neuronal-like phenotype [10]. PC12 cells express endogenous RSK2 at a relatively high level, whereas the level of GluR2 is low. Therefore, the PC12 cell line is a good model for our present study. As anticipated from our previous data in Rsk2-KO neurons [9], silencing of the Rsk2 gene led to an up-regulation of ERK1/2 activity and to an increase of GluR2 expression, both being, however, stronger than those observed in Rsk2-KO neurons [8,9]. Increase of GluR2 expression was observed at both the protein and mRNA levels, indicating that it was predominantly the transcription of the Gria2 gene that was affected by the absence of RSK2 activity. Since it was previously reported that the ERK pathway plays a role in the regulation of GluR2 expression [27], we hypothesized that the increased GluR2 expression level in the Rsk2-KO mice and in Rsk2 knockdown PC12 cells might be related to the elevated ERK1/2 activity. As expected, inhibition of ERK1/2 activity in the Rsk2 knockdown PC12 cells with the U0126 inhibitor reversed the effect of RSK2 depletion on GluR2 expression, clearly implicating ERK1/2 in the elevated GluR2 expression.
Previous investigations demonstrated a positive regulatory role for Sp1 in the transcriptional regulation of Gria2 gene via cis-acting elements in its proximal promoter regions [11]. Multiple signaling pathways converge on Sp1, including ERK, Akt, and c-Jun N-terminal protein kinase [28]. ERK-regulated phosphorylation sites on Sp1 were identified at T453 and T739 in response to growth factor regulation of the vascular endothelial growth factor gene [29]. Phosphorylation of Sp1 on T453 and T739 increases Sp1 DNA binding activity [29]. Our data show that levels of phosphorylation at both Sp1 sites is strongly increased in Rsk2 knockdown PC12 cells, whereas treatment with the MEK inhibitor U0126 inhibits phosphorylation at both of these sites, and also abolishes the increased level of GluR2 expression in RSK2 deficient cells. Thus, our results confirm the implication of ERK1/2 in Sp1 phosphorylation at both T453 and T739 Sp1 sites, and indicate that the increased levels of Sp1 phosphorylation and GluR2 expression in RSK2 deficient cells are related to the elevated ERK1/2 activity. Finally, the evidence that a significant reduction of Sp1 expression, with a Sp1-specific siRNA, prevents the increased level of GluR2 expression in RSK2 deficient cells further supports the critical role of Sp1 in the up regulation of GluR2 expression. How phosphorylated Sp1 activates the Gria2 promoter remains now to be investigated. It is possible that Sp1 phosphorylation may change its interaction with other transcription factors. Sp1 phosphorylation induced by the ERK pathway was also shown to cause the release of a histone deacetylase co-repressor complex from a gene promoter [30].
Previously, it was shown that changes in the expression levels of RSK2 can modify both the expression levels and the subcellular distribution of GluR2 subunits [8]. Our present data, by showing that RSK2 can modulate indirectly GluR2 expression by increasing or decreasing ERK1/2 activity, provides a potential mechanism for these modifications. RSK2 was also shown previously to directly interact with GluR2 [31]. Together the data provide increasing evidence that RSK2 plays an important role in the regulation of the function of AMPA receptors.
Gria1 (encoding GluR1) and Gria2 promoters share common features, such as multiple transcription start sites, GC rich, lack of a TATA box, and presence of Sp1 binding site in vitro [11,32]. However, RSK2 deficiency causes only an enhancement of GluR2 but not GluR1 or GluR3 protein expression levels both in hippocampal neurons [8] and PC12 cells (present study), which further underscores the specific role of the GluR2 AMPA receptor subunit. In the present study, the expression of GluR1 is even slightly down regulated in RSK2 depleted PC12 cells. Our data provide evidence that the regulation of Gria1 and Gria3 genes expression might be under the control of distinct signals that remain to be identified.
The vast majority of male patients affected with Coffin-Lowry syndrome have severe-to-profound intellectual disability, and currently there is no cure for this disorder. Our finding of an abnormally increased level of GluR2 expression in RSK2 deficient neurons could lead to therapeutic strategies for CLS. Indeed, in the Fragile X syndrome (FXS), observations suggesting increased metabotropic glutamate receptor (mGluR5) expression led to preclinical studies showing that the inhibition of mGluR5 can ameliorate multiple mutant phenotypes in mouse and drosophila models of FXS, and clinical trials based on this therapeutic strategy are underway [33]. Modulation of GluR2 levels may be a useful pharmaceutical approach for improving cognitive function. However, it is first necessary to validate the mechanism described in the present report in Rsk2-KO mice, and the present results in PC12 cells will facilitate this validation. This will need the use of lentiviral mediated gene and ShRNA transfer, and preliminary experiments have been undertaken. It is also essential to further dissect the molecular mechanism of Gria2 transcriptional control, including the identification by which Sp1 activates the Gria2 promoter. In addition to Sp1, Gria2 expression is also influenced at the transcriptional level by another positive (Nrf-1) and a negative (Re1/Nrse-like silencer) regulatory element in the 5' proximal region of the promoter [11]. Myers et al. [32] showed that the Gria2 Re1/Nrse-like silencer binds the RE1-silencing transcription factor (NRSF/REST), and that co-transfection of REST into neurons reduced Gria2 promoter activity in a silencer-dependent manner. The nuclear respiratory factor 1 (NRF-1) is a transcription factor that was also shown to regulate Gria2 gene transcription in neuroblastoma cells [34]. It is clearly necessary to understand the roles of the different regulatory elements and their interplay in neurons. In addition, although there is evidence that rapid (AMPAR) excitatory synaptic transmission is affected in Rsk2 mutant mice [8], the extent of glutamate transmission alteration, and whether synaptic plasticity is modified, remain to be explored. These studies will provide more insights into the feasibility of targeting Gria2 transcription in Coffin-Lowry syndrome treatment.

