Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. 2008 Oct; 28(20): 6262–6277.
Published online 2008 Aug 11. doi:  10.1128/MCB.00923-08
PMCID: PMC2577430

Protein Arginine Methyltransferase 5 Suppresses the Transcription of the RB Family of Tumor Suppressors in Leukemia and Lymphoma Cells[down-pointing small open triangle]


The proper epigenetic modification of chromatin by protein arginine methyltransferases (PRMTs) is crucial for normal cell growth and health. The human SWI/SNF-associated PRMT5 is involved in the transcriptional repression of target genes by directly methylating H3R8 and H4R3. To further understand the impact of PRMT5-mediated histone methylation on cancer, we analyzed its expression in normal and transformed human B lymphocytes. Our findings reveal that PRMT5 protein levels are enhanced in various human lymphoid cancer cells, including transformed chronic lymphocytic leukemia (B-CLL) cell lines. PRMT5 overexpression is caused by the altered expression of the PRMT5-specific microRNAs 19a, 25, 32, 92, 92b, and 96 and results in the increased global symmetric methylation of H3R8 and H4R3. An evaluation of both epigenetic marks at PRMT5 target genes such as RB1 (p105), RBL1 (p107), and RBL2 (p130) showed that promoters H3R8 and H4R3 are hypermethylated, which in turn triggers pocket protein transcriptional repression. Furthermore, reducing PRMT5 expression in WaC3CD5 B-CLL cells abolishes H3R8 and H4R3 hypermethylation, restores RBL2 expression, and inhibits cancer cell proliferation. These results indicate that PRMT5 overexpression epigenetically alters the transcription of key tumor suppressor genes and suggest a causal role of the elevated symmetric methylation of H3R8 and H4R3 at the RBL2 promoter in transformed B-lymphocyte pathology.

Enzymes that modify chromatin via arginine methylation have different effects on gene expression, depending on whether they catalyze either the symmetric or asymmetric dimethylation of histones (20, 31). Recent work has clearly established a link between the aberrant expression of chromatin-modifying enzymes and cancer; however, the impact protein arginine methyltransferases (PRMTs) have on cell growth and proliferation has only begun to be appreciated. The role PRMTs play in cancer epigenetics remains unexplored, and it is essential to unravel how histone arginine methylation influences cancer. Previous studies have shown that the expression of EZH2, the catalytic subunit of PRC2/3, is altered in cancer cells (19, 32, 40, 41). Similarly, several reports have shown that the interplay and cross-talk between chromatin-modifying enzymes is necessary for the efficient regulation of gene expression, and that changes that affect the activity and/or targeting of chromatin remodelers can trigger cancer (13, 23).

Histone methylation has emerged as an important epigenetic modification in the control of chromatin structure and gene expression, and there is little doubt that different epigenetic marks act either synergistically or antagonistically to specify transcriptional performance in various chromatin regions (13). The methylation of histones can be found on both lysine and arginine residues, and both modifications are introduced by SET and PRMT enzymes, respectively. Just like SET proteins, which either can induce or repress transcription, PRMTs either can induce or inhibit transcription depending on the modified residue (2, 25). For example, the asymmetric methylation of H3R17 and H4R3 is associated with active transcription, whereas the symmetric methylation of H3R8 and/or H4R3 is a mark for gene silencing (6, 26, 29, 30).

Members of the highly conserved family of PRMTs can be classified into either type I enzymes, which catalyze ω-NG-monomethylation and ω-NG,NG′-asymmetric dimethylation, or type II chromatin modifiers that introduce ω-NG-monomethylation and ω-NG,NG′-symmetric dimethylation (31). To date there are 11 PRMTs, and except for PRMT2, PRMT10, and PRMT11, they all have the ability to catalyze arginine methylation (6, 9, 21, 28, 29, 30, 35). Among all PRMTs, only PRMT4/CARM1 and PRMT5 have been shown to interact with ATP-dependent chromatin remodelers (22, 28, 29, 36). As part of the NUMAC complex, PRMT4/CARM1 induces the transcription of its target genes, whereas PRMT5-containing hSWI/SNF/mSIN3A-HDAC complexes are associated with transcriptional repression (28, 29, 44). Although the function of PRMT5 in gene regulation has been characterized, its involvement in cancer remains unclear.

We have previously determined, by using PRMT5 knockdown cell lines, that RBL1 (p107) gene transcription is enhanced, suggesting a possible direct link between PRMT5 and pocket proteins (29). We also have shown that when PRMT5 protein levels are increased, immortalized NIH 3T3 cells hyperproliferate and acquire the ability to grow in an anchorage-independent manner (29). To further understand the mechanisms by which PRMT5 induces these profound cellular changes and to investigate the relevance of the relationship between PRMT5 and RB-related proteins, we examined their expression in a panel of transformed lymphoid cell lines, including transformed WaC3CD5, Mec1, and Mec2 chronic lymphocytic leukemia (B-CLL) cell lines. In comparison to other B-cell disorders, B-CLL is a leukemia characterized by the slow proliferation of B cells that express CD5, CD19 or CD20, CD23, CD25, CD69, and CD71 and display low levels of immunoglobulin D (IgD), CD22, and CD79b (7). Although patients with B-CLL survive for long periods of time, other patients succumb to the disease very rapidly, and currently there are no known cures for these patients. Besides what is known about receptor expression, telomere length, telomerase activity, and the mutation of IgVH and p53, nothing is known about the levels of epigenetic modifiers in B-CLL cells. Using transformed B-CLL cell lines, we examined the levels of PRMT5 and the epigenetic marks it introduces in H3 and H4. Our results show that PRMT5 is overexpressed in all B-CLL cell lines examined, and this increase in PRMT5 expression is accompanied by the enhanced global symmetric methylation of H3R8 and H4R3. We also show that the increased translation of PRMT5 is a direct result of the aberrant expression of PRMT5-specific microRNAs (miRNAs) and correlates with the transcriptional silencing of the RB family of tumor suppressor genes. Furthermore, reducing the expression of PRMT5 interferes with the proliferation of transformed WaC3CD5 B-CLL cells and derepresses the transcription of pocket protein genes. Thus, the present study shows that the misregulated expression of PRMT5 not only alters global H3R8 and H4R3 methylation in transformed B lymphocytes, including B-CLL cells, but also results in the epigenetic silencing of the RB family of tumor suppressor genes.


Plasmid constructions.

Plasmids used to produce lentiviral particles that express antisense PRMT5 (AS-PRMT5) were described previously (30).

Cell culture, B-cell isolation and activation, generation of lentiviral particles, and infection of transformed B cells.

Normal and transformed B cells were cultured in RPMI 1640 supplemented with 1 mM sodium pyruvate and 10% fetal bovine serum (FBS). Normal B cells were isolated from tonsilar tissues obtained from Nationwide Children's hospital through the Cooperative Human Tissue Network as described previously (30). Normal B cells purified by this method were analyzed by fluorescence-activated cell sorting using anti-CD19-phycoeryrthrin antibody and generally were 90 to 95% pure. To activate resting B lymphocytes, purified B cells were resuspended at a density of 5 × 106 cells/ml in RPMI 1640 containing 10% FBS, and then cells were induced to enter the cell cycle by adding 15 ng/ml of recombinant human interleukin-4 (IL-4) (Invitrogen, Inc.) and 15 μg/ml of goat anti-human IgG + IgM (Jackson ImmunoResearch, Inc.). To measure the efficiency of B-cell activation, bromodeoxyuridine (BrdU) incorporation was carried out for various periods of time, as described previously (29). To check the level of pocket proteins in activated B lymphocytes, Western blot analysis was conducted after 4 days of activation. To generate control lentivirus or lentiviral particles for knocking down PRMT5 expression, 293T cells were transfected by calcium phosphate with a combination of vectors containing 5 μg of pVSV-G, 15 μg of pCMVΔ8.2, and 20 μg of pRRL-puro or pRRL-AS-PRMT5 (34). Control PRMT5 and AS-PRMT5 lentiviral supernatants were harvested 48 h posttransfection, clarified by centrifugation at 2,000 rpm for 4 min, and used to infect WaC3CD5 B-CLL cells as described previously (30). To monitor the expression of PRMT5 and its target genes as well as the proliferation of WaC3CD5 cells, cells infected with either control or AS-PRMT5 lentiviral particles were harvested every 2 days for 4 to 6 days after infection and either counted or lysed to prepare protein extracts and RNA.

