Protein Arginine Methyltransferase 1 Interacts with and Activates p38α to Facilitate Erythroid Differentiation

Protein arginine methylation is emerging as a pivotal posttranslational modification involved in regulating various cellular processes; however, its role in erythropoiesis is still elusive. Erythropoiesis generates circulating red blood cells which are vital for body activity. Deficiency in erythroid differentiation causes anemia which compromises the quality of life. Despite extensive studies, the molecular events regulating erythropoiesis are not fully understood. This study showed that the increase in protein arginine methyltransferase 1 (PRMT1) levels, via transfection or protein transduction, significantly promoted erythroid differentiation in the bipotent human K562 cell line as well as in human primary hematopoietic progenitor CD34+ cells. PRMT1 expression enhanced the production of hemoglobin and the erythroid surface marker glycophorin A, and also up-regulated several key transcription factors, GATA1, NF-E2 and EKLF, which are critical for lineage-specific differentiation. The shRNA-mediated knockdown of PRMT1 suppressed erythroid differentiation. The methyltransferase activity-deficient PRMT1G80R mutant failed to stimulate differentiation, indicating the requirement of arginine methylation of target proteins. Our results further showed that a specific isoform of p38 MAPK, p38α, promoted erythroid differentiation, whereas p38β did not play a role. The stimulation of erythroid differentiation by PRMT1 was diminished in p38α- but not p38β-knockdown cells. PRMT1 appeared to act upstream of p38α, since expression of p38α still promoted erythroid differentiation in PRMT1-knockdown cells, and expression of PRMT1 enhanced the activation of p38 MAPK. Importantly, we showed for the first time that PRMT1 was associated with p38α in cells by co-immunoprecipitation and that PRMT1 directly methylated p38α in in vitro methylation assays. Taken together, our findings unveil a link between PRMT1 and p38α in regulating the erythroid differentiation program and provide evidence suggesting a novel regulatory mechanism for p38α through arginine methylation.


Introduction
Circulating red blood cells are derived from hematopoietic progenitors through an intricate process of erythropoiesis which is essential for maintaining a vital physiological condition. In addition, stimulating erythropoiesis has a wide application in treating anemia caused by chemotherapy, chronic renal failure and hematological diseases [1]. Despite extensive studies, our understanding of regulatory events involved in erythroid differentiation remains incomplete.
A variety of cytokines acts in concert to regulate erythroid differentiation [2]. Among these, erythropoietin (EPO) is the primary cytokine regulating various stages of erythropoiesis in our body. Intracellular signaling pathways such as the JAK/STAT [3], mitogen-activated protein kinase (MAPK) [4] and PI3 kinase/Akt [5] cascades have been shown to mediate these extracellular signals. Activation of the p38 MAPK pathway contributes critically to the erythroid differentiation of leukemia cell lines induced by various agents [6,7] and of primary CD34 + hematopoietic progenitors induced by EPO [8]. There exist multiple isoforms of p38; however the distinctive role of each isoform in erythropoiesis is still elusive. Moreover, although some downstream effectors of p38 MAPK, such as activating transcription factor-2 (ATF-2) and cyclic AMP response element binding protein (CREB), have been reported to participate in the induction of the c-globin gene [9], the detailed upstream regulatory events and downstream effectors of the p38 MAPK pathway during erythroid differentiation are yet to be elucidated.
Posttranslational modifications expand the chemical diversity of side chains of amino acids and provide dynamic and reversible modulations of the activity of proteins to meet the need of cellular functions. Protein arginine methylation, which adds one or two methyl groups to the guanidine nitrogen of arginine residues, is emerging as a pivotal posttranslational modification involved in various cellular processes including transcription regulation, DNA repair, RNA processing and signal transduction [10,11]. Protein arginine methyltransferase 1 (PRMT1) is the first cloned [12] and the predominant PRMT in mammalian cells [13]. PRMT1 can regulate gene expression by methylating transcription factors and histones. The methylation of NIP45 by PRMT1 facilitates its interaction with NF-AT and thus stimulates cytokine gene expression [14]. The transcriptional activity of RUNX1 is potentiated by PRMT1 methylation, which abrogates its association with the repressor SIN3A [15]. PRMT1 can modify histone H4 on Arg3, which permits the subsequent acetylation of histone H3 and thus activates gene expression [16]. In addition, arginine methylation is reported to affect the enzymatic activity of the modified proteins. For example, PRMT1-mediated methylation of the NS3 protein of hepatitis C virus inhibits its helicase activity [17]. Evidence supporting the crosstalk between arginine methylation and phosphorylation is emerging. Arginine methylation of FOXO transcription factors and anti-apoptotic BAD proteins by PRMT1 inhibits their phosphorylation by Akt [18,19]. PRMT1 transiently methylates estrogen receptor a and triggers its interaction with PI3K and Src tyrosine kinase [20]. The methylation of Axin by PRMT1 enhances its phosphorylation by GSK3b (glycogen synthase kinase 3b), leading to an increased stability [21].
Given the broad range of cellular functions regulated by protein arginine methylation, this study explored the involvement of this posttranslational modification in modulating erythroid differentiation. Our results provide solid evidence for a novel role of PRMT1 in promoting erythroid differentiation and suggest arginine methylation as a novel mechanism for regulating p38a functions.