Transfection of PC12 Cells with shRNA Vectors
We first tested two mouse shRNAs (SureSilencing, Qiagen S.A. Courtaboeuf, France) cells. We then tested a shRNA directed against the human Rsk2 gene (according to the nomenclature: RPS6KA3 gene) (SureSilencing TM shRNA Plasmids for Human RPS6KA3, Clone ID7) and matching completely with the rat Rsk2 gene, which decreased Rsk2 expression dramatically. Thus, the shRNA clone ID7 was used in all subsequent studies (and named Rsk2-shRNA). Approximately 10 3 to 10 4 undifferentiated cells were plated in each of six well plates. After differentiation the cells were used for transfection. Briefly, 5 µg of Rsk2-or scrambled-shRNA, diluted in Opti-MEM ® I, were added to each well and transfection was performed with Lipofectamine 2000 according to the manufacturer's instructions. Cells were incubated for 6 h at 37 °C in a CO 2 incubator, then the medium was changed for

Immunofluorescence
Untransfected and Transfected PC12 cells grown on slides were washed three times with 1X Phosphate Buffered Saline (PBS), fixed for 10min with 4% PFA and 2.5 M glycine and washed again 3 times (1X PBS). Slides were then immersed in PBS-Triton 0.1% for 15 min and rinsed again 3 times with PBS. The cells were then incubated with the primary monoclonal antibody (Goat-anti-RSK2, 1:500, Santa Cruz Biotechnology) overnight, followed by incubation with a fluorescence-conjugated secondary antibody for 1 h, and finally dehydrated and mounted with KAISER's glycerol gelatin (Merck, Fontenay Sous Bois, France).

Statistical Analysis
All data are expressed as mean ± SEM and comparisons between groups were made by a Student's t-test. Differences with p < 0.05 were considered significant.

Conclusions
Our findings suggest an indirect role of RSK2 in modulating the expression of AMPA receptor GluR2 subunit and, therefore, in regulating glutamatergic neurotransmission and probably synaptic plasticity. Future studies are warranted to elucidate the molecular mechanisms that underline the interplay between the Rsk2 gene and the glutamatergic system.