Antibodies, Western blotting, and immunofluorescence.

For Western blot analysis, nuclear, cytosolic, or whole-cell lysates were separated on sodium dodecyl sulfate-8 to 10% polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane, and specific proteins were detected by enhanced chemiluminescence using previously described antibodies (30, 42). Antibodies to detect RB1 were purchased from BD Biosciences, while antibodies to detect RBL1 and RBL2 were purchased from Santa Cruz Biotechnology Inc. Immunofluorescence analysis was performed by fixing 1 × 104 cells in a 4% paraformaldehyde solution for 15 min at room temperature, followed by a 10-min incubation in 0.1% Triton X-100. Fixed and permeabilized cells were blocked in 10% goat serum for 2 h, rinsed with 1× phosphate-buffered saline (PBS), and incubated with either preimmune or immune antibodies raised against PRMT5, H3(Me2)R8, and H4(Me2)R3. Samples were washed with 1× PBS and incubated with fluorescein isothiocyanate-labeled goat anti-rabbit antibody for 1 h at 37°C. Excess secondary antibody was removed by rinsing samples with 1× PBS, and genomic DNA was stained with 4′,6′-diamidino-2-phenylindole (DAPI) before samples were mounted and visualized by fluorescence microscopy using a Zeiss axioscope at 100× magnification.

RT and real-time PCR.

To measure steady-state levels of PRMT5 mRNA and its target genes, total RNA was isolated using TRIzol reagent, and 20-μl reverse transcription (RT) reactions containing 2.5 μM random hexamer primers and 1 μg of total RNA were carried out using a kit as specified by the manufacturer (Applied Biosystems, Inc.). Next, 10-μl real-time PCRs, which include 1 μl of RT reaction mix, 1× TaqMan universal master mix, 0.8 μM of each 5′ and 3′ gene-specific primer, and 0.2 μM probe, were performed using the TaqMan system according to the manufacturer's instructions (Applied Biosystems, Inc.). Primer sets for RB1 (forward, 5′-AGGACCCAGAGCAGGACAG-3′; reverse, 5′-AGGTTCTTCTGTTTCTTCAAACTCA-3′; Roche universal probe no. 34), RBL1 (forward, 5′-GAACCACCAAAGTTACCACGA-3′; reverse, 5′-ATTAAACAGATCCTTAACACTGCAAG-3′; Roche universal probe no. 38), RBL2 (forward, 5′-TTGTTGGGTGCTTTTTATATATGC-3′; reverse, 5′-TTTCCATAAACTAAGTCCAAAGCA-3′; Roche universal probe no. 62), E2F1 (forward, 5′-ACCCTGACCTGCTGCTCTT-3′; reverse, 5′-GCCAGGTACTGATGGTCAGTTT-3′; Roche universal probe no. 75), and RR2 (forward, 5′-CAGCAAGCGATGGCATAGT-3′; reverse, 5′-AGCGGGCTTCTGTAATCTGA-3′; Roche universal probe no. 22) were used. The primer sets and probes used to detect PRMT5 and suppressor of tumorigenicity 7 (ST7) mRNAs by real-time RT-PCR were described previously (30). For internal controls, the expression of GAPDH and β-ACTIN mRNAs was measured using 1× premixed GAPDH and β-ACTIN primer/probe sets, respectively (Applied Biosystems, Inc.).

Polyribosome profiling.

Whole-cell extracts were prepared essentially as described previously (30). Briefly, 2 × 107 normal or transformed B cells were washed twice with 1× PBS and resuspended in 250 μl of lysis buffer containing 20 mM HEPES (pH 7.5), 100 mM KCl, 10 mM MgCl2, 0.25% NP-40, 100 μg/ml cycloheximide, 100 U/ml RNasin, 1 mM dithiothreitol, and protease inhibitors. Next, cells were disrupted by passing them several times through a 27.5-gauge needle before loading whole-cell lysate onto a 15 to 40% sucrose gradient (4.8 ml). Samples were spun in an SW55 rotor at 43,000 rpm for 2.5 h at 4°C, and 200-μl fractions were collected. The absorbance of each fraction was determined at 254 nm, total RNA was purified using TRIzol reagent, and PRMT5 mRNA levels in each fraction were determined by real-time RT-PCR using β-ACTIN mRNA as the internal control.


Labeled probes to detect various miRNAs were generated using the mirVana miRNA probe construction kit (Ambion, Inc.). A single-stranded DNA oligonucleotide including the desired miRNA sequence and the T7 promoter sequence at the 3′ end was hybridized to the T7 promoter primer and extended using Klenow DNA polymerase for 30 min at 37°C. The double-stranded DNA (dsDNA) template generated then was in vitro transcribed using T7 RNA polymerase in the presence of [α-32P]CTP for 30 min at 37°C. The reaction was stopped by adding DNase I, and 32P-labeled full-length RNA was purified by electrophoresis on a 12% polyacrylamide gel. RNase protection assays (RPAs) were performed using the mirVana miRNA detection kit (Ambion, Inc.), which consists of hybridizing 5 × 104 cpm of each miRNA-specific probe with 10 μg of total RNA from either normal or transformed lymphoid cells at 42°C. As controls, miRNA-specific probes either were digested alone or incubated with 10 μg of yeast tRNA before digestion with a 1:100 dilution of RNase A+T1 in a 150-μl reaction mixture containing RNase digestion buffer. Reaction mixtures were incubated for 45 min at 37°C before they were stopped by adding 225 μl of RNase inactivation-precipitation solution and 225 μl of ethanol. Protected dsRNAs were harvested by centrifugation after samples were incubated for 2 h at −80°C, and pellets were resuspended in gel loading buffer. Samples were heat denatured and separated on a 15% polyacrylamide-urea gel. Individual bands representing protected miRNA products were visualized by a PhosphorImager and quantitated using ImageQuant v5.0.

RNA interference and transfection assays.

Small interfering RNAs (siRNA) were generated using a Silencer siRNA construction kit (Ambion, Inc.). To determine the extent to which in vitro-synthesized siRNAs can interfere with PRMT5 expression, 2.5 or 5.0 μg of dsRNA template consisting of wild-type or mutant miR-19a, miR-25, miR-32, miR-92b, miR-96, wild-type siPRMT5, or wild-type RBL2 was electroporated into 5 × 106 transformed WaC3CD5 or Mec1 B-CLL cells using an Amaxa Biosystems electroporator. Next, cells were incubated in 3 ml of RPMI 1640 containing 10% FBS, and whole-cell extracts were prepared 30 h later in 100 μl of radioimmunoprecipitation assay lysis buffer before PRMT5 and pocket proteins were detected by Western blotting. Primers used to generate wild-type and mutant miR-92b and miR-96 dsRNAs were described previously (30). To synthesize wild-type and mutant miR-19a, miR-25, and miR-32 dsRNAs, T7-driven single-stranded DNA oligonucleotides (corresponding to the following underlined sequences) were used according to the manufacturer's instructions (Ambion, Inc.): wild-type miR-19a sense, 5′-AATGTGCAAATCTATGCAAAACTGAC CTGTCTC-3′; antisense, 5′-AATC AGTTTTGCATAGATTTGCACACCTGTCTC-3′; mutant miR-19a sense, 5′-AAACCTACGCTGTTAGTCTGTGTGACCTGTCTC-3′; antisense, 5′-AATCACACAGACTAACAGCGTAGGTCCTGTCTC-3′; wild-type miR-25 sense, 5′-AACATTGCACTTGTCTCGGTCTGACCTGTCTC-3′; antisense, 5′-AATCAGACCGAGACAAGTGCAATGCCTGTCTC-3′; mutant miR-25 sense, 5′-AACCAACGTACGGTCTCGGTCTGACCTGTCTC-3′; antisense, 5′-AATCAGACCGAGACCGTACGTTGGCCTGTCTC-3′; wild-type miR-32 sense, 5′-AATATTGCACATTACTAAGTTGCCCTGTCTC −3′; antisense, 5′-AAGCAACTTAGTAATGTGCAATACCTGTCTC-3′; mutant miR-32 sense, 5′-AAACCCTAGAACGAACAAGTTGCCCTGTCTC-3′; antisense, 5′-AAGCAACTTGTTCGTTCTAGGGTCCTGTCTC-3′; wild-type miR-197 sense, 5′-AAGCTGGGTGGAGAAGGTGGTGAACCTGTCTC-3′; antisense, 5′-AATTCAC CACCTTCTCCACCCAGCCCTGTCTC-3′; and mutant miR-197 sense, 5′-AAGACATTGTACGAAGGTGGTGAACCTGTCTC-3′; antisense, 5′-AATTCACCACCTTCTTTCAATCTACCTGTCTC-3′. To knock down the expression of PRMT5 and RBL2, T7-driven single-stranded DNA oligonucleotides were generated according to previously published sequences (3, 15).

