HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 26, 19866-19876, June 30, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19866    most recent M000023200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Tang, J. Articles by Herschman, H. R. Search for Related Content PubMed PubMed Citation Articles by Tang, J. Articles by Herschman, H. R. Social Bookmarking                      What's this?

Protein-arginine Methyltransferase I, the Predominant Protein-arginine Methyltransferase in Cells, Interacts with and Is Regulated by Interleukin Enhancer-binding Factor 3^*

Jie Tang^§¶, Peter N. Kao, and Harvey R. Herschman^§**

From the  Molecular Biology Institute and the ^§ Department of Biological Chemistry, UCLA School of Medicine, Los Angeles, California 90095 and the  Division of Pulmonary and Critical Care Medicine, Stanford University Medical Center, Stanford, California 94305

Received for publication, January 5, 2000, and in revised form, March 27, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Arginine methylation is a common post-translation modification found in many proteins. Protein-arginine methyltransferase I (PRMT1) contributes >90% of type I protein-arginine methyltransferase activity in cells and tissues. To expand our knowledge on the regulation and role of PRMT1 in cells, we used the yeast two-hybrid system to identify proteins that interact with PRMT1. One of the interacting proteins we cloned is interleukin enhancer-binding factor 3 (ILF3), also known as M phase phosphoprotein 4. ILF3 is closely related to nuclear factor 90 (NF90). Using an immunofluorescence analysis, we determined that ILF3 and PRMT1 co-localize in the nucleus. Moreover, PRMT1 and ILF3 co-precipitate in immunoprecipitation assays and can be isolated together in "pull-down" experiments using recombinant fusion proteins. ILF3 is a robust substrate for methylation by PRMT1 and can modulate PRMT1 activity in in vitro methylation assays. Deletion studies demonstrated that the COOH-terminal region of ILF3, which is rich in arginine, glycine, and serine, is responsible for the strong interaction between PRMT1 and ILF3 and is the site of ILF3 methylation by PRMT1. Although ILF3 and NF90 are highly similar, they differ in their carboxyl-terminal regions. Because of this difference, NF90 does not interact with PRMT1, is a much poorer substrate than ILF3 for PRMT1-dependent methylation, and does not modulate PRMT1 enzyme activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Arginine methylation is a common post-translation protein modification in eucaryotic cells, discovered more than 30 years ago (1, 2). There are at least three types of protein-arginine N-methyltransferase activities that transfer methyl groups from S-adenosyl-L-methionine to the guanidino group of arginine residues (3). The formation of monomethylarginine and asymmetric N^G,N^G-dimethylarginine is catalyzed by type I enzymes. Substrates for type I protein-arginine methyltransferases include many RNA-binding proteins (4), RNA-transporting proteins (4), transcription factors (3, 5), nuclear matrix proteins (3), and cytokines (6). Type I arginine protein methylation modifies the activities of transcription factors (3), modulates the affinity of nucleic acid-binding proteins for nucleic acids (3), regulates interferon signaling pathways (7), and alters targeting of nuclear proteins (8). Type II protein-arginine methyltransferases catalyze the formation of both monomethylarginine and symmetric N^G,N'^G-dimethylarginine (3). Myelin basic protein is the only known substrate for type II arginine methyltransferase activity. Type III protein-arginine methyltransferase catalyzes the monomethylation of the internal guanidino nitrogen atom to form -N^G-monomethylarginine (9).

Protein-arginine methyltransferase I (PRMT1)¹ was the first type I protein-arginine N-methyltransferase in mammalian cells to be cloned and characterized (10). PRMT1 is the predominant type I enzyme in tissues and contributes most of the type I protein-arginine methyltransferase activity in mammalian cells (11). The calculated molecular mass of the PRMT1 polypeptide, based on the cDNA-derived protein sequence, is 40.5 kDa (10). However, PRMT1 in cells exists as a large complex of 300-400 kDa (10, 12). PRMT1 can interact with a variety of proteins, including TIS21 (10), BTG1 (10), and interferon-/ receptor 1 (7). TIS21 and the related protein BTG1 are members of a family of proteins thought to be involved in negative regulation of cell growth (13, 14). Both TIS21 and BTG1 interact with and regulate the activity of PRMT1 (10). The interaction between PRMT1 and the cytoplasmic domain of interferon-/ receptors is proposed to be important to the interferon signaling pathways.

We performed a yeast two-hybrid interaction screen for PRMT1-interacting proteins to identify proteins that might associate with PRMT1 in cells and to study the regulation of PRMT1. This screen identified interleukin enhancer-binding factor 3 (ILF3) as a protein that could potentially interact with PRMT1. Interleukin enhancer-binding factor proteins belong to a group of proteins that can bind to the antigen receptor response element 2 (ARRE-2) sequence (15). This group of proteins includes ILF1 (15), nuclear factor 45 (NF45, also referred as ILF2) (16, 17), and nuclear factor 90 (NF90) (18). ILF3, a protein with an apparent molecular mass of ~110 kDa (19), is closely related in sequence to NF90. In fact, NF90 and ILF3 have both been referred to as ILF3 in different reports (16, 18, 19). However, ILF3 and NF90 are distinct protein species (17, 19, 20) whose relationships have not been clearly addressed previously. In this report, the 110-kDa protein is referred to as ILF3, and the 90-kDa protein is termed NF90.

