Protein Arginine Methyltransferase 5 (PRMT5) Signaling Suppresses Protein Kinase Cδ- and p38δ-dependent Signaling and Keratinocyte Differentiation*

Background: MAPK signaling is an important mechanism controlling keratinocyte differentiation. Results: PRMT5 and p38δ interact as part of a multiprotein signaling complex, and PRMT5 and p38δ produce opposing actions in regulating differentiation. Conclusion: PRMT5 modulates p38δ MAPK kinase phosphorylation and signaling. Significance: This is a novel mechanism that links p38δ MAPK signaling and PRMT5 signaling. PKCδ is a key regulator of keratinocyte differentiation that activates p38δ phosphorylation leading to increased differentiation as measured by an increased expression of the structural protein involucrin. Our previous studies suggest that p38δ exists in association with protein partners. A major goal is to identify these partners and understand their role in regulating keratinocyte differentiation. In this study we use affinity purification and mass spectrometry to identify protein arginine methyltransferase 5 (PRMT5) as part of the p38δ signaling complex. PRMT5 is an arginine methyltransferase that symmetrically dimethylates arginine residues on target proteins to alter target protein function. We show that PRMT5 knockdown is associated with increased p38δ phosphorylation, suggesting that PRMT5 impacts the p38δ signaling complex. At a functional level we show that PRMT5 inhibits the PKCδ- or 12-O-tetradecanoylphorbol-13-acetate-dependent increase in human involucrin expression, and PRMT5 dimethylates proteins in the p38δ complex. Moreover, PKCδ expression reduces the PRMT5 level, suggesting that PKCδ activates differentiation in part by reducing PRMT5 level. These studies indicate antagonism between the PKCδ and PRMT5 signaling in control of keratinocyte differentiation.

Mitogen-activated protein kinases (MAPK) are dual-specificity serine/threonine kinases that drive intracellular signal transduction (1). MAPK family kinases share sequence similarity and conserved structural domains and include the extracellular signal-regulated kinases (ERK), Jun N-terminal kinases (JNK), and p38 MAPK. ERK is activated by mitogens and growth factors (2), whereas JNK and p38 kinases are typically activated in response to cellular stress (3). MAPKs play a central role in control of keratinocyte cell fate, and the balance between ERK and p38␦ activity is a key determiner of keratinocyte survival status. Enhanced ERK activity is associated with survival, whereas enhanced p38␦ activity is associated with differentiation and apoptosis (4 -10). Three p38 isoforms, p38␣, p38␤, and p38␦, are expressed in keratinocytes (5,8). Among these, p38␦ has a key role as a positive regulator of keratinocyte differentiation (7). p38␦ mediates the response to differentiating agents including phorbol ester, calcium, okadaic acid, and green tea polyphenol (7,10 -15).
It is clear that MAPK signaling is regulated by cross-talk from other signaling cascades; however, this regulation is not well understood. This is because the key cascades that cross-talk with MAPK signaling are not well defined, are likely cell typeand context-specific, and the impact of this cross-talk on biological outcome is not well understood. In this study we demonstrate a role for protein arginine methyltransferase five (PRMT5) 2 in modulating MAPK signaling in keratinocytes. PRMT5 is an enzyme that dimethylates protein-bound arginine residues (16). Protein methylation is receiving increasing attention as an important post-translational modification. Protein arginine methyl transferases (PRMTs) are evolutionarily conserved enzymes that catalyze transfer of methyl groups from S-adenosyl methionine to the guanidino nitrogen of proteinbound arginine. Eight functional PRMT proteins are encoded in the mammalian genome (17). These enzymes mono-and dimethylate arginine residues in proteins and are classified as type I (PRMT1, 2, 3, 4, 6, and 8) and type II (PRMT5 and 7) enzymes. Type I enzymes catalyze formation of asymmetrically dimethylated arginine (16,18).