Chromatin immunoprecipitation (ChIP) assay.

Chromatin was isolated from both normal and transformed B cells as described previously (42). To amplify genomic sequences of the RB family of tumor suppressor genes (ST5, HOXA2, and ALDOA), the following primer pairs and probes were used: RB1 (forward, 5′-TTGAAATTATTTTTGTAACGGGAG-3′; reverse, 5′-CAGCGAGCTGTGGAGGAG-3′; Roche universal probe no. 55), RBL1 (forward, 5′-CGGAGGAAAAAC GGACTTT-3′; reverse, 5′-GGGACGTGTTGTCATCCAC-3′; Roche universal probe no. 16), RBL2 (forward, 5′-ATTTTTGGCCCCCTTGAA-3′; reverse, 5′-GCACCCGTAGTCTTGAGCAC-3′; Roche universal probe no. 3), ST5 (forward, 5′-CGCCACGAAAGGTCAGAG-3′; reverse, 5′-CTTAAGCTCCGATACCTGCTG-3′; Roche universal probe no. 17), HOXA2 (forward, 5′-AACACCCTAGCAGCGATATTCT-3′; reverse, 5′-GGGCAAGGCCTAGGAAAA-3′; Roche universal probe no. 6), and ALDOA (forward, 5′-TCCAGCCTGAGGTCCTCTAA-3′; reverse, 5′-GAGAATGGTCCTTCATCTCGTC-3′; Roche universal probe no. 69).

Statistical analysis.

Results were expressed as the means ± standard deviations unless otherwise specified. Paired t tests and analysis of variance (ANOVA) were used to generate P values for comparisons between two groups and when multiple samples within different groups were used, respectively.


Expression of PRMT5 and the level of its epigenetic marks are altered in transformed B-CLL cell lines.

We have previously shown that the overexpression of PRMT5 induces hyperproliferation and cellular transformation, and more recently we have determined that the increased expression of PRMT5 is associated with mantle cell lymphoma (MCL) pathology (29, 30). To further confirm the relevance of our findings and to determine if the misregulation of PRMT5 is widespread in other B-lymphocyte disorders such as B-CLL, which is distinct and extremely different from MCL and Burkitt's lymphoma, nuclear and cytosolic extracts from normal CD19+ B lymphocytes and transformed WaC3CD5, Mec1, and Mec2 B-CLL cell lines were analyzed by Western blotting (Fig. (Fig.1A).1A). For comparison, we also analyzed two other transformed B-cell lines, JeKo (MCL) and Raji (Burkitt's lymphoma), and found that PRMT5 protein levels were elevated in both nuclear and cytosolic compartments of all transformed lymphoid cell lines examined. PRMT5 is found in both the nuclear and cytosolic fractions of transformed cells, while its localization is restricted primarily to the cytosol in normal B lymphocytes, as evidenced by the presence of BRG1, BRM, and BAF57 nuclear proteins in the nuclear fraction and α-tubulin in the cytosolic fraction.

FIG. 1.
PRMT5 is overexpressed in transformed B-cell lines. (A) Western blot analysis of PRMT5 and human SWI/SNF subunits using 20 μg of nuclear (N) and cytosolic (C) extracts from either normal cells or the indicated transformed cell lines. The expression ...

To solidify our Western blot results, we used immunofluorescence to measure the level of PRMT5 and the epigenetic methylation marks it introduces in chromatin (Fig. (Fig.1B).1B). We found that both PRMT5 and the global symmetric methylation of H3R8 and H4R3 are enhanced in WaC3CD5 and Mec1 B-CLL cell lines. In accord with these findings, when core histones were isolated from normal and transformed B lymphocytes and analyzed by Western blotting, there was a clear increase in the symmetric methylation of H3R8 and H4R3 (Fig. (Fig.1C),1C), suggesting that the elevated expression of PRMT5 alters the global symmetric methylation of histones H3 and H4 in both acute and chronic B-cell disorders.

PRMT5 mRNA translation is enhanced in transformed WaC3CD5, Mec1, and Mec2 B-CLL cell lines.

Our recent studies have shown that PRMT5 protein expression is augmented in the patient-derived Mino and JeKo MCL cell lines as a direct result of enhanced translation (30). To determine if PRMT5 protein expression is altered in B-CLL cells via a similar mechanism, we analyzed both steady-state levels of PRMT5 mRNA and its association with polyribosomes (Fig. (Fig.2).2). Real-time RT-PCR revealed that PRMT5 mRNA is reduced 1.5- to 5.5-fold (P < 10−4) in transformed lymphoid cells (Fig. (Fig.2A),2A), indicating that there is a lack of correlation between PRMT5 mRNA and protein levels in transformed B-CLL cell lines. To determine if this decrease in mRNA levels is unique to PRMT5, we measured the levels of another well-characterized PRMT, PRMT6 (Fig. (Fig.2B).2B). Our findings clearly show that unlike PRMT5, PRMT6 mRNA levels are enhanced 2.8- to 4-fold (P = 0.01) in transformed B cells, indicating that the observed decrease in PRMT5 mRNA expression is specific.

FIG. 2.
Expression of PRMT5 is enhanced at the translational level in transformed lymphoid cells. (A) To measure the steady-state levels of PRMT5 mRNA in normal and transformed B cells, real-time RT-PCR was performed three times in triplicates using total RNA. ...

Since we have previously shown that PRMT5 protein levels are elevated in Mino and JeKo MCL cell lines via enhanced translation, we sought to determine if the altered expression of PRMT5 in WaC3CD5, Mec1, and Mec2 B-CLL cell lines is achieved via a similar mechanism (Fig. (Fig.2C).2C). When we measured PRMT5 mRNA translational efficiency by polyribosome profiling using sucrose gradient sedimentation and the RNA collected from each gradient fraction was analyzed by real-time RT-PCR, we found that PRMT5 mRNA translation is enhanced fivefold in WaC3CD5 (P = 0.008), Mec1 (P = 10−2), and Mec2 (P = 0.006) B-CLL cell lines compared to the translation of normal B cells. Similar results were observed when RNA from JeKo (fivefold; P = 0.002) and Raji (4.6-fold; P = 10−2) gradient fractions were analyzed by real-time RT-PCR (Fig. (Fig.2C).2C). These results show that increased PRMT5 expression via enhanced translation is common among transformed acute and chronic B-cell disorders.

Misregulation of an elaborate miRNA program is involved in inducing PRMT5 translation.