Mouse ILF3 is encoded by an open reading frame of 911 amino acid residues (19). The cDNA of murine ILF3 is encoded by a single copy gene located on chromosome 9 near D9Mit160 (19). ILF3 is cyclically hyperphosphorylated in M phase during mitosis and is also referred to as M phase phosphoprotein 4 (MPP4) (21). ILF3 is closely related in sequence to the Xenopus p122 protein, a double-stranded RNA-binding protein that is also a component of a CCAAT box transcription factor (22), and to human NF90 (17). NF90 is virtually identical to ILF3 in the first 600 amino acid residues (17, 19). Like ILF3, NF90 contains two double-stranded RNA-binding motifs and binds double- and single-stranded RNAs (23). NF90 and NF45 are purified as a dimeric complex that binds specifically to ARRE-2 (20), an important cis-acting element found in many cytokine genes. Both NF90 and NF45 migrate from the nucleus to the cytoplasm in mitotic cells (24). NF90 appears to be involved in transcriptional regulation of promoters that contain nuclear factor of activated T cells-binding sites (17, 20). In cells, NF90 can interact with a variety of proteins, including double-stranded RNA-dependent protein kinase (25) and DNA-dependent protein kinase (26). NF90 can be phosphorylated by the latter two proteins in vitro (25, 26). NF90 and ILF3 have been proposed as factors in the regulation of transcription, DNA repair, and cell cycle controls (21, 22, 24, 25, 26).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Constructions-- The yeast two-hybrid bait plasmid pLexA(L)-PRMT1 has been described previously (12). cDNA clones that encode proteins interacting with PRMT1 were identified in yeast two-hybrid interactions from the pPC86 cDNA library prepared from FAO hepatoma cell poly(A⁺) mRNA (12). The full-length rat ILF3 cDNA in pBluescript, which can be released by single EcoRI digestion, was recovered by screening a rat PC12 cDNA library as described previously (12) using the cDNA insert from pPC86-40A as the probe. pGEX(SN)-40A (ILF3 Lys³⁵⁵-Arg⁹¹⁰) was constructed by inserting the 40A SalI/NotI fragment from pPC86-40A into pGEX(SN). pGEX(SN)-ILF3 was constructed by (i) PCR subcloning the 5'-half of the ILF3 cDNA with a 5'-primer (5'GGGTGGGTCGACCATGGCATTGTATCATCATC) and a 3'-primer (5'-CAGGGAAAAGCTTCTCTAATG) using pBluescript-ILF3 as the template and (ii) cutting the fragment with SalI and HindIII and then inserting the fragment into pGEX(SN)-40A at the SalI and HindIII sites. To construct the carboxyl-terminal deletion of ILF3, an ILF3 cDNA fragment was amplified by PCR using a 5'-primer (5'CGATGTTTAATACCACT) and a 3'-primer (5'TACTTAGCGGCCGCTCAGACAGGCACAGGGGCTCT) with pPC86-40A as the template. The amplified fragment was digested with SalI/NotI and then cloned into pGEX(SN) to generate pGEX(SN)-40AC. The SalI/HindIII fragment of the ILF3 cDNA from pGEX(SN)-ILF3 was then cloned into pGEX(SN)-40AC at the SalI and HindIII sites to generate pGEX(SN)-ILF3C, which encodes the carboxyl-terminal deletion of ILF3 from Met¹ to Val⁶²¹. pLexA(L)-NF90 was constructed by inserting the NF90 cDNA into pLexA117. pLexA117 was derived from pLexA(L) (12) by replacing the SphI cassette in pLexA(L) with the SphI cassette from pBTM117 (27). The NF90 cDNA was subcloned from pBluescript-NF90 (10) by PCR using a 5'-primer (5'CCCGAATTCCATGCGTCCAATGCGAATTTTTGTG) and a 3'-primer (5'TACTTAGCGGCCGCCAGCACCGGCATTCATGTAGCCTC) to create EcoRI and NotI sites at the 5'- and 3'-ends. The amplified fragment was inserted into pLexA117 at the EcoRI and NotI sites to create pLexA-NF90. pLexA(L)-PRMT3 was generated by inserting the SalI/NotI fragment of PRMT3-(214-528) cDNA from pBTM117-PRMT3-(214-528) into pLexA-117. pGEX4T3-NF90, which encodes a GST-NF90 fusion protein, was constructed by cloning the EcoRI/NotI fragment from pLexA(L)-NF90 into pGEX4T3. The yeast two-hybrid prey plasmids pPC86-ILF3 and pPC86-ILF3C were constructed by inserting the SalI/NotI fragments from pGEX(SN)-ILF3 and pGEX(SN)-ILF3C into pPC86, respectively. pcDNA3.1()-ILF3 was constructed by inserting the XhoI/BamHI fragment of ILF3 cDNA from pBluescript-ILF3 into pcDNA3.1() (Invitrogen). pcDNA3.1(+)-ILF3C was constructed by subcloning the SalI/NotI fragment from pGEX(SN)-ILF3C into pBluescript to create a KpnI site at the 5-end and then inserting the KpnI/NotI fragment from pBluescript-ILF3C into pcDNA3.1(+). pcDNA3.1()-NF90 was constructed by inserting the EcoRI fragment of NF90 cDNA from pBluescript-NF90 into pcDNA3.1(). Sequences of the cDNA fragments amplified by PCR were confirmed by DNA sequencing.

Biotinylation of Anti-PRMT1 IgG-- IgG proteins from anti-PRMT1 antiserum were purified on a protein A-Sepharose 4B column. Briefly, 1 ml of antiserum was bound to a 1.5-ml protein A-Sepharose 4B column that had been equilibrated with phosphate-buffered saline (PBS). The column was extensively washed with PBS until no protein was detectable in the eluate. The absorbed IgG proteins were eluted with 250 mM glycine (pH 2.5). The eluted proteins were neutralized in 2 M Tris-HCl buffer (pH 7.6) and then desalted on a Sephadex G-25 column using 100 mM bicarbonate buffer (pH 8.0).

To biotinylate anti-PRMT1 IgG protein, purified IgG (1 mg/ml) in 100 mM bicarbonate buffer (pH 8.0) was incubated with 40 µl of 10 mM sulfosuccinimidobiotin for 4 h at room temperature. Then, IgG proteins were separated from sulfosuccinimidobiotin on a Sephadex G-25 column using PBS to elute proteins from the column. IgG proteins were shown to be biotinylated by immunoblotting Rat1 cell lysate with the biotinylated IgG as the first antibody and using horseradish peroxidase-labeled streptavidin as the secondary antibody on Western blots.

Immunofluorescent Labeling of ILF3/NF90 and PRMT1-- Immunolocalization of proteins in cells was performed as described previously (12). To double-stain HeLa cells with anti-PRMT1 and anti-ILF3/NF90 antibodies, HeLa cells in four-well chamber slides were washed twice with PBS and fixed with 3% paraformaldehyde for 5 min, followed by 10 min of incubation in 3% paraformaldehyde containing 0.1% Triton X-100. Cells were first stained with anti-NF90 antibody (1:150) using fluorescein isothiocyanate-labeled anti-rabbit IgG as the secondary antibody and then were blocked again with goat serum. The cells were subsequently stained with biotinylated anti-PRMT1 IgG and Texas Red-labeled streptavidin. The stained cells were mounted in fluoromount and examined with both a Zeiss fluorescence microscope and a Bio-Rad confocal microscope.