PRMT5 is the sole type II member of the PRMT family that catalyzes formation of symmetrically dimethylated arginine (SDMA) (16). PRMT5 was discovered by yeast two-hybrid screening as Janus kinase-interacting protein 1 (16). PRMT5 dimethylates a variety of histone and non-histone proteins. Histone targets include histones H3 and H4 (19,20), whereas non-* This work was supported, in whole or in part, by National Institutes of Health Grant R01 AR046494 (to R. L. E.). 1  histone targets include small heterodimer partner (21), myelin basic protein (22), and a host of others. PRMT5 interacts in a number of protein complexes that regulate RNA processing, signal transduction, and transcription (19,(23)(24)(25)(26)(27)(28)(29). PRMT5 is a critical determinant of circadian period in Arabidopsis (30), and as a component of the androgen receptor cofactor complex, PRMT5 positively modulates androgen receptor-driven transcription independent of its methyltransferase activity (31,32). PRMT5 modulates enhanced GFR-mediated ERK activation (33) and is required for p53 expression and induction of p53 targets (34). PRMT5 also binds to death receptor 4 (35). An important study shows that PRMT1 modulates p38 MAPK regulation of differentiation in megakaryocytes (36). In addition to these functions in signal transduction, PRMT5 also participates in the assembly of the transcriptional repressor complex on various eukaryotic promoters (37). Thus, PRMT5 and protein arginine dimethylation are emerging as important regulators of cell function. Involucrin is a keratinocyte structural protein that functions as a precursor of the cornified envelope and is expressed in the suprabasal layers of epidermis (38,39). Regulation of involucrin gene expression has been extensively studied as a model for understanding regulation of differentiation-dependent gene expression in epidermis (5). A PKC, Ras, MEKK1, MEK3/ MEK6 signaling cascade has been implicated as a key control pathway in regulating involucrin expression (5). In this study we use this system to study cross-talk between PRMT5 and MAPK signaling in regulating keratinocyte differentiation. We show that PRMT5 reduces involucrin expression in normal human keratinocytes (KERn). These studies further show that PRMT5 is part of a p38␦-ERK signaling complex and that PRMT5 modification of proteins in this complex is associated with reduced p38␦ phosphorylation. The net impact is that PRMT5 antagonizes p38␦-dependent keratinocyte differentiation.
Primary Keratinocyte Culture and Adenovirus Production-Newborn foreskin epidermis was separated from dermis with dispase, and the keratinocytes were dispersed with trypsin and cultured in KSFM supplemented with epidermal growth factor and pituitary extract (44,45). Adenoviruses encoding wild-type FLAG-p38␦, PKC␦, and empty control adenovirus (Ad5-FLAG-p38␦, Ad5-PKC␦, Ad5-EV) were prepared as previously described by propagation in 293 cells and purification by cesium chloride centrifugation (9,46). For infection, keratinocytes were treated with multiplicity of infection 10 of adenovirus in the presence of 2.5 g/ml Polybrene for 4 h before the addition of fresh KSFM.
Affinity Purification of FLAG-p38␦ and Mass Spectrometry-Keratinocytes were infected with Ad5-FLAG-p38␦ and after 24 h extracts were prepared in lysate buffer (20 mM Tris-HCl, pH 7.4, containing 150 mM NaCl, 1 mM Na 2 EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM ␤-mercaptoethanol, 1 mM Na 3 VO 4 , 1 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, and proteinase inhibitor mixture). Anti-FLAG M2 affinity gel was washed with a large volume of lysis buffer followed by two bed volumes of 20 mM glycine-HCl, pH 2.8, and then re-equilibrated with lysis buffer. Cell lysate (120 mg of total protein) was incubated with M2 gel for 12 h at 4°C with shaking and then packed in to PD10 column and washed with lysis buffer until A 280 Ͻ 0.02 to remove unbound protein. FLAG-p38␦-associated proteins were eluted with 20 mM glycine-HCl, pH 2.8, and the collected fractions (200 l) were neutralized by the addition of 1 M Tris before storage at Ϫ20 C.