The aberrant expression of specific miRNAs has been tightly linked to altered gene expression (1). For instance, we have found that miR-92b and miR-96 both are required for the proper regulation of PRMT5 expression (30). To date, 59 miRNAs have been identified and predicted to anneal to the 3′-untranslated region (3′UTR) of PRMT5 mRNA. To gain a better understanding of PRMT5 mRNA translation, we sought to identify additional miRNAs whose expression might be altered in transformed B lymphocytes. Using RPAs, we examined the level of the 15 PRMT5-specific miRNAs that showed the best seed sequence complementarity and discovered that the expression of miR-19a (P < 10−4), miR-25 (P < 10−4), miR-32 (P < 10−4), miR-92 (P = 0.003), miR-92b (P = 0.004), and miR-96 (P = 10−3) was reduced 1.5- to 5-fold in transformed B lymphocytes (Fig. (Fig.3A).3A). The only exception noted was in the case of miR-19a, whose expression in JeKo cells was increased twofold compared to that of normal B cells. Furthermore, while the expression of miR-19b did not fluctuate in normal and transformed B lymphocytes, we were unable to detect the other eight PRMT5-specific miRNAs, suggesting that their expression is restricted to non-lymphoid cell lineages. Thus, it appears that not all PRMT5-specific miRNAs are altered in their expression and that the abrogation of the level of a select few might be sufficient to enhance PRMT5 translation.

FIG. 3.
PRMT5-specific miRNAs are expressed differently in normal and transformed B cells, and their reexpression can inhibit PRMT5 translation. (A) RPAs were performed on 10 μg of total RNA isolated from the indicated cells using miR-19a, miR-19b, miR-25, ...

Our previous work has shown that the reexpression of miR-92b or miR-96 in JeKo and Raji cell lines causes a significant decrease in PRMT5 protein levels (30). To assess if miR-92b, miR-96, and the newly identified miR-19a, miR-25, and miR-32 have a direct effect on PRMT5 expression in transformed B-CLL cell lines, we electroporated WaC3CD5 and Mec1 cell lines with either wild-type or mutant versions of PRMT5-specific miRNAs whose expression has been found to be altered (Fig. (Fig.3B).3B). The electroporation of wild-type miR-19a in WaC3CD5 and Mec1 cell lines reduced PRMT5 protein expression by 70 and 90%, respectively. Although the reexpression of miR-32 did not result in a large reduction of PRMT5 protein expression in WaC3CD5 (15 to 25%) and Mec1 (15 to 50%), we noticed a greater decline in PRMT5 protein levels when miR-25, miR-92b, and miR-96 were electroporated into WaC3CD5 (70 to 85%) and Mec1 (75 to 80%) B-CLL cell lines. The electroporation of PRMT5-specific miRNAs had no effect on β-actin expression, and similarly the electroporation of miR-197, which does not show any sequence complementarity to PRMT5 mRNA, did not affect PRMT5 expression. These results show that the decrease in PRMT5 protein expression mediated by miR-19a, miR-25, miR-32, miR-92b, and miR-96 is highly specific.

PRMT5 targets the RB family of tumor suppressor genes.

In a previous study, we were able to determine by cDNA microarray analysis that PRMT5 is involved in the transcriptional regulation of 270 genes (29). In the same report, we were able to determine that the upregulation of PRMT5 induces the silencing of a large set of genes, including the retinoblastoma-like-1 (RBL1) tumor suppressor gene. To investigate if PRMT5 is directly involved in the transcriptional regulation of RBL1, we conducted ChIP experiments using cross-linked chromatin from normal B cells and transformed JeKo, Raji, WaC3CD5, Mec1, and Mec2 lymphoid cell lines (Fig. (Fig.4A).4A). When anti-PRMT5 antibody was used to immunoprecipitate chromatin from normal and transformed B cells, there was a clear increase in PRMT5 recruitment to the RBL1 promoter-proximal region in all transformed lymphoid cells. Real-time PCR revealed that PRMT5 recruitment to the RBL1 promoter was enriched 4- to 9.5-fold (P < 10−4) in transformed lymphoid cell lines compared to that of normal B cells. We have established that when PRMT5 is recruited to induce transcriptional silencing, both of its substrates, H3R8 and H4R3, become symmetrically methylated (29, 30). To examine if both epigenetic marks are introduced in the RBL1 promoter, we performed ChIP assays in normal and transformed B cells using anti-H3(Me2)R8 and anti-H4(Me2)R3 antibodies (Fig. (Fig.4B).4B). The results show that the symmetric methylation of H3R8 was increased 2.7- to 3.8-fold (P < 10−4), while the epigenetic modification of H4R3 was enhanced twofold (P = 0.0003) in transformed B-cell lines.

FIG. 4.
PRMT5 is recruited to and epigenetically modifies the RB family of tumor suppressor genes. ChIP assays were performed on cross-linked chromatin from either normal or transformed B cells using preimmune (PI) or immune anti-PRMT5 (A, C, and E) antibody ...

Pocket proteins are well known for their ability to regulate the expression of genes involved in G1-to-S transition through their interaction with the E2F family of transcription factors (8, 17). Since RBL1 is one of three members of the RB family of tumor suppressors, we wanted to determine if PRMT5 was involved in the epigenetic regulation of other members of the family. Remarkably, ChIP analysis showed that PRMT5 is associated with the promoter regions of RB1 and RBL2 (Fig. 4C and E). The recruitment of PRMT5 to the RB1 promoter was enhanced 3- to 4.7-fold (P < 10−4), while PRMT5 association with RBL2 was increased 3- to 5.2-fold (P < 10−4) in transformed lymphoid cell lines. When we examined the levels of H3R8 and H4R3 symmetric methylation at the RB1 and RBL2 promoters, we found that both marks were increased (Fig. 4D and F). The epigenetic modification of H3R8 was enriched fourfold (P < 10−4) at the RB1 promoter and 4.5- to 6.4-fold (P < 10−4) at the RBL2 promoter, while the symmetric methylation of H4R3 was enhanced three- to fourfold at the RB1 (P < 10−4) and RBL2 (P < 10−4) promoters. Since we had previously shown that ST7, another PRMT5 target gene, is suppressed in JeKo MCL cells, we wanted to evaluate the PRMT5 recruitment and symmetric methylation of H3R8 and H4R3 at its promoter (see Fig. S1 in the supplemental material). We found that PRMT5 association with the ST7 promoter was enhanced 3.3- to 5-fold (P = 0.0005) in transformed lymphoid cells (see Fig. S1A in the supplemental material), while the symmetric methylation of H3R8 was increased 3.9- to 7.5-fold (P < 10−4) and the symmetric methylation of H4R3 was enriched 3.8- to 5-fold (P = 0.0002) (see Fig. S1B in the supplemental material).

The recruitment of PRMT5 to the RB family of tumor suppressor genes and the symmetric methylation of their promoter histones H3 and H4 are specific, because when cross-linked chromatin from either normal or transformed B cells was immunoprecipitated with anti-PRMT5, anti-H3(Me2)R8, or anti-H4(Me2)R3, there were no significant changes in the promoter region of ST5 (Fig. 4G and H). Furthermore, to demonstrate the specificity of PRMT5 recruitment and the enrichment of its epigenetic marks in the promoter region of the RB family of tumor suppressors, we examined the recruitment of another PRMT family member, PRMT6, which has been shown to be involved in regulating HOXA2 gene expression (18), and we also monitored the methylation of H3K4, an epigenetic mark known to be associated with active promoters. When anti-PRMT6 antibody was used to immunoprecipitate chromatin from either normal or transformed B cells, there was no noticeable enrichment in PRMT6 recruitment at pocket protein promoters, while PRMT6 association with the HOXA2 promoter was enriched three- to fourfold (P = 0.01) in transformed B cells (see Fig. S2 in the supplemental material). Similarly, when we used an antibody that recognizes triply methylated H3K4, we discovered that contrary to the 3.8- to 5.4-fold (P = 0.02) enrichment observed at the ALDOA control promoter, the methylation of H3K4 was excluded from pocket protein promoters in transformed lymphoid cells (see Fig. S3 in the supplemental material). Collectively, these results show that PRMT5 specifically targets the RB family of tumor suppressor genes and epigenetically alters their promoter histones H3 and H4 in transformed lymphoid cell lines.

PRMT5 recruitment induces transcriptional repression of RB family members.