Erythroid Differentiation
Human chronic myelogenous leukemia K562 cells and human erythroleukamia HEL cells were obtained from BCRC (Bioresource Collection and Research Center, Taiwan). Both cell lines were cultivated in RPMI 1640 medium supplemented with 10% fetal bovine serum (Biological Industries). To induce erythroid differentiation, cells were treated with cytosine arabinoside (araC, 1 mM) or sodium butyrate (NaB, 2 mM for K562 and 0.1 mM for HEL). Hemoglobin production was detected using a benzidine/ hydrogen peroxide solution [6]. The dark violet particles of oxidized benzidine were readily distinguished under a light microscope. Two to three hundred cells were examined in each assay. Surface expression of glycophorin A, a characteristic of erythroid cells, was detected by flow cytometry.

Isolation of CD34 + Hematopoietic Progenitor Cells and Induction of Erythroid Differentiation
Isolation and expansion of CD34 + cells were performed as described previously [22]. To induce differentiation toward erythroid lineage, the expanded CD34 + cells were cultivated (5 6 10 4 cells/ml) in media without a cytokine cocktail for 6 hr prior to the addition of erythropoietin (EPO; 1 U/ml). Three days later, the cells were again treated with EPO. Erythroid differentiation was examined 7 days after the first EPO treatment by surface expression of glycophorin A and benzidine staining. The TATmediated protein transduction system was applied to examine the effect of PRMT1. The purified recombinant TAT-fused HA-PRMT1 and the control TAT-fused HA-GFP proteins were added to the CD34 + cells 6 hr before the first EPO treatment and again at the time of the EPO treatment.

Ethic Statement
The human umbilical cord blood was obtained from Taoyuan General Hospital, Taiwan, with an informed written consent from donors and the prior approval from the ethic committee of Institutional Review Board (IRB). The procedure was conducted following the approved guideline of IRB. The data obtained in this study were analyzed anonymously.

Flow Cytometric Analysis
Cells were washed, incubated in PBS containing 2% BSA for 30 min at 4uC and stained with anti-human FITC-conjugated glycophorin A antibodies (Abcam; 1 ml in 200 ml of 1% BSA-PBS) for 30 minutes. After washing, cells were suspended in 1 ml of 1% BSA-PBS and subjected to flow cytometric analysis using a FACScan flow cytometer (BD Biosciences) equipped with Cellquest software (BD Biosciences).

Examination of PRMT1 Activity in CD34 + Progenitor Cells
Human primary CD34 + hematopoietic progenitor cells were resuspended in extraction buffer (50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM EGTA, 40 mg/ml leupeptin, 40 mg/ml aprotinin, 20 mg/ml pepstatin, and 1 mM PMSF) and homogenized on ice with a glass tissue grinder. The homogenates were centrifuged at 16,0006g for 60 min at 4uC, and supernatants were collected. To examine endogenous PRMT1 activity, cell homogenates (2 mg) were incubated with purified recombinant hnRNP A1, A2 or K proteins (2 mg), which are known substrates for PRMT1 [23,24,25], in 25 mM Tris-HCl, pH 8.0, and S-adenosyl-L-[methyl-3 H] methionine ( 3 H-AdoMet) at 30uC for 30 min. Reactions were stopped by the addition of SDS sample buffer and then subjected to SDS-PAGE. Gels were stained, de-stained and then soaked in the fluorographic enhancer EN 3 HANCE (Perkin Elmer). After dried, the gels were exposed to X-ray films (GE) at 270uC. Methyl incorporation was analyzed by fluorography or by scintillation counting as described previously [23,24].