Preparation of GST Fusion Proteins-- All GST fusion proteins (Amersham Pharmacia Biotech) were expressed and purified as described by the manufacturer with modifications. Briefly, Escherichia coli cells transformed with GST fusion plasmids were grown overnight to saturation in 2× YT medium. Then, the culture was diluted with fresh 2× YT medium and induced with isopropyl--D-thiogalactopyranoside as suggested by Amersham Pharmacia Biotech. GST fusion proteins were purified from E. coli as described previously (12). GST fusion proteins were separated from glutathione on a Sephadex G-25 column using PBS to elute proteins.

Yeast Two-hybrid Analysis of Protein Interactions-- Yeast two-hybrid analysis was performed as described previously (12). Bait plasmid pLexA-PRMT1, constructed in our previous study (12), was used to screen the pPC86 cDNA library from the rat FAO hepatoma cell poly(A⁺) mRNA. Several clones that interact with PRMT1 in both histidine and -galactosidase assays were identified. One of these clones is 40A. A homology search of GenBank^TM using the BLAST algorithm revealed that 40A encodes a carboxyl-terminal fragment of ILF3.

To test the interactions between NF90 and ILF3 and between PRMT1 and ILF3/NF90, bait plasmids and prey plasmids were transformed into L40 yeast, and transformed yeast cells were plated on appropriate selective media and incubated at 30 °C. Individual colonies were patched and assayed for -galactosidase activity as described previously (12).

Northern Analysis of RNA from HeLa Cells-- Total RNA from exponentially growing HeLa cells was prepared using the Rneasy total RNA purification kit from QIAGEN Inc. RNA was subjected to electrophoresis and Northern blotting as described previously (10, 12).

Protein Concentration Determinations-- Protein concentration was determined by the bicinchoninic acid assay (Pierce). Bovine serum albumin was used as the standard.

SDS-PAGE and Western Blot Analysis-- Protein samples were subjected to SDS-PAGE and immunoblot analysis or silver staining as described previously (12).

In Vitro Protein-arginine Methyltransferase Assay-- Methyltransferase activity was assayed at 37 °C for times and in final volumes as specified in each figure legend. The methyl-donating substrate was S-adenosyl-L-[methyl-³H]methionine ([³H]AdoMet; specific activity = 75 Ci/mmol; NEN Life Science Products). The methyl-accepting substrates were GST-GAR (12) and hypomethylated, adenosine dialdehyde-treated Rat1 cell lysate.² Methylation reactions were stopped by adding SDS-PAGE sample buffer and resolved by SDS-PAGE. The methylated proteins were visualized by fluorography as described previously (12).

Preparation of Hypomethylated Rat1 Cell Lysates-- Rat1 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum as described previously (10). Exponentially growing Rat1 cells were treated with 20 µM adenosine dialdehyde for 2 days. Cells were washed twice with PBS and harvested in PBS with protease inhibitors (Roche Molecular Biochemicals catalog no. 1836170). Cells were lysed by brief low power sonication, and the cell lysate was subjected to 5 min of centrifugation. The supernatant fraction was collected as the adenosine dialdehyde-treated Rat1 cell lysate (10).² To inactivate the endogenous protein-arginine methyltransferases, adenosine dialdehyde-treated Rat1 cell lysate was incubated at 70 °C for 6 min.

In Vitro Translation of PRMT1, NF90, and ILF3-- [³⁵S]Methionine-labeled PRMT1, ILF3, and NF90 were generated by coupled in vitro transcription and translation from pcDNA3.1-PRMT1, -ILF3, -ILF3C, and -NF90 using the TNT T7 coupled reticulum lysate system (Promega) according to procedures described previously (10).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Identification of ILF3 as a Protein That Can Interact with PRMT1-- A yeast strain expressing the LexA-PRMT1 fusion protein was transformed with a rat liver FAO cell cDNA library containing cDNAs fused to the Gal4 transcription activation domain. One of the most frequently obtained clones, 40A, encodes the COOH-terminal two-thirds of rat ILF3. An ILF3 cDNA that covers the entire coding region² was cloned by screening a rat PC12 cDNA library (12) using clone 40A as a probe. Because of a 3-base pair deletion, the rat ILF3 protein is a single amino acid shorter than the murine ILF3 protein (Fig. 1). Both ILF3 and NF90 contain nuclear localization sequences and two double-stranded RNA-binding motifs (Fig. 1). Rat ILF3, mouse ILF3, human ILF3, and human NF90 are strikingly similar in amino acid sequence (Fig. 1). However, in the COOH-terminal region, only ILF3 contains a domain that includes many asymmetric arginine dimethylation consensus sequences (RXR and RGG). In contrast, NF90 is altered in several of these consensus sequences and is prematurely terminated when compared with ILF3. Human ILF3 is also referred to as MPP4 (21).

View larger version (92K): [in this window] [in a new window]   Fig. 1.   Sequence alignments of rat ILF3, mouse ILF3 (AF098967), human ILF3 (AF167570), and human NF90 (NM_004516). The amino acid sequence of rat ILF3 is derived from the rat ILF3 cDNA, which was cloned and sequenced as described under "Experimental Procedures." Reexamination of the sequence of human NF90 has resulted in the corrected sequence shown above (P. N. Kao, unpublished data). Amino acid alignment was performed using CLUSTAL W. Conserved amino acids are indicated by dark shading for identical residues and by light shading for similar residues. Box I represents the bipartite nuclear localization signal. Boxes II and III represent the two double-stranded RNA-binding sequence motifs. The first arrow indicates the start of clone 40A. The second arrow indicates the last amino acid residue in the carboxyl-terminal truncated ILF3C molecule. Sequences in boldface are putative type I arginine methylation sequences that can potentially be methylated by type I protein-arginine methyltransferases.

Interleukin Enhancer-binding Factor 3 and Nuclear Factor 90 Are Present in the Same Cell-- NF90 and ILF3 have virtually identical sequences in their overlapping regions (Fig. 1). In mouse, there is only one gene encoding ILF3 (19). However, a BLAST search (28) identified mouse expressed sequence tag cDNA clones encoding fragments for both NF90 (GenBank^TM accession number AA049057) and ILF3 (GenBank^TM accession number W81847). To determine whether cells contain multiple mRNA species for ILF3 and NF90, total RNA from HeLa cells was isolated and analyzed by Northern blotting (Fig. 2A). An NF90-specific probe detected one mRNA species, whereas a probe common to both NF90 and ILF3/MPP4 detected both the NF90 message and a weaker mRNA signal. These data suggest that there are multiple mRNAs derived from the ILF3 gene. To determine whether clonal cell populations express ILF3 and NF90 proteins, HeLa lysates were subjected to SDS-PAGE and Western blotting with an antibody to NF90 (17) (Fig. 2B). As previously reported (21), two bands were observed. The 110-kDa band corresponds to the molecular mass of ILF3; the 90-kDa band corresponds to the molecular mass of NF90.