Affinity-purified FLAG-p38␦-associated proteins, prepared as outlined above, were concentrated by trichloroacetic acid precipitation, and samples were separated by SDS-PAGE followed by Coomassie Blue staining. Gel bands were excised and digested with trypsin, and the tryptic peptides were analyzed by liquid chromatography-tandem mass spectroscopy using a Bruker Omniflex benchtop MALDI-TOF MS/MS. For protein identification, the monoisotopic mass maps of tryptic peptides were searched in databases using the MASCOT search engine.
Immunoprecipitation and Immunoblot-Total extracts in lysis buffer were incubated with appropriate primary antibody and protein A/G-agarose overnight at 4 ºC and washed 4 times with Tris-buffered saline, pH 7.4. After the final wash the beads were collected, boiled with sample buffer, and centrifuged, and supernatant was electrophoresed and transferred to nitrocellulose membrane. The blots were blocked with 5% skimmed milk in Tris-buffered saline containing 0.1% Tween 20 and incubated with appropriate primary and horse-radish peroxidase-conjugated secondary antibody, and antibody binding was visualized by chemiluminescent detection (Amersham Biosciences).
Keratinocyte Electroporation-Keratinocytes were electroporated with plasmids using an Amaxa electroporator and the VPD-1002 nucleofection kit (Amaxa, Cologne, Germany). Keratinocytes were harvested with trypsin and replated 1 day before use. On the day of electroporation, 1.2 million replated cells were harvested with trypsin and resuspended in KSFM. The cells were collected at 2000 ϫ g, washed with 1 ml of sterile phosphate-buffered saline, and suspended in 100 l of keratinocyte nucleofection solution. The cell suspension, which included 3 g of plasmid or siRNA, was mixed by gentle up and down pipetting and electroporated using the T-018 program. KSFM (500 l) was added, and the suspension was transferred to a 60-mm cell culture dish.

RESULTS
Identification of FLAG-p38␦-associated Proteins-Our previous studies describe a p38␦-ERK signaling complex in keratinocytes that regulates keratinocyte differentiation (6). Although these studies show that p38␦ interacts with ERK1/2, it is likely that p38␦ also interacts with other proteins and that these proteins influence function. To identify additional interaction partners, extracts were prepared from FLAG-p38␦-expressing keratinocytes, and p38␦-associated proteins were purified by FLAG affinity column chromatography. The column was loaded with 120 mg of total cell lysate and washed with loading buffer, and affinity-bound proteins were eluted with glycine buffer. Fig. 1A shows a Coomassie-stained gel illustrating the affinity purification of FLAG-p38␦. The fractions show the lysate (L) that was loaded onto the column, the column wash (W), and column elution fractions (F1, etc.). The Coo-massie-stained gel shows the presence of FLAG-p38␦ in elution fractions F2, F3, and F4 but most prominently in F3 (Fig. 1A). Anti-FLAG immunoblot of proteins derived from the total lysate and elution fractions reveals FLAG-p38␦ in elution fractions F2, F3, and F4 (Fig. 1B). These fractions were pooled and analyzed by matrix-assisted laser desorption ionization mass spectrometry, which revealed the presence of several proteins. Among these was PRMT5. Fig. 1C indicates that PRMT5-derived tryptic peptides identified by mass spectrometry. PRMT5 is a protein arginine methyltransferase that modifies arginine in target proteins to form SDMA (16). This covalent post-translational modification acts to alter target protein function (16). FIGURE 1. FLAG-p38␦ complex includes PRMT5. A, KERn were infected with multiplicity of infection 10 of Ad5-FLAG-p38␦, and after 24 h extracts were prepared for FLAG affinity chromatography. Total cell lysate (L), column wash (W), and fractions eluted from the FLAG affinity column (F1, etc.) were electrophoresed, and the gel was Coomassie-stained. Molecular mass markers are indicated. B, shown is detection of FLAG-p38␦. Samples were electrophoresed and immunoblotted with anti-FLAG. FLAG-p38␦ was absent from the wash fraction. F3 contains the highest concentration of FLAG-p38␦. C, amino acid sequence of PRMT5 tryptic fragments were identified by mass spectrum analysis. D, co-precipitation of p38␦ and PRMT5 is shown. Total cell extracts were prepared from KERn and then immunoprecipitated (IB) with anti-IgG (control), anti-p38␦, or anti-PRMT5. The precipitates were then electrophoresed for immunoblot detection of p38␦ and PRMT5. Similar results were observed in each of three experiments.