Real-time Reverse Transcription PCR
Total RNA was extracted using an illustra RNAspin Mini Isolation Kit (GE Healthcare). Equal amounts of total RNA from each sample were subjected to cDNA synthesis using reverse transcriptase (Fermentas Life Sciences). Real-time PCR was performed in an ABI StepOne System (Applied Biosystems) using the SYBR Green detection protocol as outlined by the manufacturer. Samples were examined in triplicate in each experiment and analyzed by using the 2 2DDCT method. Expression of GAPDH was used as an internal control for cDNA content. The DCT of a target gene after stimulation was compared to the DCT of the same target gene before stimulation. The results were presented as fold changes. Gene-specific amplification was conducted with the following primer sets: for delta-aminolevulinate synthase 2 (ALAS2), 59-GCAGCACTCAACAGCAAG-39 (sense) and 59-ACAGGACGGCGACAGAAA-39 (antisense); for c-globin, 59-CCATAAAGCACCTGGATGATC-39 (sense) and 59-ATCTG-GAGGACAGGGCACTG-39 (antisense); for porphobilinogen deaminase (PBGD), 59-CGCCTCCCTCTAGTCTCTGCTTCT-39 (sense) and 59-GTTGCCACCACACTGTCCGTCTG-39 (antisense); for erythroid Krüppel-like factor (EKLF), 59-CGGCAAGAGCTACAC-CAAG-39 (sense) and 59-CCGTGTGTTTCCGGTAGTG-39 (antisense); for nuclear factor-erythroid-derived 2 (NF-E2), 59-ACAGAGAGCCAGCTAGAGCT- 39 (sense) and 59-TGGGCTGCCACCTTGTTT-39 (antisense); for GATA1, 59-  (1 mM) or NaB (2 mM). Transient expression of PRMT1 significantly enhanced erythroid differentiation, as detected by benzidine staining. (B) R2-1 and R2-3 cells, which stably expressed HA-PRMT1, also exhibited higher erythroid differentiation. E3-6 cells were an empty vector control. (C, D) The expression of various erythroid-related genes and transcription factors was examined by real-time PCR after araC treatment. The b-actin gene was a control whose expression was not affected by araC treatment. Data were calculated as fold changes using vector control cells without treatment as a reference. (E) The enzymatically impaired G80R PRMT1 mutant, when introduced by transient transfection, failed to enhance differentiation. (F) The stable clone that expressed the G80R mutant also failed to stimulate ALAS2 gene expression. All results shown are representative of three separate experiments. Differentiation results are presented as means 6 S.E. of three repeats; *, p,0.05; **, p,0.01; ***, p,0.005 compared with vector control cells. doi:10.1371/journal.pone.0056715.g001

In vitro Protein Kinase Assay
The p38 kinase assay was performed using a nonradioactive p38 kinase assay kit (Cell Signaling). Briefly, active p38 MAPK was immunoprecipitated with an immobilized anti-phospho-p38 antibody and assayed for kinase activity with ATF2 peptides (1 mg) as the substrate. Reactions were stopped with SDS sample buffer and subjected to Western blotting with a specific anti-phospho-ATF2 antibody according to the manufacturer's instruction.

Methylation of p38a by PRMT1
Methylation of recombinant His-p38a by GST-PRMT1 was carried out by in vitro methylation assay using 3 H-AdoMet as a methyl donor and detected by fluorography.

Statistical Analysis
All results shown are representative of three separate experiments. Results are presented as means 6 S.E. Statistical comparison between two groups was analyzed using Student's ttest. Values of p,0.05 were considered significant. Each differentiation experiment was performed in triplicate and repeated independently for at least three times.