View larger version (43K): [in this window] [in a new window]   Fig. 2.   ILF3 and NF90 both exist in human HeLa cells. A, Northern analysis of ILF3 and NF90 expression using either an NF90-specific probe or a probe common to both ILF3 and NF90. The template for the NF90-specific probe was generated by PCR amplification of nucleotides 1-238 using 5'-pCGCCGCCTGCCCGCCCGCCCGCTC and 5'-pTTTTGCTTATCTTTCAAATTGTGATAGAG as primers and pBluescript-NF90 as the template. The -32P-radiolabeled NF90-specific probe was generated by transcribing the antisense strand from the probe template using 5'-pTTTTGCTTATCTTTCAAATTGTGATAGAG as the primer and Klenow fragment to catalyze the reaction. Template for the probe common to both ILF3 and NF90 was PCR-amplified from pBluescript-ILF3 with 5'-pCTCCAATGCGAATTTTTGTGAATG and 5'-GGGATGGCCAGGTTGTCGGCC as primers. The -32P-radiolabeled probe for ILF3 and NF90 was generated by a random hexamer labeling reaction of the probe template. 10 µg of total RNA purified from HeLa cells was resolved on agarose gel and probed as described previously (12). B, Western blot of ILF3 and NF90 in HeLa cells. 50 µg of HeLa cell lysate was subjected to electrophoresis on 8% SDS-polyacrylamide gel; transferred to a nitrocellulose membrane; and probed with anti-NF90 antiserum, which recognizes both ILF3 and NF90. The 110-kDa band corresponds to the molecular mass of ILF3, and the 90-kDa band corresponds to the molecular mass of NF90.

ILF3/MPP4 is known to undergo cyclic phosphorylation (21). The difference between the 110- and 90-kDa bands is unlikely to result from different phosphorylation states of the same protein because even when the lysate is treated with phosphatase, the difference between these two bands still exists (21). Our results and the results of others thus suggest that ILF3 and NF90 can exist in the same cell.

ILF3 and PRMT1 Associate in Yeast Two-hybrid Interaction Analysis-- The interaction of ILF3 and PRMT1 was identified through a yeast two-hybrid screen. In this screen, the 40A clone extends from Lys³⁵⁵ to Arg⁹¹⁰. 40A contains a region rich in the RGG and RXR sequences that are potential substrates for PRMT1 (3, 29). To determine the importance of the RGG/RXR region in the interaction of ILF3 and PRMT1, we generated a deletion construct (40AC) in which the cDNA encoding the ILF3 Lys³⁵⁵-Val⁶²¹ fragment was fused to the Gal4 transcription activation domain. pLexA-PRMT1 was transformed into yeast along with different prey plasmids (Fig. 3A). As expected, clone 40A (ILF3 Lys³⁵⁵-Arg⁹¹⁰) formed a strong complex with PRMT1. 40A did not interact with PRMT3. When the RGG/RXR-rich COOH-terminal domain (Arg⁶²²-Arg⁹¹⁰) of 40A was deleted, the interaction between 40AC (ILF3 Lys³⁵⁵-Val⁶²¹) and PRMT1 was dramatically decreased.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Yeast two-hybrid analysis of interactions between PRMT1 and ILF3 and between ILF3 and NF90. A, pLexA-PRMT1 and pLexA-PRMT3 are two-hybrid bait plasmids that express PRMT1 and PRMT3 fused to the LexA DNA-binding domain, respectively. Yeast (L40) cells were transformed with pLexA-PRMT1 along with prey plasmids pPC86-40A, pPC86-40AC, and pPC86-PRMT3 (as a positive control) (12). These latter plasmids express fusion proteins between the Gal4 transcription activation domain (Gal4AD) and the corresponding proteins. Interaction between PRMT3 and 40A was analyzed by a similar transformation. pPC86 empty vector was included as a control for nonspecific interactions. B, to analyze interactions between ILF3 domains and NF90, yeast cells were transfected with pLexA-NF90 and pPC86-40A. As a non-interacting control, plasmids pLexA-NF90 and pPC86-BTG1 were transformed into yeast cells. Transformed yeast cells were plated on synthetic complete minimal-Ura/Lys/Trp/Leu plates and incubated at 30 °C. Individual colonies were patched and assayed for -galactosidase activity using a yeast colony filter assay. Blue color develops when the bait prey fusion proteins physically interact in the yeast nucleus and activate the transcription of a lacZ gene under the control of LexA-responsive elements.

NF90 and ILF3 Associate in Yeast Two-hybrid Interaction Analysis-- To determine whether full-length NF90 and ILF3 can form a complex in yeast, we generated pLexA-NF90 and transformed both this plasmid and pPC86-40A (i.e. pPC86-ILF3 Lys³⁵⁵-Arg⁹¹⁰) into yeast cells (Fig. 3B). The positive LacZ expression demonstrates that ILF3 and NF90 can form a complex in yeast two-hybrid interaction analysis.