To determine if the increased association of PRMT5 with pocket protein promoters and the enhanced symmetric methylation of H3R8 and H4R3 alters the transcription of the RB family of tumor suppressors, we analyzed their mRNA levels in both normal and transformed B cells (Fig. (Fig.5A).5A). When real-time RT-PCR was performed, we found that RBL1 mRNA was repressed 1.7- to 2.5-fold (P < 10−4), RB1 mRNA was decreased 2.2- to 14.7-fold (P < 10−4), and RBL2 mRNA was inhibited 2.3- to 6.7-fold (P < 10−4) in transformed B lymphocytes. Similarly, when we measured ST7 mRNA expression in transformed B cells, we found that its levels were reduced 2.9- to 20-fold (P = 0.009) (see Fig. S1C in the supplemental material). To determine if this decrease in mRNA levels of RB family members is accompanied by a decrease in protein expression, we analyzed whole-cell extracts by Western blotting (Fig. (Fig.5B).5B). With the exception of RBL2, for which there was a significant drop in protein expression, the levels of RB1 and RBL1 were either unchanged, as is the case for RB1 in Mec1 and Mec2, or showed a moderate increase in transformed lymphoid cells. As expected, the levels of β-actin were unaffected in all samples. These results suggest that while there is a good correlation between RBL2 mRNA transcriptional repression and RBL2 protein expression, the same is not true for RB1 and RBL1.

FIG. 5.
Expression of pocket proteins in normal and transformed B cells. (A) The mRNA expression of the RB family of tumor suppressors was measured in normal and transformed B cells by real-time RT-PCR. Graphs show normalized changes in expression for each gene ...

Because RB1 and RBL1 protein expression was higher in transformed lymphoid cells, in which it was obvious that their mRNA levels were reduced compared to those of normal B cells, we reasoned that this might be directly related to the decreased expression of RB1- and RBL1-specific miRNAs. There are 10 miRNAs predicted to bind to the 3′UTR of RB1, and 14 miRNAs are capable of annealing to the 3′UTR of RBL1. When we checked the expression of all RB1- and RBL1-specific miRNAs using total RNA from normal and transformed B cells, we discovered that the majority either were not expressed or showed no change in their expression; however, the levels of the RBL1-specific miR-22 were decreased 3.6- to 11-fold (P = 0.01), while the expression of the RB1-specific miR-649 was two- to fivefold lower in transformed lymphoid cells compared to the expression of normal B cells (Fig. 5C and D).

Several reports have previously shown that the levels of RBL1 and RBL2 fluctuate in a cell cycle-dependent manner, while RB1 protein expression remains almost constant (8, 14, 43). As cells transit from G0/G1 to S phase, RBL2 protein levels decrease and RBL1 levels increase, and it is believed that these changes, accompanied by the hyperphosphorylation of RB1, promote progression through the cell cycle. To determine if the changes in the mRNA and protein expression of the RB family of tumor suppressor genes are the result of cell cycle-dependent events or if these PRMT5-induced changes are part of the mechanism by which cells become transformed, we induced normal B cells to enter the cell cycle and measured both mRNA and pocket protein expression. The stimulation of normal B cells with human IL-4 in the presence of goat anti-human IgG + IgM resulted in 43% activation as measured by BrdU incorporation (Fig. (Fig.6A).6A). When total RNA and whole-cell extracts from resting and activated B cells were analyzed, RBL1 mRNA and protein expression were induced (Fig. 6B and C). The evaluation of RB1 expression at the transcriptional as well as the translational level revealed that while RB1 mRNA was repressed twofold (P < 10−4), protein expression was unaltered. In stark contrast, B-cell activation had no effect on RBL2 mRNA and protein expression. Taken together, these findings suggest that while the decreased expression of RB1 mRNA in transformed lymphoid cells is cell cycle dependent, the PRMT5-induced transcriptional repression of RBL1 and RBL2 is transformation dependent.

FIG. 6.
RBL2 mRNA and protein levels do not fluctuate in proliferating normal B cells. (A) To monitor normal B-cell activation and proliferation, the fraction of BrdU-positive cells was determined after 0, 2, and 4 days of treatment with recombinant human IL-4 ...

Reducing PRMT5 expression slows down growth of transformed B-CLL cells.

The overexpression of PRMT5 is a hallmark of various transformed cell lines of different origins (30). To evaluate the effect of PRMT5 on transformed B-CLL cell growth and proliferation, we reduced its expression by infecting WaC3CD5 cells with recombinant lentivirus that expresses AS-PRMT5 (Fig. (Fig.7).7). Western blot analysis showed that 4 days postinfection, PRMT5 levels were reduced fourfold compared to those of control cells infected with lentiviral particles containing vector alone (Fig. (Fig.7A).7A). Decreased PRMT5 expression was paralleled by a 1.5-fold (P = 0.002) decrease in WaC3CD5 cell proliferation (Fig. (Fig.7B).7B). Having found that PRMT5 is directly involved in the transcriptional regulation of the RB family of tumor suppressor genes and that RBL2 protein expression was repressed in transformed lymphoid cells, we wanted to determine if knocking down PRMT5 expression could trigger the transcriptional and translational derepression of RBL2. RNA was collected from control and test WaC3CD5 cells and was used to measure RBL2 mRNA levels (Fig. (Fig.7C).7C). Real-time RT-PCR showed that the transcription of RBL2 was unaffected 2 days postinfection; however, by day 4 RBL2 mRNA was enhanced 2.3-fold (P = 0.002). Similarly, when we analyzed whole-cell extracts, we found that RBL2 protein expression was induced 2.8-fold, while the expression of RB1 and RBL1 was unaffected (Fig. (Fig.7D).7D). These results provide more evidence that PRMT5 is directly involved in the epigenetic regulation of RBL2 and demonstrate that reducing PRMT5 levels in transformed lymphoid cells brings about RBL2 derepression, which in turn inhibits cell growth and proliferation.

FIG. 7.
Reducing expression of PRMT5 impairs cancer cell growth. (A) To reduce PRMT5 protein expression, the transformed WaC3CD5 B-CLL cell line was infected with either recombinant lentivirus containing either vector alone or vector including AS-PRMT5 DNA. After ...

To confirm the AS-PRMT5 results and to further examine the role played by PRMT5 in regulating the RBL2 pathway, we used siRNA to knock down PRMT5 expression either alone or in combination with RBL2 and then measured protein expression, cell proliferation, the mRNA levels of target genes regulated by RBL2, and the gene-specific as well as global symmetric methylation of H3R8 and H4R3. The siRNA-mediated knock down of PRMT5 showed that reducing the expression of PRMT5 by 80 to 90% results in a 2.4- to 2.6-fold increase in RBL2 expression (Fig. (Fig.8A,8A, compare lanes 3 and 6 to lane 1). The derepression of RBL2 is accompanied by a twofold decrease in WaC3CD5 cell growth by day 4 (Fig. (Fig.8B,8B, compare black and white bars). When RBL2 protein expression was reduced by 70 to 80% in cells that have low levels of PRMT5 (Fig. (Fig.8A,8A, compare lanes 4 and 7 to lanes 3 and 6, respectively) using an siRNA that has been shown to efficiently knock down RBL2 expression (3), there was a clear rescue of wild-type WaC3CD5 cell growth (Fig. (Fig.8B,8B, compare the hatched and white bars at day 4). To investigate the impact of lowering PRMT5 protein levels on the growth-regulatory pathway controlled by RBL2, we measured the mRNA level of two RBL2 target genes, E2F1 and ribonucleotide reductase 2 (RR2) (10), in WaC3CD5 cells electroporated with either control siRNA, siPRMT5, or siPRMT5 and siRBL2 (Fig. (Fig.8C).8C). Analysis by real-time RT-PCR revealed that in the presence of siPRMT5, E2F1 was repressed 3.6-fold (P = 0.0002), and the expression of RR2 was decreased 2.1-fold (P < 10−4). Moreover, when RBL2 expression was knocked down in cells in which PRMT5 expression also was reduced, the expression of E2F1 and RR2 was restored to normal levels. These results are consistent with the rescue of cell growth observed when RBL2 levels were reduced in the presence of siPRMT5 (Fig. (Fig.8B,8B, compare the hatched and black bars to the white bar at day 4) and further solidify the role played by PRMT5 in regulating the RBL2 pathway.