Protein Arginine Methyltransferase 1 Enhances Erythroid Differentiation of K562 Cells
We found in our preliminary examination that the cellular profile of protein arginine methylation significantly changed during erythroid differentiation of K562 cells. Since PRMT1 is the predominant protein arginine methyltransferase, we investigated the role of PRMT1 in erythroid differentiation. K562 cells were transiently transfected with the control vector or pcDNAHA2-PRMT1 to express the N-terminally tagged HA-PRMT1 proteins ( Figure 1A, upper panel) and induced for erythroid differentiation. Benzidine staining for hemoglobin showed that overexpression of HA-PRMT1 significantly promoted both araC-induced and sodium butyrate (NaB)-induced erythroid differentiation ( Figure 1A, lower panels). K562 cell clones R2-1 and R2-3, which stably expressed HA-PRMT1 [22], also exhibited a higher efficiency in araC or NaB induction of erythroid differentiation, compared to the parental or vector control cells ( Figure 1B). The effect of PRMT1 was further examined in a different erythroleukemia cell line HEL. Transient overexpression of HA-PRMT1 also stimulated NaB-induced erythroid differentiation in HEL ( Figure S1). These results suggest that the stimulatory impact of PRMT1 is not restricted to K562 cells.
To confirm the effect of PRMT1 expression on erythroid differentiation, we examined the transcripts of c-globin and two key enzymes involved in heme synthesis, i.e., delta-aminolevulinate synthase 2 (ALAS2) and porphobilinogen deaminase (PBGD).
Real-time PCR results showed that araC treatment stimulated the expression of these genes in a time-dependent manner in the vector control cells, and stable expression of PRMT1 indeed significantly increased the transcript levels of ALAS2, PBGD and c-globin ( Figure 1C). GATA1, NF-E2 and EKLF are transcription factors critically involved in the differentiation toward the erythroid lineage [26]. We found that the expression of these three factors in control cells were increased during differentiation and overexpression of PRMT1 significantly enhanced the transcript levels of these transcription factors ( Figure 1D). The stimulatory effects were similarly observed in R2-3 cells (Figure The mutation of Gly80 to Arg80 impairs the methyltransferase activity of PRMT1 [22]. Transient expression of the PRMT1 G80R mutant ( Figure 1E, upper panel) was unable to enhance araC-or NaB-induced erythroid differentiation in K562 cells like the wild type PRMT1 did ( Figure 1E, lower panel). The G80R mutant also failed to stimulate ALAS2 gene expression ( Figure 1F). These results suggest that the methyltransferase activity of PRMT1 is essential for its stimulatory effect on erythroid differentiation.

Knockdown of PRMT1 Significantly Decreases Erythroid Differentiation
To examine the role of endogenous PRMT1 in erythroid differentiation, we used PRMT1-deficient clones PRMT1 KD-1 and KD-2 (Figure 2A), which were established previously by selecting clones stably expressing PRMT1 shRNAs [22]. Knocking down endogenous PRMT1 expression suppressed both araCinduced and NaB-induced erythroid differentiation (Figure 2A). Cell-surface expression of glycophorin A, a characteristic marker of erythrocytes, induced by araC and NaB was also suppressed in PRMT1-deficient cells ( Figure 2B and Figure S3A). PRMT1deficient cells also exhibited reduced expression of ALAS2, PBGD, c-globin and GATA1 transcripts ( Figure 2C and Figure S3B). Furthermore, when PRMT1 was re-introduced into PRMT1deficient cells by transfection, the differentiation of these cells was restored to a level comparable to that in the parental and vector control cells ( Figure 2D). These results provide solid evidence that endogenous PRMT1 positively regulates the differentiation toward the erythroid cell lineage.

The Activity of p38 MAPK is Required for the Effect of PRMT1 on Erythroid Differentiation
The p38 MAPK pathway has been shown to play a pivotal role in regulating erythroid differentiation [4]. We examined whether PRMT1 potentiated the activation of p38 MAPK. p38 MAPK is activated via dual phosphorylation on specific threonine and tyrosine residues by the upstream MKKs (MAP kinase kinases) [27]. By monitoring the level of phosphorylated p38 using Western blotting, we found that araC and NaB treatment stimulated the activation of p38 MAPK in parental K562 cells, and in the PRMT1-expressing R2-1 clone the activation was significantly enhanced 2 to 3 fold ( Figure 3A). Similar results were observed in R2-3 cells ( Figure S4A). The Erk1/2 MAPK was not activated by either araC or NaB in parental K562 cells ( Figure S4B). Ectopic expression of HA-PRMT1 did not appear to affect activation of Erk1/2 to any significant extents ( Figure S4B).
We further performed an in vitro kinase assay using ATF2 as the protein substrate, and demonstrated that the kinase activity of p38 MAPK was greatly suppressed in PRMT1-knockdown KD-1 and KD-2 cells ( Figure 3B and Figure S4C). The pharmacological inhibitor of p38 MAPK, SB203580, suppressed araC-and NaB-induced erythroid differentiation in a dosedependent manner in K562 cells, confirming an essential role of p38 MAPK in erythroid differentiation; notably, SB203580 limited the enhancement of differentiation observed in PRMT1expressing R2-1 cells, suggesting that the differentiationstimulating effect of PRMT1 expression was dependent on p38 activation ( Figure 3C). Similar results were also observed in R2-3 cells ( Figure S4D). Together, these results suggest that PRMT1 acts by enhancing the activation of p38 MAPK to promote erythroid differentiation.