These findings suggest that PRMT5 is a binding partner for p38␦. To confirm this, we prepared total cell extracts from human keratinocytes and immunoprecipitated endogenous p38␦ with anti-p38␦. The precipitate was electrophoresed, and the gel was stained with anti-PRMT5. Fig. 1D shows that PRMT5 co-precipitates with p38␦ whether the precipitation is performed with anti-p38␦ or anti-PRMT5, thereby providing additional evidence for p38␦/PRMT5 intracellular interaction.
Role of PRMT5 in Regulating Keratinocyte Differentiation-Our previous studies demonstrate that a PKC␦ signaling cascade increases p38␦ activity, which leads to increased cell differentiation (5, 6, 10, 47, 48). Because PRMT5 interacts with p38␦, we hypothesized that it may regulate differentiation-associated responses that depend upon p38␦ activity. To assess this we monitored the impact of PRMT5 on involucrin (hINV) expression. hINV is a marker protein that is increased during keratinocyte differentiation and in response to differentiation stimuli (5,6,10,47,48). We began by assessing the impact of PRMT5 knockdown and overexpression on hINV mRNA and protein levels. Fig. 2A shows that treatment with PRMT5pooled (Santa Cruz, sc-41073) or individual (PRMT5-1 or PRMT5-2) siRNA (35) reduces PRMT5 levels, and this is associated with a marked increase in hINV protein levels. Fig.  2B shows this is due to changes in mRNA levels. Treatment with PRMT5-pooled (P), PRMT5-1 (1) or PRMT5-2 (2) siRNA reduces PRMT5 mRNA, and this is associated with increased hINV mRNA. Consistent with these findings, the converse is also observed, as overexpression of PRMT5 reduces hINV mRNA and protein level (Fig. 2, D and E).
PRMT5 is a protein arginine methyltransferase that catalyzes modification of arginine residues in target proteins to form SDMA, a modification that modulates target protein function (16,18). We, therefore, examined the impact of PRMT5 knockdown and overexpression on SDMA formation. Extracts from PRMT5-siRNA-treated and PRMT5-overexpressing cells were prepared for anti-SDMA immunoblot. Proteins ranging in molecular mass from 50 to 300 kDa are detected (Fig. 2, C and  E). Immunostaining of a limited number of proteins covering a broad range of molecular weights is typical of anti-SDMA reagents (49). Using this as a crude assay of PRMT5 activity, we observe a 10 -15% reduction in average SDMA intensity in PRMT5-siRNA-treated cells and a similar 10 -15% increase in extracts derived from PRMT5-overexpressing cells. We do not presently know which proteins are identified in this assay, but these findings suggest that increased PRMT5-dependent SDMA formation is associated with suppression of involucrin gene expression.
Regulation of hINV Expression by PKC␦ and PRMT5-The above studies indicate that PRMT5 suppresses hINV mRNA and protein levels. We next compared the action of PRMT5 against two treatments that increase differentiation, PKC␦ overexpression and TPA treatment. PKC␦ promotes keratinocyte differentiation and increases involucrin gene expression via activation of p38␦ (4,5,47). pINV-2473 encodes the fulllength hINV promoter linked to a luciferase reporter gene such that increased promoter activity is reflected in increased luciferase activity (44). Keratinocytes were transfected with pINV-2473 in the presence of control-siRNA or PRMT5-siRNA and empty or PKC␦-encoding expression vector. In the presence of control-siRNA, PKC␦ increases hINV promoter activity 3-fold (Fig. 3A). In contrast, when PMRT5-siRNA is present and PRMT5 is reduced (Fig. 3C), the induction is 6.5-fold (Fig. 3A). Moreover, the PKC␦ induction of hINV promoter activity is suppressed (Fig. 3B) in PRMT5 overexpressing (Fig. 3C) cells.