FIG. 8.
siRNA-mediated knockdown of PRMT5 results in RBL2 derepression and the slow growth of transformed B cells. (A) Transformed WaC3CD5 B-CLL cells were treated either with control (Ctrl) siRNA, siPRMT5, or siPRMT5 combined with siRBL2, and 25 μg of ...

To assess the effect of siRNA-mediated PRMT5 knock down on the gene-specific and global symmetric methylation of H3R8 and H4R3, we conducted ChIP and immunofluorescence experiments (Fig. (Fig.9).9). When we monitored the PRMT5 recruitment and methylation of histones H3R8 and H4R3 at the RBL2 promoter, we found that knocking down the expression of PRMT5 resulted in a 12-fold (P = 0.0002) drop in PRMT5 recruitment, 6.8-fold (P = 0.001) decrease in H3R8 methylation, and 2.3-fold decline in H4R3 methylation (Fig. (Fig.9A).9A). Similarly, when we used immunofluorescence to examine the global symmetric methylation of H3R8 and H4R3 in wild-type WaC3CD5 cells electroporated with either control siRNA or siPRMT5, there was a significant decrease in PRMT5 protein expression and its epigenetic marks (Fig. (Fig.9B).9B). Collectively, these studies clearly show that PRMT5 is able to induce the growth and proliferation of WaC3CD5 cells by transcriptionally repressing RBL2 and that interference with PRMT5 protein levels alleviates RBL2 suppression, which in turn slows down WaC3CD5 cell growth.

FIG. 9.
siRNA-mediated knockdown of PRMT5 abolishes gene-specific and global H3R8 and H4R3 symmetric methylation. (A) ChIP analysis was used to assess the association of PRMT5 with the RBL2 promoter and the level of H3R8 and H4R3 methylation in WaC3CD5 cells ...


Recent studies have revealed that there is a myriad of epigenetic marks that can either induce or repress gene expression, and understanding how the vast array of histone modifications dictate the transcriptional outcome has become more clear. However, little progress has been made in clarifying the role some of these histone posttranslational marks play in disease. In this report, we show that the increased expression of PRMT5 is common among transformed lymphoid cell lines, and that the epigenetic modification of histones H3R8 and H4R3 is elevated in transformed B-CLL cells. We also show that the enhanced expression of PRMT5 in transformed WaC3CD5, Mec1, and Mec2 B-CLL cell lines is a direct result of the decreased expression of a select set of PRMT5-specific miRNAs, which in turn results in increased PRMT5 translation as measured by polyribosome profiling combined with real-time RT-PCR.

The efficiency of each miRNA to regulate PRMT5 expression was evaluated in two transformed B-CLL cell lines, and results from these studies indicate that while there is variation in their ability to knock down PRMT5 expression, they all interfere with its translation. We also provide evidence that clearly demonstrates that PRMT5 controls the growth and proliferation of transformed cells in part by epigenetically targeting the RB family of tumor suppressor genes. Our findings show that PRMT5 is enriched at the promoter of all RB family members, and this recruitment is accompanied by the increased symmetric methylation of H3R8 and H4R3 and the transcriptional repression of pocket protein transcripts. Although there is a perfect correlation between decreased RBL2 mRNA and protein expression, RB1 and RBL1 exhibit elevated protein levels, probably as a result of the decreased expression of specific miRNAs involved in their translational regulation. In fact, the RPA analysis of most RB1- and RBL1-specific miRNAs revealed that miR-649, which is specific for RB1 mRNA, and miR-22, which is specific for RBL1 mRNA, both are downregulated in transformed lymphoid cells (Fig. 5C and D). In complete agreement with our previous work using mantle cell lymphoma cell lines, reducing the expression of PRMT5 by either antisense RNA or siRNA derepresses RBL2 expression and inhibits the growth and proliferation of transformed WaC3CD5 B-CLL cells, highlighting the importance of PRMT5 and histone arginine methylation in the control of lymphoid cancer cell growth and proliferation.

Different miRNAs are required for proper PRMT5 translational regulation.

Recent genome-wide miRNA screens have provided great insight into various miRNonome programs that cells utilize to control the expression of specific target genes (1). Disparities in global miRNA profiles also have provided a means to differentiate normal cells from tumorigenic cells and have permitted the classification of different human cancers. More recently, the impact of specific miRNAs on certain growth-regulatory and developmental pathways has been solidified, and it is becoming evident that these small RNA molecules play a major role in controlling cell growth and differentiation (1, 5).

Our findings show that PRMT5 protein levels are augmented in transformed cells of lymphoid origin and establish a clear link between the altered expression of specific miRNAs and enhanced PRMT5 translation. In addition to miR-92b and miR-96, which previously have been shown to be repressed in JeKo and Mino MCL cell lines, we showed that the levels of four additional miRNAs also were altered in JeKo (MCL), Raji (Burkitt's lymphoma), and three different transformed B-CLL cell lines. However, miR-19a expression was not decreased in JeKo cells, suggesting that the repression of all six miRNAs is not required for the upregulation of PRMT5 translation. Consistently with this notion is the observation that not all PRMT5-specific miRNAs are repressed to the same extent, indicating that perhaps it is the misregulation of a combination of miRNAs that is responsible for altered PRMT5 translation. Furthermore, the observation that the reintroduction of a combination of miRNAs, miR-19a and miR-25, can bring about the derepression of RBL2 and partial inhibition of cell growth (data not shown) highlights the importance the identified miRNAs play in regulating PRMT5 expression.

The RB family of tumor suppressor genes is directly regulated by PRMT5 and is transcriptionally repressed in transformed lymphoid cells.

Pocket proteins are known for their ability to regulate cell cycle progression by targeting the E2F family of transcription factors (4, 43). Several studies have shown that pocket proteins achieve the transcriptional repression of E2F target genes by physically interacting with E2F transcription factors and blocking their transactivation potential, and/or by recruiting corepressors in association with chromatin remodelers such as mSIN3/HDAC-containing BRG1 and BRM complexes, DNMT1, and SUV39H1/HP1 (11, 24, 27, 33, 46). The release of the inhibitory effect of RB1 and its related family members is thought to be via a mechanism that involves phosphorylation by specific cyclins and cyclin-dependent kinases (8, 37). We have shown that the transcription of the RB family of tumor suppressors is repressed in transformed lymphoid cell lines and that there is a lack of correlation between mRNA and protein levels for RB1 and RBL1. However, this was not the case for RBL2, because low levels of RBL2 mRNA paralleled decreased levels of protein expression.

Previous work showed that both lymphoid and myeloid cell lines progress through the cell cycle in the presence of large amounts of wild-type RB1 and that RBL2 can block their proliferation (17). These results suggest that despite the existence of functional redundancy, RBL2 contributes more significantly to hematopoietic cell growth and proliferation. This is further supported by previous studies by Vairo et al. (38), who demonstrated that the RBL2-E2F4 complex inhibits T-cell growth via a mechanism distinct from that used by the RB1-E2F1 complex. Our studies using transformed lymphoid cell lines show that PRMT5 targets the RB family of tumor suppressor genes and triggers their transcriptional repression by inducing the symmetric methylation of promoter histones H3R8 and H4R3. However, despite the decreased levels of RB1 and RBL1 mRNA, transformed B cells proliferate in the presence of high levels of RB1 and RBL1 proteins, suggesting that the progression of these transformed lymphoid cells through the cell cycle is independent of RB1 and RBL1. In marked contrast, RBL2 mRNA and protein levels are reduced in these transformed B cells, suggesting that the PRMT5-induced epigenetic suppression of RBL2 is critical for transformed B-cell growth and proliferation.