p38a, but not p38b, Mediates the Stimulatory Effect of PRMT1 on Erythroid Differentiation
Although p38 MAPK has been shown to play a crucial role in regulating erythroid differentiation, the role of each isoform has not been fully elucidated. p38a and p38b are the most closely related members [27]. We first selectively knocked down either endogenous p38a or p38b using specific shRNAs; stable cell clones were obtained that showed a significant deficiency in either p38a or p38b proteins ( Figure 4A, upper panel). Knockdown of p38a resulted in dramatically reduced erythroid differentiation induced by araC, while knockdown of p38b had no effect ( Figure 4A, lower panel), indicating a crucial role of the a but not the b isoform of p38. Consistent with the above finding, expression of Flag-p38a but not Flag-p38b enhanced erythroid differentiation ( Figure 4B). In vitro kinase assays demonstrated that the Flag-p38a and Flag-p38b proteins expressed in K562 cells were similarly activated when cells were stimulated with sorbitol, a known activator for p38 MAPK  [28], indicating that both were functionally active kinases ( Figure 4C). These results indicate that p38a plays a positive role in regulating erythroid differentiation, whereas p38b does not participate in this regulation.
To further establish the link between p38a and PRMT1, we transiently expressed HA-PRMT1 in K562-derived cell clones, and found that expression of PRMT1 promoted erythroid differentiation in the control cells but not in the p38a-deficient clones ( Figure 4D), suggesting that the stimulatory effect of PRMT1 requires the presence of p38a. Notably, knockdown of p38b did not affect the stimulatory effect of PRMT1 ( Figure 4D), confirming that p38b is not involved in PRMT1-mediated modulation of erythroid differentiation. Expression of p38a in PRMT1-knockdown cell clones was still able to enhance erythroid differentiation ( Figure 4E), suggesting that p38a acts downstream of PRMT1. Consistently, the expression of p38b did not have apparent effects in either control or PRMT1-knockdown cells ( Figure 4E). These results demonstrate that p38a, but not p38b, mediates the stimulatory effect of PRMT1 on erythroid differentiation.

PRMT1 Binds with and Methylates p38a
To investigate the possibility that p38a is a direct target of PRMT1, we first performed immunoprecipitation in K562 cells. When p38a was immunoprecipitated using a specific antibody, the endogenous PRMT1 protein was co-precipitated ( Figure 5A, upper). Consistently, when PRMT1 was immunoprecipitated, p38a was associated with the precipitates (Figure 5A, lower). These results demonstrate the physical interaction between PRMT1 and p38a, and suggest that PRMT1 may activate the p38a pathway by directly targeting p38a. An in vitro methylation assay was carried out to test whether PRMT1 can methylate p38a. The results showed that methyl incorporation into the recombinant p38a proteins increased in a time-dependent manner, while no methylation was detected in the absence of either PRMT1 or p38a ( Figure 5B). These results showed for the first time that p38a can serve as a substrate for PRMT1. Together, our results suggest that PRMT1 may enhance the activation of p38a by directly binding to and methylating p38a.