TPA is a pharmacological agent that increases PKC activity and stimulates differentiation (44). We used TPA treatment as a second method of PKC␦ activation. Keratinocytes were transfected with pINV-2473 and treated with TPA in the presence of control-or PRMT5-siRNA. TPA at 5 and 10 ng/ml increases hINV promoter activity, and a further 2-fold increase in promoter activity is observed in PRMT5 knockdown cells (Fig. 4A). In the converse experiment in which PRMT5 is overexpressed, the TPA-dependent increase in hINV promoter activity is suppressed by 2-fold (Fig. 4B). These studies suggest that PRMT5 opposes the pro-differentiation action of TPA and PKC␦.
It is well established that PKC␦ enhances keratinocyte differentiation (4 -6, 47). We wondered whether part of this mechanism may involve PKC␦ suppression of PRMT5 expression. To test this we manipulated PKC␦ level and monitored the impact on PRMT5 expression. Fig. 5A shows that PRMT5 RNA level is increased 4-fold when PKC␦ mRNA levels is reduced by treating KERn with PKC␦-pooled (Santa Cruz) or individual (PKC␦-1 or PKC␦-2) (40) siRNA. Moreover, the reduction in PKC␦ level is associated with increased PRMT5 protein expression and SDMA formation (Fig. 5B). In the inverse experiment, PKC␦ overexpression reduces PRMT5 RNA and protein level (Fig. 5, C and D), and this is associated with reduced SDMA formation (Fig. 5D). As noted in Fig. 2, the anti-SDMA blots shown in Fig. 5, B and D, provide a generally index of PRMT5 activity (49).

PRMT5-dependent Covalent Modification of Proteins in p38␦ Complex Is Associated with Increased p38␦ Phosphorylation-
Given that PRMT5 knockdown enhances total p38 phosphorylation, a key question is whether PRMT5 specifically modulates phosphorylation of the p38␦ isoform. This could be expected, as previous studies indicate that p38␦ is the key p38 isoform involved in regulation of differentiation (6,7,9). A second key issue is whether any such ability of PRMT5 to regulate p38␦ phosphorylation is associated with altered SDMA modification of proteins in the p38␦ signaling complex. To assess this, we performed two types of experiments. We first treated cells with control or PRMT5-pooled PRMT5-1 or PRMT5-2 siRNA for 48 h and then immunoprecipitated p38␦ and assessed phosphorylation status. In Fig. 7A shows that p38␦ is phosphorylated in PRMT5 knockdown but not control cells. As a second method to confirm this result, we monitored the phosphorylation state of vector-expressed FLAG-p38␦. Cells were treated with control-or PRMT5-siRNA and then infected with Ad5-EV or Ad5-FLAG-p38␦. At 24 h post-infection, FLAG-p38␦ was immunoprecipitated with anti-FLAG, and the p38␦ phosphorylation state was measured. Fig. 7B confirms that PRMT5 knockdown is associated with increased p38␦ phosphorylation.
We next monitored whether the p38␦ complex contains SDMA modified proteins. We infected cells with Ad5-EV or Ad5-FLAG-p38␦ and after 24 h immunoprecipitated with anti-FLAG and monitored the level of SDMA in the precipitate. Fig. 7C shows that SDMA modified proteins are present in the p38␦ complex. We next assessed whether this SDMA modification requires PRMT5. We treated cells with control-or PRMT5-siRNA and after 48 h prepared extracts for immunoprecipitation with anti-p38␦. Fig. 7D shows that a SDMA-modified 60-kDa protein is associated with the p38␦ complex. Moreover, PRMT5 knockdown reduced the level of this protein or the extent of its SDMA modification. This protein has a molecular mass of 60 kDa, suggesting that it is a protein other than p38␦. As a control for assay integrity, we confirm appropriate immunoprecipitation of p38␦ and PRMT5. Taken together, these findings suggest that PRMT5 suppresses p38␦ phosphorylation and that this is associated with increased SDMA modification of a 60-kDa protein present in the p38␦ complex.