Endogenous PRMT1 and ILF3 Co-localize and Immunoprecipitate Together in Mammalian Cells-- The ILF3/PRMT1 interaction was first demonstrated in the yeast-two hybrid assay (Fig. 2). To test whether PRMT1 and ILF3 interact in mammalian cells, we first examined their co-localization. HeLa cells were stained first with rabbit anti-ILF3/NF90 antibody (17) and fluorescein isothiocyanate-conjugated anti-rabbit IgG and then stained with biotinylated anti-PRMT1 antibody (12) and Texas Red-conjugated streptavidin. ILF3/NF90 stains green, and PRMT1 stains red (Fig. 4A). When ILF3 and PRMT1 co-localize in cells, the co-localized compartments will be yellow. PRMT1 and ILF3 co-localize in the nucleus. Neither antigen is present in the nuclear speckles.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   PRMT1 and ILF3/NF90 interaction in mammalian cells. A, PRMT1 and ILF3/NF90 co-localize in the nucleus of HeLa cells. HeLa cells were cultured in four-well chamber slides and stained first with anti-ILF3/NF90 antibody (green color) and then anti-PRMT1 antibody (red color) as described under "Experimental Procedures." Stained cells were examined with a Bio-Rad confocal microscope. When cells were stained only with anti-ILF3/NF90 antibody, only green staining was detected (not shown). When cells were stained only with anti-PRMT1 antibody, only red staining was detected (not shown). When cells were stained with both anti-PRMT1 and anti-ILF3/NF90 antibodies, the ILF/NF90 staining was green (right panel), and the PRMT1 staining was red (center panel); and co-staining of these proteins in cells is the combination of red and green, resulting in yellow (right panel). B, PRMT1 enzyme activity co-immunoprecipitates with ILF3/NF90. HeLa cell lysates were first precleared with protein A-Sepharose 4B at a lysate/protein A bead volume ratio of 10:1 for 30 min at 4 °C. 50 µg of the precleared HeLa lysate was incubated with 2 µl of the various antibodies for 30 min at 4 °C. Then, 30 µl of protein A-Sepharose 4B beads (50% suspension in PBS) was added to the reaction mixture, and the reaction mixture was incubated for an additional 90 min at 4 °C. The protein A-Sepharose beads were recovered and washed with PBS containing 0.5% Triton X-100, 0.5% Tween 20, and 0.05% SDS three times (10 min each). PRMT1 activity associated with the protein A-Sepharose beads was assayed at 37 °C for 90 min in 60 µl of reaction mixture containing 2 µg of the methyl-accepting substrate GST-GAR and 4.4 µCi of the methyl-donating substrate [3H]AdoMet. The reaction mixture was then subjected to SDS-PAGE, and the methylated GST-GAR was visualized by exposing the gel to film at 80 °C for 3 days. Lane 1 is a positive control in which proteins were immunoprecipitated with anti-PABII antibody. Lane 2 is a negative control immunoprecipitated by preimmune serum. Lane 3 shows the GAR methyltransferase activity associated with proteins co-immunoprecipitated by the anti-ILF3/NF90 antibody. C, GST-PRMT1 pull-down of ILF3/NF90 from a HeLa cell lysate. Lane 1 shows the ILF3 and NF90 signal from 10 µg of HeLa cell lysate immunoblotted with anti-ILF3/NF90 antibody. 50 µg of this HeLa lysate was diluted to 500 µl with PBS and incubated with 3 µg of GST protein (lane 2) or GST-PRMT1 (lane 3) for 30 min at 4 °C. Then, 30 µl of glutathione-Sepharose beads (50% suspension) was added, and the reaction mixture was incubated at 4 °C for an additional 90 min. The glutathione-Sepharose beads were recovered by centrifugation and washed three times with PBS containing 0.5% Triton X-100, 0.5% Tween 20, and 0.1% SDS. Proteins associated with the glutathione-Sepharose beads were subjected to SDS-PAGE and immunoblotting with anti-ILF3/NF90 antibody.

To demonstrate an intracellular interaction between ILF3/NF90 and PRMT1, we carried out a co-immunoprecipitation reaction. HeLa cell lysates were subjected to immunoprecipitation reactions with preimmune serum, antibody to PABII (29), or antibody to ILF3/NF90 (17). The immunoprecipitates were then assayed for protein-arginine methyltransferase activity using GST-GAR as substrate (12). Poly(A)-binding protein II, the positive control, has previously been shown to interact with PRMT1 (30). Antiserum to ILF3/NF90 also coprecipitated protein-arginine methyltransferase activity (Fig. 4B).

As an additional demonstration of interaction between ILF3/NF90 and PRMT1, we used GST-PRMT1 "pull-down" experiments. Recombinant GST-PRMT1 or GST protein was allowed to interact with HeLa cell lysate. Complexes were then recovered with glutathione-Sepharose 4B beads. The presence of ILF3/NF90 in the complexes was analyzed by electrophoresis and immunoblotting with anti-ILF3/NF90 antibody. GST-PRMT1, but not GST protein, could precipitate ILF3/NF90 protein from cell lysate (Fig. 4C).

ILF3, but Not NF90, Interacts with PRMT1-- The RGG/RXR region of ILF3 appears to contribute strongly to the interaction between PRMT1 and ILF3 (Fig. 3). Because NF90 does not contain the RGG/RXR region found in ILF3, we thought that (despite their substantial similarity) ILF3 and NF90 might show quite different interactions with PRMT1. We first examined the interactions of ILF3 and NF90 with PRMT1 in GST pull-down experiments. In this experiment, we examined the ability of in vitro translated, [³⁵S]Met-labeled PRMT1 to be "pulled down" by recombinant GST-ILF3, GST-ILF3C (Met¹-Val⁶²¹), GST-40A (ILF3 Lys³⁵⁵-Arg⁹¹⁰), GST-40AC (ILF3 Lys³⁵⁵-Val⁶²¹), GST-NF90, and negative control GST protein. GST-ILF3 strongly interacted with PRMT1 (Fig. 5A, lane 1). In contrast, NF90 and GST-ILF3C, which are missing the RGG/RXR-rich COOH-terminal region of ILF3, did not interact with PRMT1 (Fig. 5A, lanes 2 and 3). 40A, the original fragment of ILF3 isolated by two-hybrid analysis, interacted strongly with PRMT1 (Fig. 5A, lane 4). However, when the RGG/RXR-rich COOH-terminal region of 40A was eliminated, the resulting 40AC molecule did not react with PRMT1 (Fig. 5A, lane 5). GST was a negative control in this experiment, and GST-TIS21 was a positive control (10). Thus, the RGG/RXR-rich COOH-terminal region of ILF3 appears to play a key role in the ILF3/PRMT1 interaction.

View larger version (43K): [in this window] [in a new window]   Fig. 5.   ILF3, but not NF90, interacts with PRMT1. A, in vitro translated PRMT1 precipitated by purified recombinant GST-ILF3 fusion proteins. [35S]Methionine-labeled PRMT1 was generated by in vitro transcribing and translating pcDNA3.1-PRMT1 using the TNT T7 coupled in vitro translation kit. 2 µl of in vitro translated PRMT1 was diluted to 500 µl with PBS and incubated with 3 µg of purified GST-ILF3 (lane 1), GST-ILF3C (lane 2), GST-NF90 (lane 3), GST-40A (lane 4), or GST-40AC (lane 5) at 4 °C for 30 min. GST protein (lane 6) was used as a negative control, GST-TIS21 (lane 7) was used as a positive control. Then, 30 µl of glutathione-Sepharose beads (50% suspension in PBS) was added to the reaction mixture, and the mixture was further incubated for 90 min at 4 °C. Proteins associated with the glutathione-Sepharose beads were recovered by centrifugation, and the beads were washed three times with PBS containing 0.5% Triton X-100, 0.5% Tween 20, and 0.1% SDS (10 min). [35S]Methionine-labeled PRMT1 bound to the glutathione-Sepharose beads was subjected to SDS-PAGE and fluorography as described previously (12). B, in vitro translated ILF3, ILF3C, and NF90 precipitated by purified recombinant PRMT1. [35S]Methionine-labeled ILF3, ILF3C, and NF90 were in vitro translated from pcDNA3.1-ILF3, pcDNA3.1-ILF3C, and pcDNA3.1-NF90, respectively, using a TNT T7 coupled in vitro transcription and translation kit. Equal amounts of [35S]methionine-labeled ILF3, ILF3C, and NF90 were diluted to 500 µl with PBS, pulled down with 3 µg of GST-PRMT1 or 3 µg of GST protein, and visualized, using the conditions described for A, by fluorography.