PRMT1 Enhances the EPO-induced Erythroid
Differentiation of Primary Human CD34 + Hematopoietic Progenitor Cells through Activation of p38 MAPK EPO is the primary cytokine that regulates various steps in the differentiation of erythrocytes in the body [2]. To study the role of PRMT1 in a more physiologically relevant cell model, EPOinduced erythroid differentiation of primary CD34 + hematopoietic progenitor cells isolated from human umbilical cord blood was examined. The methyltransferase activity of PRMT1 in CD34 + cell homogenates was readily measured using hnRNP K as the substrate; methyl incorporation increased in a dose-dependent manner ( Figure 6A). Using three known PRMT1 substrate, hnRNP K, hnRNP A1 and hnRNP A2, in this in vitro methylation assay, we showed that upon the EPO treatment of CD34 + cells, the PRMT1 activity was rapidly stimulated and reached its peak at 2 hr ( Figure 6B). Considering individual variations, we performed assays on CD34 + cells isolated from different donors and quantified the methylation of three PRMT1 substrates 2 hr after EPO treatment; the results showed that homogenates from all donors contained an EPO-induced methyltransferase activity, although the fold of stimulation varied ( Figure 6C). While the methyltransferase activity in CD34 + cells was stimulated by EPO, the protein level remained stable ( Figure 6D), suggesting that the activity of PRMT1 may be regulated in response to EPO signaling. To evaluate the role of PRMT1, we then introduced PRMT1 into CD34 + progenitor cells by the TAT-mediated protein transduction method [22,29]. The TAT-fused PRMT1 and the control TAT-GFP proteins entered the cells rapidly and remained in the cells for up to 24 hr ( Figure 6E). TAT-PRMT1 significantly stimulated EPO-induced erythroid differentiation, as measured by either benzidine staining ( Figure 6F) or surface expression analysis of glycophorin A ( Figure 6G). Taken together, our results demonstrate that PRMT1 plays a positive role in modulating erythroid differentiation not only in K562 cells but also in human primary CD34 + hematopoietic progenitor cells induced by EPO.