DISCUSSION
Epidermal Differentiation-The epidermal keratinocyte, the major cell type present in the epidermis, undergoes a complex and choreographed program of differentiation. This process requires a balance among keratinocyte proliferation, differentiation, and apoptosis. Ultimately, this process leads to the formation of a multilayered epidermis that contains a proliferative basal zone beneath several layers of cells that are at various stages in the differentiation process (50). Involucrin is a marker of this process that is increased in differentiated cells (5) and is an important component of the scaffolding of the cornified envelope (43,(51)(52)(53). Understanding how this process is regulated to produce a multilayered tissue is a major goal in epidermal biology, as disorders in this process are observed in a host of epidermal disease states and also in cancer.
p38␦-ERK Signaling Complex-Previous studies show that a p38␦-ERK complex regulates expression of key apoptosis, differentiation, and proliferation-associated genes in keratinocytes (4 -6, 10, 11). p38␦ activity is increased, and ERK1/2 activity is reduced in differentiated keratinocytes (6). The importance of interaction within this complex is illustrated by the role of p38␦ and ERK1/2. p38␦ and ERK1/2 are associated either directly or indirectly in this complex, and increased p38␦ phosphorylation and reduced ERK1/2 activity is observed in FIGURE 4. Impact of PRMT5 on TPA-dependent hINV promoter activity. A, KERn were electroporated with 3 g of control-or PRMT5-pooled (Santa Cruz) siRNA. After 48 h cells were re-electroporated with 3 g of pINV-2473 and involucrin promoter luciferase reporter construct. After an additional 6 h, TPA was added, and after an additional 18 h cells were lysed for luciferase activity assay. The values are the mean Ϯ S.E., n ϭ 3. The asterisks indicate a significant difference as determined by Student's t test, p Ͻ 0.001, between control-siRNA and PRMT5-siRNA at 5 and 10 ng/ml TPA. B, KERn were electroporated with 3 g of pcDNA3 or pcDNA-PRMT5 and after 48 h re-electroporated with 3 g of pINV-2473. After an additional 6 h the cells were treated with TPA for 18 h, and then extracts were prepared for luciferase activity assay. The results are presented as the mean Ϯ S.E., n ϭ 3. The asterisks indicate a significant difference as determined by Student's t test, p Ͻ 0.001, between pcDNA3 and pcDNA3-PRMT5 at 5 and 10 ng/ml TPA. response to treatment with differentiation stimuli (4 -7). We propose that this complex serves to integrate incoming signals and that the net output then determines cell fate. For example, activation of PKC␦ increases p38␦ activity and cell differentiation and death, whereas activation of ERK1/2 by delivery of constitutively active Raf antagonizes this response and promotes cell survival (5). The fact that a p38␦-ERK complex is involved in this regulation suggests that other proteins may also be partners in this regulatory complex.
An important goal is identification of additional partners that are part of this complex and influence the signaling activity. To achieve this we expressed FLAG-p38␦ in keratinocytes and identified proteins associated with p38␦ by FLAG-affinity chromatography and mass spectrometry. One of the proteins that co-purified with p38␦ was PRMT5. This interaction was confirmed by co-immunoprecipitation of PRMT5 and p38␦ from keratinocyte extracts using a p38␦-specific antibody. This confirmation is important, as it has been reported in one study that PRMT5 can interact with anti-FLAG (16). Protein arginine methyltransferases are a family of proteins that dimethylate arginine residues in target proteins (16). Arginine modification in this context influences target protein structure, function, and activity (16). PRMTs have a role in numerous cellular processes including regulation of cell signaling and gene expression (17,   32, 33, 54 -57). However, additional studies are needed to explore the physiological mechanisms whereby PRMTs regulate cells survival, differentiation, and apoptosis. Despite the recognized importance of PRMT-dependent modification in a host of cell types, the role of PRMT in regulating keratinocyte function has not been studied.