In an experiment that was essentially the reciprocal of the one described above, [³⁵S]Met-labeled ILF3, ILF3C (Met¹-Val⁶²¹), and NF90 were in vitro translated and subjected to pull-down experiments with GST-PRMT1 or GST protein. Using similar amounts of radioactively labeled proteins, far more ILF3 was captured than either of the other two proteins (Fig. 5B). NF90 and ILF3C, which do not contain the RGG/RXR-rich COOH-terminal domain, did not interact well with GST-PRMT1. It seems likely that the GST-PRMT1 pull-down of NF90 (Fig. 4C) is due to the interaction between NF90 and ILF3 (see Fig. 3B and "Discussion").

ILF3 and NF90 Are PRMT1 Substrates-- The COOH-terminal region of ILF3 contains multiple RGG and RXR sequences, the substrates for type I protein-arginine methyltransferases (3). To determine whether ILF3 and NF90 are substrates for protein-arginine methyltransferases, we used GST-ILF3, GST-NF90, and GST-ILF3C as substrates in a methylation reaction with recombinant GST-PRMT1. GST-ILF3 was by far the best methylation substrate (Fig. 6, lane 3). If a similar amount of ILF3C was used as substrate, no substantial methylation was observed (Fig. 6, lane 2). NF90, which has a naturally occurring deletion of most of the RGG/RXR-rich COOH-terminal region of ILF3, was a much poorer methylation substrate than ILF3 (Fig. 6, lane 1).

View larger version (29K): [in this window] [in a new window]   Fig. 6.   ILF3 and NF90 are substrates of PRMT1. 1.5 µg each of purified GST-ILF3, GST-ILF3C, and GST-NF90 were methylated by 0.5 µg of purified GST-PRMT1 in 60-µl reaction mixtures in PBS containing 4.4 µCi of [3H]AdoMet. The reactions were performed at 37 °C for 90 min, and the reaction mixtures were then subjected to SDS-PAGE and fluorography for 5 days.

Recombinant GST-ILF3 Protein Can Modulate PRMT1 Enzyme Activity-- PRMT1 interacts with several other molecules, e.g. BTG1 and TIS21 (10). These molecules modulate the enzyme activity of PRMT1. We therefore investigated whether recombinant ILF3 and NF90 proteins can regulate PRMT1 activity. Preliminary experiments demonstrated that GST-40A (i.e. GST-ILF3 Lys³⁵⁵-Arg⁹¹⁰) could activate recombinant PRMT1 in in vitro methylation assays (data not shown). To test whether full-length ILF3 can modulate the activity of PRMT1 and whether this regulation is dependent on the COOH-terminal arginine- and glycine-rich domain (Arg⁶²²-Arg⁹¹⁰), we first used purified GST-PRMT1 to methylate the hypomethylated substrates present in lysates from adenosine dialdehyde-treated, heated, hypomethylated Rat1 cells. Adenosine dialdehyde treatment inhibits methylation in cells and results in production of hypomethylated substrates; the heat step inactivates the endogenous PRMT activity for the subsequent in vitro methylation assays.² The methylation reactions were carried out in the presence and absence of recombinant GST-ILF3, GST-ILF3C, or GST-NF90 (Fig. 7A). GST-ILF3, but not GST-ILF3C or GST-NF90, enhanced the methylation activity of GST-PRMT1 (comparing lanes 4 and 5-7), especially at 32-, 40-, 45-, 67-, 86-0, and 105-kDa bands. The increased activity for these proteins was not due to GST protein or to the fact that ILF3 is a PRMT1 substrate because GST-heterogeneous nuclear ribonucleoprotein A2 (another PRMT1 substrate) did not enhance the GST-PRMT1 activity in this assay (data not shown).

View larger version (39K): [in this window] [in a new window]   Fig. 7.   GST-ILF3 modulates PRMT1 activity in in vitro methyltransferase assays. A, the activity of purified GST-PRMT1 can be modulated by GST-ILF3 in in vitro methylation assays using hypomethylated proteins present in adenosine dialdehyde-treated Rat1 cell lysate as the methyl-accepting substrates. GST-NF90 (5 µg; lane 1), GST-ILF3 (5 µg; lane 2), and GST-ILF3C (5 µg; lane 3) were methylated by 1 µg of GST-PRMT1 in a final volume of 80 µl of PBS containing 6.6 µCi of [3H]AdoMet. These samples had no cell lysate added, i.e. no endogenous substrates. In lanes 4-7, 20 µg of heat-inactivated, hypomethylated Rat1 cell lysate was methylated by 1 µg of GST-PRMT1 in a final volume of 80 µl of PBS containing 6.6 µCi of [3H]AdoMet in the absence of any other GST fusion protein (lane 4) or in the presence of 0.25 µg of GST-NF90 (lane 5), 0.25 µg of GST-ILF3 (lane 6), or 0.25 µg of GST-ILF3C (lane 7). Heating inactivates the endogenous protein-arginine methyltransferase activity (11). Methylation reactions were performed at 37 °C for 90 min and stopped by adding SDS-PAGE sample buffer. The mixtures were then subjected to SDS-PAGE and fluorography for 2 days. The arrows indicate the proteins whose methylation by GST-PRMT1 was changed when recombinant ILF3 was present in the methylation reaction. B, the activity of endogenous PRMT1 activity in Rat1 cells can be modulated by ILF3. Substrates present in hypomethylated Rat1 cell lysate (20 µg), prepared from adenosine dialdehyde-treated cells, were methylated by the endogenous protein-arginine methyltransferase in a final volume of 80 µl of PBS containing 6.6 µCi of [3H]AdoMet in the absence of any GST fusion protein (lane 1) or in the presence of GST-NF90 (0.25 µg; lane 2), GST-ILF3 (0.25 µg; lane 3), or GST-ILF3C (0.25 µg; lane 4). Methylation reactions were performed, and the products were analyzed as described for A. The arrows indicate the proteins whose methylation by endogenous PRMT1 was changed when recombinant ILF3 was present in the methylation reaction.