Discussion
Despite intensive studies, the molecular events that regulate erythropoiesis are not fully elucidated. This study reveals for the first time that PRMT1, the predominant enzyme responsible for cellular protein arginine methylation, plays a crucial role in promoting erythroid differentiation in bipotent leukemia cells as well as in human primary CD34 + hematopoietic progenitor cells. Moreover, we have identified a specific isoform of p38 MAPK, p38a, that acts downstream of PRMT1 in mediating the stimulatory effect on erythroid differentiation. Furthermore, we show that PRMT1 is associated with p38a in cells and demonstrate that p38a is a direct substrate for PRMT1 in vitro. These results highlight a novel regulatory mechanism involving protein arginine methylation in modulating the signaling pathways leading to the development of the erythroid lineage.
A number of studies have reported that p38 MAPK is involved in erythroid differentiation [6,7,8], yet the upstream regulatory events required for the activation of p38 MAPK in this developmental program are still poorly understood. The p38 MAPKs are modulated by a phosphorylation cascade which leads to dual phosphorylation on threonine/tyrosine residues and the activation of the catalytic activity [27]. Dephosphorylation of MAPKs by specific phosphatases turns off the signaling event [30]. In addition, the MAPK signaling complex binds to scaffold proteins, which regulates the specificity of downstream events by influencing the enzyme's location, catalytic activity, duration of activation and substrate selection [27]. The association between PRMT1 and p38a discovered in this study suggests another potential regulatory mechanism for MAPK signaling. It is conceivable that methylation of p38a by PRMT1 may make it a better substrate for MAP kinase kinases (MKKs), reduce its dephosphorylation, or affect its interaction with scaffold or other regulatory proteins.
Methylation by PRMT1 produces mono-and asymmetric dimethyl-arginines which has been reported to affect phosphorylation through various mechanisms. When occurs in the consensus motif for Akt-mediated phosphorylation (RxRxxS/T), arginine methylation of FOXO1 transcription factor and BAD (BCL-2 antagonist of cell death) counteracts their phosphorylation by Akt and abrogates Akt-mediated signaling [18,19]. Arginine methylation may provide the binding surface for interacting partners and thus positively or negatively regulates phosphorylation events. PRMT1-mediated methylation of ASK1 (apoptosissignal-regulating kinase 1) promotes its interaction with the negative regulator thioredoxin and thus blocks the H 2 O 2 -induced activation of this kinase [31]. Methylation of Axin, a regulator of Wnt singaling, by PRMT1 promotes its interaction with and phosphorylation by glycogen synthase kinase 3b which leads to stabilization of Axin [21]. Further investigation is required to determine the effects of arginine methylation of p38a at the molecular level.
PRMT1 may have multiple targets in mediating the regulation of erythropoiesis. Although we have demonstrated that p38a is a PRMT1 substrate, given the scaffolding mechanism in MAPK signaling, we cannot rule out the possibility that PRMT1 also interacts with other components in the p38 signaling pathway. Moreover, our results show that PRMT1 promotes multiple aspects of the erythroid differentiation program, including the characteristic erythroid cell phenotypes and the expression of transcriptional regulators critical for lineage-specific differentiation. In addition to genes encoding c-globin and two key enzymes for heme synthesis, transcription factors GATA1, NF-E2 and ELKF are also up-regulated under PRMT1 overexpression. As both NF-E2 and ELKF are known downstream targets of GATA1 [32,33], this raises the possibility that GATA1 may be the primary target transcription factor regulated by the PRMT1/p38a pathway. One of the known mechanisms through which PRMT1 activates transcription is by methylating the histone H4 on Arg3, which permits the subsequent acetylation of histones. This is demonstrated in the retinoic acid-induced expression of transglutaminase type 2 (TGM2) gene [34] and the DMSO-induced expression of b-globin gene [16]. The possibility that PRMT1mediated up-regulation of transcription during erythroid differentiation involves histone H4 Arg3 methylation is not ruled out.
Our analysis of CD34 + hematopoietic progenitor cells shows that despite the increase in the methyltransferase activity following EPO treatment, the PRMT1 protein levels maintain steady, suggesting that the enzymatic activity of PRMT1 is regulated during erythroid differentiation. Up-regulation of PRMT1 activity has been reported in NGF-stimulated PC12 cells [35], during fatal liver development, and in the early stage during liver regeneration [36]. The molecular events regulating PRMT1 activity during these cellular processes are not clear. The activity of PRMT1 can be regulated by protein interactions. The interacting proteins, TIS21/BTG2 and BTG1, have been shown to modulate the substrate selectivity of PRMT1 [12]. CCR4-associated factor 1 (hCAF1) regulates PRMT1 activity in a substrate-dependent manner [37]. TR3, an orphan receptor, causes an inhibition of the methyltransferase activity of PRMT1 by directly binding to the catalytic domain [38]. In addition, when phosphorylated, the methyltransferase activity of CARM1 (PRMT4) and PRMT5 is inhibited [39,40]. It is not yet clear whether the enzymatic activity of PRMT1 is regulated by posttranslational modifications.
Erythroid cells and megakaryocytes share common bipotent megakaryocyte/erythrocyte progenitors [41]. Differentiation of progenitors toward these two lineages is mutually exclusive. The molecular mechanisms controlling this lineage-specific differentiation have only been partially elucidated. Several molecules are reported to serve as a switch in governing commitment/ differentiation toward specific lineages. A deficiency of erythroid Krüppel-like factor (EKLF) enhances the formation of megakaryocytes and suppresses erythroid cells [42,43]. c-Myc 2/2 mice display significantly increased numbers of megakaryocytic progenitors and a blockage of erythroid progenitors [44]. This study unveils a novel link between PRMT1 and p38 in positively regulating the erythroid differentiation program and suggests a novel regulatory mechanism for p38a through arginine methylation. Our previous study has shown that PRMT1 plays a negative role in megakaryocytic differentiation [22]. Taken together, these studies suggest that PRMT1, mediated by p38a, serves as a novel molecular switch in lineage-specific differentiation toward megakaryocytes or erythrocytes. Our results have significant impacts on our understanding of how protein arginine methylation mediated by PRMT1 and protein phosphorylation mediated by p38a are coordinated in regulating hematopoietic differentiation. These findings also suggest that PRMT1 and p38a can potentially serve as novel targets for therapeutic purposes. Figure S1 Ectopic expression of PRMT1 enhanced erythroid differentiation of HEL cells. Cells were transiently transfected with either empty vectors or pPCDNA3HA2-PRMT1 plasmids expressing HA-PRMT1 and treated with NaB (0.1 mM). Transient expression of PRMT1 significantly enhanced erythroid differentiation, as detected by benzidine staining at 24 hr after induction. All results shown are representative of three separate experiments. Differentiation results are presented as means 6 S.E. of three repeats; *, p,0.05; **, p,0.01; ***, p,0.005 compared with vector control cells. (TIF) Figure S2 The expression of erythroid-related genes was enhanced in PRMT1-overexpressing R2-3 cells. The expression of various erythroid-related genes (A) and transcription factors (B) in PRMT1-overexpressing R2-3 cells was examined by real-time PCR after araC treatment. Data were calculated as fold changes using vector control cells without treatment as a reference. All results shown are representative of three separate experiments. Results are presented as means 6 S.E. of three repeats. *, p,0.05; **, p,0.01; ***, p,0.005 compared with vector control cells.