PRMT5 Regulates Basal Involucrin Expression-Our previous studies show that keratinocyte differentiation requires activity in a PKC␦, Ras, MEKK1, MEK3 cascade that triggers changes in activity in a p38␦-ERK complex to increase p38␦ activity relative to ERK activity (4 -7, 9 -11, 15, 47, 48, 58). Activation of this cascade is associated with cessation of cell proliferation and increased morphological differentiation (8,10,15), and prolonged stimulation can cause apoptosis (4,8). Stimulation of MAPK activity by this cascade is not associated with changes in expression of p38␦ or ERK1/2 level but is associated with increased p38␦ activity and reduced ERK1/2 activity (6). Involucrin is a marker of differentiation that is increased in differentiated cells and is expressed at a fixed basal level in rest-ing keratinocytes (5). An interesting and unexpected finding is that PRMT5 knockdown increases basal involucrin expression including increasing hINV mRNA and protein levels and promoter activity. Moreover, the converse is also observed in that PRMT5 overexpression reduces basal involucrin mRNA and protein levels.
PRMT5 Influences Stimulus-dependent Involucrin Expression-We also monitored the impact of PRMT5 on response to agents that activate p38␦ signaling to increase differentiation. Overexpression of PKC␦ and treatment with TPA enhance p38␦ phosphorylation and increase involucrin expression (4,6). TPA is a diacylglycerol analog that enhances PKC␦ activity and induces differentiation (44,48). We show that PRMT5 knockdown enhances the differentiation promoting ability of these agents and the PRMT5 overexpression suppresses the differentiation promoting response. These findings suggest that PRMT5 is a negative regulator of keratinocyte differentiation. Thus, these studies identify an important new PRMT5-mediated regulatory circuit in keratinocytes that functions to sup-FIGURE 7. PRMT5 regulates p38␦ phosphorylation and SDMA modification of p38␦-associated proteins. A, KERn were electroporated with 3 g of control, PRMT5-pooled, PRMT5-1, or PRMT5-2 siRNA, and after 48 h extracts were prepared for immunoprecipitation (IP) with anti-p38␦. The precipitated material was then monitored by immunoblot to detect p38␦ and phosphorylated p38␦. B, KERns were electroporated with 3 g of control-or PRMT5-pooled (Santa Cruz) siRNA and after 24 h infected with multiplicity of infection 10 of Ad5-EV or Ad5-p38␦ and after an additional 24 h extracts were prepared for immunoprecipitation with anti-FLAG. The precipitated proteins were electrophoresed, and the level of FLAG-p38␦ (anti-FLAG) and phosphorylated p38␦ was monitored by immunoblot. C, KERn (70% confluent) was infected with multiplicity of infection 10 Ad5-EV or Ad5-FLAG-p38␦, and after 24 h lysate was prepared, and 200 g of protein was immunoprecipitated with anti-FLAG antibody. The resulting pellet was electrophoresed for immunoblot (IB) with anti-SDMA. 20 g of total lysate (L) was electrophoresed in a parallel lane. The blot was then incubated with anti-SDMA. D, KERn were electroporated with 3 g of control-or PRMT5-pooled (Santa Cruz) siRNA, and after 48 h extracts were prepared for immunoprecipitation with anti-IgG or anti-p38␦. The precipitates were then electrophoresed for immunodetection of p38␦, PRMT5, and SDMA. The arrows indicate the major SDMA modified bands. E, PKC␦ and PRMT5 regulation of p38␦ activity; a working model. The model is explained in the "Discussion." We realize that multiple complexes may exist and that some may lack one or more of the components, and that it is likely that other proteins are also present in this complex. We also note that we do not know whether there is direct interaction between the indicated components. The arrows indicate relative change in level. The dashed arrow indicates PRMT5 SDMA modification of target proteins. press differentiation. In addition to an important role in normal cells, these findings suggest a role for PRMT5 in skin cancer. PRMT5 levels are increased (29), and PKC␦ levels are decreased in cancer (59,60). Thus, the balance between these signaling inputs is likely to be important in disease development, and we anticipate that PRMT5 may serve to enhance skin cancer cell survival. The possibility that PRMT5-related regulation is increased is an important area for further investigation in skin cancer. PRMT5 may also link the p38␦ complex to other signaling cascades, as PRMT5 associates with Janus kinases and STAT (54,56). Thus, the knowledge that PRMT5 is part of the p38␦ complex helps us in our effort to assembly the signaling network that controls keratinocyte differentiation.