To test whether ILF3 proteins can regulate endogenous PRMT1 activity, we performed in vitro methylation reactions using unheated hypomethylated Rat1 cell lysates in the presence and absence of 0.25 µg of recombinant GST-ILF3, GST-ILF3C, and GST-NF90 proteins. Only GST-ILF3 could modulate the endogenous PRMT activity (Fig. 7B, compare lanes 1 and 3). The methylation by endogenous PRMT1 of substrates at 40, 45, 67, and 86 kDa was enhanced in lysate containing GST-ILF3.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

ILF3 Interacts Specifically with PRMT1-- To date, three mammalian enzymes have been demonstrated as type I protein-arginine methyltransferases: PRMT1 (10), PRMT3 (12), and coactivator-associated arginine methyltransferase 1 (5). Among them, PRMT1 contributes most of the type I activity in cells. Immunoprecipitation of PRMT1 can deplete >90% of the total protein-arginine methyltransferase activity present in cell and tissue extracts. Despite the recent cloning of these enzymes, which catalyze the asymmetric dimethylation of arginine in proteins, the biological roles of this post-translational protein modification are still unclear. However, several recent observations have begun to suggest roles for protein arginine methylation. For example, the yeast type I arginine methyltransferase facilitates the export of certain heterogeneous nuclear ribonucleoproteins from the nucleus (31). PRMT1, the predominant type I protein-arginine methyltransferase in mammalian cells and the subject of the studies reported here, is involved in interferon signaling pathways (7). To understand the role of PRMT1 in cells and to study its regulation, we used the yeast two-hybrid system to identify PRMT1-interacting proteins. One of the proteins identified in this screen is ILF3/MPP4. ILF3 is virtually identical to NF90, except in the COOH-terminal region.

ILF3 can interact with PRMT1, but not with PRMT3, a structurally related type I protein-arginine methyltransferase. These data suggest that the ILF3/PRMT1 interaction is not fortuitous, but is likely to play a biological role. The specific interaction between ILF3 and PRMT1 has been shown by four distinct protocols: two-hybrid interaction analysis, immunofluorescent co-localization, GST pull-down experiments, and modulation of PRMT1 enzyme activity. It is particularly important to note that both coprecipitation and pull-down experiments demonstrate interactions between endogenous and not overexpressed proteins. Although it is often possible to force protein/protein interactions by overexpression, endogenous PRMT1 and ILF3 appear to exist in a macromolecular complex in living cells.

NF90 and ILF3 Are Closely Similar in Structure, Except for Their Carboxyl-terminal Regions-- Because ILF3 and NF90 proteins predicted from human cDNA sequences are nearly identical for the first 600 amino acid residues, with the exception of occasional single amino acid changes and a single four-amino acid insertion in NF90 (Fig. 1), we suspect that they are derived from the same gene. It appears that there is only one gene in mouse, located on chromosome 9, for ILF3 and NF90. In the human genome, the ILF3/NF90 gene is located on chromosome 19 between 19q11-qter and 19p11-p13.1 (18). Because there is only one gene encoding ILF3/NF90 and we can detect multiple forms of message and proteins (Fig. 2), it is likely that mRNAs encoding NF90 and ILF3 are derived from alternative splicing of the same precursor RNA. The similarities between NF90 and ILF3 include two double-stranded RNA-binding domains and nuclear translocation signals. Our Western blotting data suggest that both NF90 and ILF3 can exist in the same cells.

Although NF90 and ILF3 are closely related, they interact with different proteins and may well have different cellular functions. NF90 was first purified from Jurkat cells as a dimer with NF45 (17, 20). NF90 also forms complexes with DNA-dependent protein kinase (26) and double-stranded RNA-dependent protein kinase (25). However, in the purified NF90·NF45 complex or complexes of DNA-dependent protein kinase and double-stranded RNA-dependent protein kinase, no 110-kDa ILF3 protein can be detected (25, 26). Since NF90 and ILF3 are so similar in the rest of their structure, it seems likely that the different COOH-terminal regions of ILF3 and NF90 are responsible for the differences in the proteins with which they interact. Because NF90 and ILF3 are likely to be derived from the same gene through alternative splicing (18, 19), it is possible that the function of this alternative splicing is to generate ILF3-related proteins that can function in different pathways as a consequence of their participation in distinct protein complexes.

The Carboxyl-terminal Region of ILF3 Is Crucial to ILF3/PRMT1 Interactions-- We demonstrated in this study that the COOH-terminal 287-amino acid residue region of ILF3 is responsible for ILF3/PRMT1 protein interactions, arginine methylation of ILF3 by PRMT1, and modulation of PRMT1 activity by ILF3. When this 287-residue domain is deleted, ILF3C does not interact with PRMT1 by two-hybrid interaction or GST pull-down experiments. Similarly, when the PRMT1 substrate region in the carboxyl-terminal region of ILF3 is deleted, ILF3C is not a potent activator of PRMT1 methyltransferase activity. Thus ILF3C behaves similarly to NF90 in that it neither interacts with PRMT1 nor modulates PRMT1 enzyme activity. Like ILF3C, NF90 (which does not contain the ILF3 COOH-terminal arginine-glycine-serine-rich domain) does not interact with PRMT1 or modulate PRMT1 enzyme activity.

PRMT1 does not interact with in vitro synthesized NF90 in GST pull-down experiments (Fig. 5), yet purified GST-PRMT1 seems able to interact with NF90 and to precipitate NF90 from cell lysates (Fig. 4). How can we explain these two seemingly contradictory observations? The likely answer is due to the fact that ILF3 and NF90 appear to form a complex (Fig. 2). Because PRMT1 interacts with ILF3, GST-PRMT1 will pull down ILF3 from cell lysate. If other proteins associate with ILF3, e.g. NF90, these proteins should be coprecipitated with ILF3 as well, even though they do not themselves have strong affinity for PRMT1.

It is difficult to determine the relative amounts of either PRMT1 or ILF3 that are complexed together in cells. We and others (7) have previously shown that PRMT1 interacts with several additional molecules, including BTG1 and TIS21 (10) and the interferon-/ receptor. Thus, only a portion of the PRMT1 present in cells appears to be in a macromolecular complex with ILF3. Our coprecipitation experiments (Fig. 4B) demonstrate that only a portion of the ILF3 present in cells can be isolated in a complex with PRMT1. Although it is likely that the fractional interactions of ILF3 and PRMT1 reflect the dynamic interactions of these molecules as they participate in regulatory responses in cells, additional studies will be necessary to clarify the role of PRMT1 modulation of ILF3 activity. Determination of the precise stoichiometry of the ILF3·PRMT1 heterodimer interaction would also be of great interest in considering both the biological consequences of this interaction and composition of the large molecular mass complex in which PRMT1 is found in cells.