These findings are in agreement with studies that show that even though pocket proteins exhibit functional redundancy, they play distinct and specific roles in the control of the cycle progression of cells of different origins. For instance, there are examples in which RB1 levels have been found to be elevated in colorectal carcinomas (45). It is believed that this increase in RB1 expression promotes tumor cell growth by antagonizing the proapoptotic effects of E2F1. Therefore, it is going to be important to examine the level of E2F proteins in transformed B cells to determine if there is any fluctuation in their expression, and if these changes coincide with the expression profile of pocket proteins. More particularly, measuring the level and association of E2F4 with RBL2 will provide insight into the contribution of this E2F family member to the control of G1-to-S transition in transformed lymphoid cells. This is especially important, since it appears that the ectopic expression of Bcl-2 can retard cell cycle progression by increasing the levels of RBL2, which, in association with E2F4, causes the downregulation of E2F1 and G1 arrest in 3T3 fibroblasts (39). In this study, Bcl-2 was shown to be incapable of reducing G1-to-S progression in cells that lacked RBL2, but it was able to slow cell cycle progression in the absence of RB1, suggesting that RBL2 plays an important role in Bcl-2-induced cell cycle withdrawal. These studies highlight the relevance of RBL2 in controlling the transition from quiescence to proliferation.

We have investigated whether the PRMT5-induced transcriptional repression of RB1, RBL1, and RBL2 is caused by changes in cell proliferation or whether these fluctuations in pocket protein mRNA expression are associated with transformation. Our studies clearly demonstrate that the transcription of RB1 is repressed in a cell cycle-dependent manner, as evidenced by decreased RB1 mRNA levels in proliferating B cells, whereas the PRMT5-mediated inhibition of RBL1 and RBL2 transcription appears to be associated with transformation (Fig. (Fig.6).6). Furthermore, it appears that despite decreased levels of RB1 and RBL1 mRNAs in transformed cells, protein expression either is unaltered in the case of RB1 or is enhanced in the case of RBL1, suggesting that there are posttranscriptional mechanisms involved in regulating RBL1 expression. In contrast, both RBL2 mRNA and protein were reduced significantly in transformed B cells, while in resting and activated B cells RBL2 mRNA and protein levels were unaffected. Therefore, it appears that the PRMT5-mediated epigenetic silencing of RBL2 is involved in promoting the growth and proliferation of transformed B cells.

Evidence in support of this conclusion comes from PRMT5 knockdown experiments in WaC3CD5 in which the decreased expression of PRMT5 results in the derepression of RBL2 and the inhibition of cell growth. It is known that E2F binding sites play an important role in regulating the expression of E2F1, whose activity has been well studied and shown to be critical for cells to transit from the G1 to S phase (12, 16). We have examined the levels of E2F1 mRNA, and one of its downstream targets, RR2, in cells in which PRMT5 and RBL2 expression has been knocked down. Our results show that when PRMT5 levels are reduced, RBL2 protein expression is enhanced and results in slow cell growth and the transcriptional suppression of E2F1 and RR2 (Fig. (Fig.77 and and8).8). Remarkably, when RBL2 expression is reduced in the presence of siPRMT5, the transcription of both E2F1 and RR2 is restored to normal levels, and the proliferation of WaC3CD5 B-CLL cells also is restored to wild-type levels, emphasizing the relevance of the epigenetic modification of RBL2 promoter histones H3R8 and H4R3 in the control of transformed B-cell growth and proliferation and argue that targeting PRMT5 might help restore the expression of key target genes and reduce cancer cell growth.

Supplementary Material

[Supplemental material]


We thank T. Motiwala and S. Jacob for providing B-CLL cell lines, L. Alinari and R. Baiocchi for help with B-cell isolation, and members of the laboratory for their help and critical reading of the manuscript.

This work was supported by National Cancer Institute grants CA116093 and CA101956 and American Cancer Society grant RSG-0418201-GMC to S.S.


[down-pointing small open triangle]Published ahead of print on 11 August 2008.

Supplemental material for this article may be found at http://mcb.asm.org/.