Suppression of PRMT5 Level by PKC␦-There are numerous situations where increased pro-differentiation signaling is associated with reduced pro-survival signaling and vice versa. An example in keratinocytes is the coupling of increased p38␦ and decreased ERK1/2 activity in differentiation agent-stimulated cells (5). PKC␦ and PRMT5 appear to share such a relationship, as PKC␦ overexpression is associated with reduced PRMT5, and PKC␦ knockdown is associated with increased PRMT5. Thus, it is possible that suppression of the PRMT5 level and activity may be a mechanism whereby PKC␦ drives keratinocyte differentiation. We know that PKC␦-dependent reduction in PRMT5 is associated with reduced PRMT5 mRNA and protein levels. This suggests a potential impact of PKC␦-associated signaling on transcription of the PRMT5 gene or an impact on PRMT5 RNA stability; however, further studies will be necessary to fully understand this regulation.
Impact of PRMT5 on p38␦ Signaling Complex-To gain insight regarding the role of PRMT5 in p38␦ MAPK signaling, we examined the impact of PRMT5 on the complex. These studies are extremely interesting, as they suggest that PRMT5 influences the p38␦ phosphorylation state. First, we show that PRMT5 is a partner in the p38␦ complex. PRMT5 is found associated with FLAG-p38␦ when analyzed by mass spectrometry and also co-immunoprecipitates with p38␦ present in total cell extracts. That this interaction is real is supported by functional studies. At a functional level, PRMT5 influences molecular events in this complex. Overexpression of PRMT5 results in reduced p38␦ phosphorylation of p38␦, and this is associated with increased SDMA modification of a 60-kDa protein that also co-precipitates with p38␦. Moreover, this regulation is reversed in PRMT5 knockdown cells.
The identity of the 60-kDa protein is not presently known, but it appears to be a bona fide target of PRMT5, as PRMT5 knockdown results in reduced anti-SDMA detection of this band. This is likely due to decreased SDMA labeling of this protein but could also be due to a reduction in protein level. It is interesting that this 60-kDa band is also a major SDMA-modified protein in total keratinocyte cell extracts, suggesting that it may be a major PRMT5 target that is involved in a range of processes. It should also be noted that we do not intend to imply that all p38␦ complexes included all of these partners (ERK1/2, PRMT5, 60-kDa protein, etc.), as it is likely that multiple subcomplexes of varying composition exist and that these complexes may have unique intracellular functions. It is also worth noting that no previous reports described an impact of PRMT5 on MAPK signaling. The only available report relates an impact of PRMT1, a type I PRMT that asymmetrically dimethylates arginine (16,18), on p38␣ function in megakaryocyte differentiation (36).
Based on the studies presented in this proposal, we propose that PRMT5 is a new component of this signaling complex. The role of PRMT5 in this complex is to inhibit p38␦ activity and suppress differentiation by catalyzing SDMA modification at the arginine residue(s) of a 60-kDa protein (and perhaps other proteins) that are present in this complex. We propose that SDMA modification of these targets alters the complex in a way that leads to reduced p38␦ phosphorylation. Thus, when PRMT5 levels are high and PRMT5 is active, differentiation-dependent gene (hINV) expression is suppressed. Under conditions where cells are exposed to a differentiation stimulus, PKC␦ activity triggers signaling events that drive differentiation. We propose that PKC␦ acts to increase p38␦ phosphorylation by two potential mechanisms (Fig. 7E). One mechanism is stimulating signaling events that lead to increased p38␦ phosphorylation (6). Our present studies suggest that a second mechanism is suppression of PRMT5 mRNA and protein levels, thereby reducing PRMT5 modification of proteins in the p38␦ complex. Our studies show that reduced PRMT5 levels are associated with increased p38␦ phosphorylation. Thus, the present studies identify a new signaling protein present in the p38␦ complex that suppresses differentiation-associated p38␦ activity.