It also would be of great interest to determine whether ILF3 homodimers and ILF3·NF90 homodimers interact with PRMT1 and whether the biological activities of these two molecules are differentially regulated by PRMT1. Until antisera that distinguish between these two complexes are available, however, we cannot investigate this issue.

Importance of ILF3 and PRMT1 Interactions-- PRMT1, the first identified protein-arginine methyltransferase (3), is the predominant type I arginine methyltransferase in mammalian cells.² PRMT1 was first isolated because of its ability to form complexes with the TIS21 protein (10). In this study, we identified ILF3 as a PRMT1-interacting protein that regulates PRMT1 activity. Like ILF3, TIS21 protein has the ability to modulate PRMT1 activity in in vitro methyltransferase assays.

Both TIS21 and ILF3 are cell cycle-regulated (21, 32, 33). TIS21 is rapidly induced and then degraded when cells undergo the transition from the resting, non-proliferating G₀ stage to a commitment to enter the cell cycle (32, 33). TIS21 is also induced when cells become committed to apoptosis (34) and when neuronal precursors are committed to differentiation (35). ILF3 is cyclically phosphorylated during mitosis (21). Because (i) both TIS21 and ILF3 can regulate PRMT1 activity and (ii) the expression or structure of TIS21 and ILF3 is regulated during the cell cycle, it seems likely that the activity of PRMT1 may be regulated in the cell cycle as well. There are several reports indicating that the methylation levels of proteins in cells change as cells progress through different stages of the cell cycle (36, 37). A recent report described a new protein methyltransferase interacting with JAK2 kinase and members of the p21^CDC42/Rac-activated kinase family (38). The homologue of this methyltransferase in Saccharomyces pombe or in Saccharomyces cerevisiae is essential for normal cell cycle progression; its removal results in cell cycle abnormalities (39, 40). These observations support the hypothesis that protein methylation may be involved in cell cycle control. As the players begin to be identified and their interactions characterized, the role of regulated protein arginine methylation in this process may become clearer.

The methylation consensus sequences in the carboxyl-terminal region of ILF3 are immediately followed by Ser-Gly-rich sequences. The PHD secondary structure prediction program predicts that the Arg⁶²²-Arg⁹¹⁰ region in ILF3 is likely to adopt loop structures and to be exposed on the exterior of the protein. This exposed region could be readily phosphorylated by kinases active in the M phase. Because this potential phosphorylation domain is adjacent to the methylation consensus sequence, it is tempting to speculate that arginine methylation events may cooperate with protein phosphorylation events either to control the activity of ILF3/MPP4 or to control the affinity of ILF3/MPP4 for its interacting proteins, e.g. PRMT1.

Because ILF3 is so similar to NF90 and contains the two double-stranded RNA-binding motifs and other N-terminal sequences, we assume that ILF3 has RNA and DNA binding activities that may be similar to those of NF90, despite the differences in the macromolecular complexes in which NF90 and ILF3 exist. NF90, along with NF45, can bind to a specific DNA sequence (17, 20). Recently, a protein methyltransferase, coactivator-associated arginine methyltransferase 1, has been shown to activate a transcriptional complex (5). NF90 plays a role in transcription from genes that have an ARRE-2 cis-acting element present in their regulatory regions (17, 20). We tested whether overexpression of PRMT1 or inhibition of methyltransferases in cells has any effect on the ARRE-2 reporter gene transcription. At present, we have not observed any effect of these potential modifiers of ILF3/MPP4 methylation on expression from ARRE-2 reporter genes (data not shown). Additional experiments will be needed to pursue the question of ILF3 arginine methylation on ILF3 biological activity.

	ACKNOWLEDGEMENTS

We thank the members of the Herschman laboratory for helpful discussions.

	FOOTNOTES

^* This work was supported in part by UCLA Asthma, Allergy, and Immunologic Diseases Center Grant AI34567 from NIAID and NIEHS (to H. R. H.) and by National Institutes of Health Grant AI39624 (to P. N. K.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF220102.

^¶ Postdoctoral trainee supported by United States Public Health Service Institutional Research Award T32 CA09056.

^** To whom correspondence should be addressed: 341A Molecular Biology Inst., UCLA, 611 Charles E. Young Dr. East, Los Angeles, CA 90095. Tel.: 310-825-8735; Fax: 310-825-1447; E-mail: hherschman@ mednet.ucla.edu.

Published, JBC Papers in Press, April 3, 2000, DOI 10.1074/jbc.M000023200

	ABBREVIATIONS

The abbreviations used are: PRMT, protein-arginine methyltransferase; ILF, interleukin enhancer-binding factor; ARRE-2, antigen receptor response element-2; NF, nuclear factor; MPP4, M phase phosphoprotein 4; PCR, polymerase chain reaction; GST, glutathione S-transferase; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; [³H]AdoMet, S-adenosyl-L-[methyl-³H]methionine; GST-GAR, glutathione S-transferase-fibrillarin glycine arginine domain fusion protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. Guan, N. Altan-Bonnet, A. M. Parrott, C. J. Arrigo, Q. Li, M. Khaleduzzaman, H. Li, C.-G. Lee, T. Pe'ery, and M. B. Mathews Nuclear Factor 45 (NF45) Is a Regulatory Subunit of Complexes with NF90/110 Involved in Mitotic Control Mol. Cell. Biol., July 15, 2008; 28(14): 4629 - 4641. [Abstract] [Full Text] [PDF]

				P. Bermejo-Alvarez, D. Rizos, D. Rath, P. Lonergan, and A. Gutierrez-Adan Epigenetic differences between male and female bovine blastocysts produced in vitro Physiol Genomics, January 17, 2008; 32(2): 264 - 272. [Abstract] [Full Text] [PDF]

				I. Goulet, G. Gauvin, S. Boisvenue, and J. Cote Alternative Splicing Yields Protein Arginine Methyltransferase 1 Isoforms with Distinct Activity, Substrate Specificity, and Subcellular Localization J. Biol. Chem., November 9, 2007; 282(45): 33009 - 33021. [Abstract] [Full Text] [PDF]