1. Bartel, D. P. 2004. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116281-297. [PubMed]
2. Bedford, M. T., and S. Richard. 2005. Arginine methylation an emerging regulator of protein function. Mol. Cell 18263-272. [PubMed]
3. Blais, A., C. J. van Oevelen, R. Margueron, D. Acosta-Alvear, and B. D. Dynlacht. 2007. Retinoblastoma tumor suppressor protein-dependent methylation of histone H3 lysine 27 is associated with irreversible cell cycle exit. J. Cell Biol. 1791399-1412. [PMC free article] [PubMed]
4. Brehm, A., and T. Kouzarides. 1999. Retinoblastoma protein meets chromatin. Trends Biochem. Sci. 24142-145. [PubMed]
5. Chapman, E. J., and J. C. Carrington. 2007. Specialization and evolution of endogenous small RNA pathways. Nat. Rev. Genet. 8884-896. [PubMed]
6. Chen, D., H. Ma, H. Hong, S. S. Koh, S. M. Huang, B. T. Schurter, D. W. Aswad, and M. R. Stallcup. 1999. Regulation of transcription by a protein methyltransferase. Science 2842174-2177. [PubMed]
7. Chiorazzi, N., and M. Ferrarini. 2003. B cell chronic lymphocytic leukemia: lessons learned from studies of the B cell antigen receptor. Annu. Rev. Immunol. 21841-894. [PubMed]
8. Classon, M., and E. Harlow. 2002. The retinoblastoma tumour suppressor in development and cancer. Nat. Rev. Cancer 2910-917. [PubMed]
9. Cook, J. R., J. H. Lee, Z. H. Yang, C. D. Krause, N. Herth, R. Hoffmann, and S. Pestka. 2006. FBXO11/PRMT9, a new protein arginine methyltransferase, symmetrically dimethylates arginine residues. Biochem. Biophys. Res. Commun. 342472-481. [PubMed]
10. DeGregori, J., G. Leone, A. Miron, L. Jakoi, and N. R. Nevins. 1997. Distinct roles for E2F proteins in cell growth control and apoptosis. Proc. Natl. Acad. Sci. USA 947245-7250. [PMC free article] [PubMed]
11. Dunaief, J. L., B. E. Strober, S. Guha, P. A. Khavari, K. Alin, J. Luban, M. Begemann, G. R. Crabtree, and S. P. Goff. 1994. The retinoblastoma protein and BRG1 form a complex and cooperate to induce cell cycle arrest. Cell 79119-130. [PubMed]
12. Dyson, N. 1998. The regulation of E2F by pRB-family proteins. Genes Dev. 122245-2262. [PubMed]
13. Fischle, W., Y. Wang, and C. D. Allis. 2003. Histone and chromatin cross-talk. Curr. Opin. Cell Biol. 15172-183. [PubMed]
14. Giacinti, C., and A. Giordano. 2006. RB and cell cycle progression. Oncogene 255220-5227. [PubMed]
15. Gonsalvez, G. B., L. Tian, J. K. Ospina, F. M. Boisvert, A. I. Lamond, and A. G. Matera. 2007. Two distinct arginine methyltransferases are required for biogenesis of Sm-class ribonucleoproteins. J. Cell Biol. 178733-740. [PMC free article] [PubMed]
16. Hake, S. B., A. Xiao, and C. D. Allis. 2004. Linking the epigenetic “language” of covalent histone modifications to cancer. Br. J. Cancer 90761-769. [PMC free article] [PubMed]
17. Harbour, J. W., and D. C. Dean. 2000. The Rb/E2F pathway: expanding roles and emerging paradigms. Genes Dev. 142393-2409. [PubMed]
18. Hoshikawa, Y., A. Mori, K. Amimoto, K. Iwabe, and M. Hatakeyama. 1998. Control of retinoblastoma protein-independent hematopoietic cell cycle by the pRB-related p130. Proc. Natl. Acad. Sci. USA 958574-8579. [PMC free article] [PubMed]
19. Hyllus, D., C. Stein, K. Schnabel, E. Schiltz, A. Imhof, Y. Dou, J. Hsieh, and U. M. Bauer. 2007. PRMT6-mediated methylation of R2 in histone H3 antagonizes H3 K4 trimethylation. Genes Dev. 213369-3380. [PMC free article] [PubMed]
20. Kleer, C. G., Q. Cao, S. Varambally, R. Shen, I. Ota, S. A. Tomlins, D. Ghosh, R. G. Sewalt, A. P. Otte, D. F. Hayes, M. S. Sabel, D. Livant, S. J. Weiss, M. A. Rubin, and A. M. Chinnaiyan. 2003. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc. Natl. Acad. Sci. USA 10011606-11611. [PMC free article] [PubMed]
21. Krause, C. D., Z. H. Yang, Y. S. Kim, J. H. Lee, J. R. Cook, and S. Pestka. 2007. Protein arginine methyltransferases: evolution and assessment of their pharmacological and therapeutic potential. Pharmacol. Ther. 11350-87. [PubMed]
22. Lee, J., J. Sayegh, J. Daniel, S. Clarke, and M. T. Bedford. 2005. PRMT8, a new membrane-bound tissue-specific member of the protein arginine methyltransferase family. J. Biol. Chem. 28032890-32896. [PubMed]
23. Le Guezennec, X. L., M. Vermeulen, A. B. Brinkman, W. A. Hoeijmakers, A. Cohen, E. Lasonder, and H. G. Stunnenberg. 2006. MBD2/NuRD and MBD3/NuRD, two distinct complexes with different biochemical and functional properties. Mol. Cell. Biol. 26843-851. [PMC free article] [PubMed]
24. Luo, R. X., A. A. Postigo, and D. C. Dean. 1998. Rb interacts with histone deacetylase to repress transcription. Cell 92463-473. [PubMed]
25. Martin, C., and Y. Zhang. 2005. The diverse functions of histone lysine methylation. Nat. Rev. Mol. Cell Biol. 6838-849. [PubMed]
26. Miao, F., S. Li, V. Chavez, L. Lanting, and R. Natarajan. 2006. Coactivator-associated arginine methyltransferase-1 enhances nuclear factor-κB-mediated gene transcription through methylation of histone H3 at arginine 17. Mol. Endocrinol. 201562-1573. [PubMed]
27. Nielsen, S. J., R. Schneider, U. M. Bauer, A. J. Bannister, A. Morrison, D. O'Carroll, R. Firestein, M. Cleary, T. Jenuwein, R. E. Herrera, and T. Kouzarides. 2001. Rb targets histone H3 methylation and HP1 to promoters. Nature 412561-565. [PubMed]
28. Pal, S., R. Yun, A. Datta, L. Lacomis, H. Erdjument-Bromage, J. Kumar, P. Tempst, and S. Sif. 2003. mSin3A/histone deacetylase 2- and PRMT5-containing Brg1 complex is involved in transcriptional repression of the Myc target gene cad. Mol. Cell. Biol. 237475-7487. [PMC free article] [PubMed]
29. Pal, S., S. N. Vishwanath, H. Erdjument-Bromage, P. Tempst, and S. Sif. 2004. Human SWI/SNF-associated PRMT5 methylates histone H3 arginine 8 and negatively regulates expression of ST7 and NM23 tumor suppressor genes. Mol. Cell. Biol. 249630-9645. [PMC free article] [PubMed]
30. Pal, S., R. A. Baiocchi, J. C. Byrd, M. R. Grever, S. T. Jacob, and S. Sif. 2007. Low levels of miR-92b/96 induce PRMT5 translation and H3R8/H4R3 methylation in mantle cell lymphoma. EMBO J. 263558-3569. [PMC free article] [PubMed]
31. Pal, S., and S. Sif. 2007. Interplay between chromatin remodelers and protein arginine methyltransferases. J. Cell. Physiol. 213306-315. [PubMed]
32. Raman, J. D., N. P. Mongan, S. K. Tickoo, S. A. Boorjian, D. S. Scherr, and L. J. Gudas. 2005. Increased expression of the polycomb group gene, EZH2, in transitional cell carcinoma of the bladder. Clin. Cancer Res. 118570-8576. [PubMed]
33. Robertson, K. D., S. Ait-Si-Ali, T. Yokochi, P. A. Wade, P. L. Jones, and A. P. Wolffe. 2000. DNMT1 forms a complex with Rb, E2F1 and HDAC1 and represses transcription from E2F-responsive promoters. Nat. Genet. 25338-342. [PubMed]
34. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
35. Strahl, B. D., S. D. Briggs, C. J. Brame, J. A. Caldwell, S. S. Koh, H. Ma, R. G. Cook, J. Shabanowitz, D. F. Hunt, M. R. Stallcup, and C. D. Allis. 2001. Methylation of histone H4 at arginine 3 occurs in vivo and is mediated by the nuclear receptor coactivator PRMT1. Curr. Biol. 11996-1000. [PubMed]
36. Tan, C. P., and S. Nakielny. 2006. Control of the DNA methylation system component MBD2 by protein arginine methylation. Mol. Cell. Biol. 267224-7235. [PMC free article] [PubMed]
37. Taya, Y. 1997. RB kinases and RB-binding proteins: new points of view. Trends Biochem. Sci. 2214-17. [PubMed]
38. Vairo, G., D. M. Livingston, and D. Ginsberg. 1995. Functional interaction between E2F-4 and p130: evidence for distinct mechanisms underlying growth suppression by different retinoblastoma protein family members. Genes Dev. 9869-881. [PubMed]
39. Vairo, G., T. J. Soos, T. M. Upton, J. Zalvide, J. A. DeCaprio, M. E. Ewen, A. Koff, and J. M. Adams. 2000. Bcl-2 retards cell cycle entry through p27(Kip1), pRB relative p130, and altered E2F regulation. Mol. Cell. Biol. 204745-4753. [PMC free article] [PubMed]
40. Varambally, S., S. M. Dhanasekaran, M. Zhou, T. R. Barrette, C. Kumar-Sinha, M. G. Sanda, D. Ghosh, K. J. Pienta, R. G. Sewalt, A. P. Otte, M. A. Rubin, and A. M. Chinnaiyan. 2002. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419624-629. [PubMed]
41. Visser, H. P., M. J. Gunster, H. C. Kluin-Nelemans, E. M. Manders, F. M. Raaphorst, C. J. Meijer, R. Willemze, and A. P. Otte. 2001. The Polycomb group protein EZH2 is upregulated in proliferating, cultured human mantle cell lymphoma. Br. J. Haematol. 112950-958. [PubMed]
42. Wang, L., R. A. Baiocchi, S. Pal, G. Mosialos, M. Caligiuri, and S. Sif. 2005. The BRG1- and hBRM-associated factor BAF57 induces apoptosis by stimulating expression of the cylindromatosis tumor suppressor gene. Mol. Cell. Biol. 257953-7965. [PMC free article] [PubMed]
43. Weinberg, R. A. 1995. The retinoblastoma protein and cell cycle control. Cell 81323-330. [PubMed]
44. Xu, W., H. Cho, S. Kadam, E. M. Banayo, S. Anderson, J. R. Yates III, B. M. Emerson, and R. M. Evans. 2004. A methylation-mediator complex in hormone signaling. Genes Dev. 18144-156. [PMC free article] [PubMed]
45. Yamamoto, H., J. W. Soh, T. Monden, M. G. Klein, L. M. Zhang, H. Shirin, N. Arber, N. Tomita, I. Schieren, C. A. Stein, and I. B. Weinstein. 1999. Paradoxical increase in retinoblastoma protein in colorectal carcinomas may protect cells from apoptosis. Clin. Cancer Res. 51805-1815. [PubMed]
46. Zhang, H. S., M. Gavin, A. Dahiya, A. A. Postigo, D. Ma, R. X. Luo, J. W. Harbour, and D. C. Dean. 2000. Exit from G1 and S phase of the cell cycle is regulated by repressor complexes containing HDAC-Rb-hSWI/SNF and Rb-hSWI/SNF. Cell 10179-89. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...