GTC
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE ADVANCED SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Genes to Cells (2009) 14, 113-128. doi:10.1111/j.1365-2443.2008.01256.x
© 2009 Blackwell Publishing or its licensors
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fukuda, T.
Right arrow Articles by Irie, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Fukuda, T.
Right arrow Articles by Irie, K.

hnRNP K interacts with RNA binding motif protein 42 and functions in the maintenance of cellular ATP level during stress conditions

 Toshiyuki Fukuda{dagger}, Takahiro Naiki{dagger}, Marie Saito and Kenji Irie*

Department of Molecular Cell Biology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba 305-8575, Japan


   Abstract
Top
 Abstract
Introduction
 Results
 Discussion
 Experimental procedures
 References
 
Heterogeneous nuclear ribonucleoprotein K (hnRNP K) is a conserved RNA-binding protein that is involved in multiple processes of gene expression, including chromatin remodeling, transcription, RNA splicing, mRNA stability and translation, together with diverse groups of molecular partners. Here we identified a previously uncharacterized protein RNA binding motif protein 42 (RBM42) as hnRNP K-binding protein. RBM42 directly bound to hnRNP K in vivo and in vitro. RBM42 also directly bound to the 3' untranslated region of p21 mRNA, one of the target mRNAs for hnRNP K. RBM42 predominantly localized within the nucleus and co-localized with hnRNP K there. When cells were treated with agents, puromycin, sorbitol or arsenite, which induced the formation of stress granules (SGs), cytoplasmic aggregates of stalled translational pre-initiation complexes, both hnRNP K and RBM42 localized at SGs. Depletion of hnRNP K by RNA interference decreased cellular ATP level following release from stress conditions. Simultaneous depletion of RBM42 with hnRNP K enhanced the effect of the hnRNP K depletion. Our results indicate that hnRNP K and RBM42 are components of SGs and suggest that hnRNP K and RBM42 have a role in the maintenance of cellular ATP level in the stress conditions possibly through protecting their target mRNAs.


   Introduction
Top
 Abstract
 Introduction
Results
 Discussion
 Experimental procedures
 References
 
Heterogeneous nuclear ribonucleoprotein K (hnRNP K) belongs to the family of heterogeneous ribonucleoproteins (hnRNPs), which participate in the processing of premRNAs and mRNA export from the nucleus (Dreyfuss et al. 1993, 2002). hnRNP K contains three K homology (KH) domains that mediate RNA/DNA binding (Braddock et al. 2002; Ostrowski et al. 2002) and three proline-rich motifs that interact with SH3 domain of the Src-family tyrosine kinases (Taylor & Shalloway 1994; Weng et al. 1994; Van Seuningen et al. 1995). hnRNP K is involved in multiple processes of gene expression including chromatin remodeling, transcription, RNA splicing, mRNA stability, and translation (Bomsztyk et al. 2004). These multiple functions of hnRNP K reflect its ability to associate with diverse groups of molecular partners such as chromatin remodeling factors, general transcription factors, transcription factors, splicing factors, and translation factors (Bomsztyk et al. 1997, 2004; Mikula et al. 2006). For example, in the process of chromatin remodeling, hnRNP K interacts with the chromatin remodeling factor Polycomb group (PcG) protein Eed, DNA methyltransferase, and scaffold attachment factor-B (Saf-B) (Shnyreva et al. 2000; Bomsztyk et al. 2004). In the process of transcription, hnRNP K interacts with TATA box-binding protein TBP and Y-box-binding protein (YB-1) (Shnyreva et al. 2000). In the process of RNA splicing, hnRNP K interacts with the splicing factors, 9G8 and SRp20, and other hnRNPs including hnRNP A1, A2/B2, G, L, D and U (Shnyreva et al. 2000; Bomsztyk et al. 2004). In the regulation of RNA stability, hnRNP K interacts with YB-1 and hnRNP E1 and regulates the stability of the renin mRNA (Skalweit et al. 2003). In the process of translational regulation, hnRNP K interacts with hnRNP E1. hnRNP K and hnRNP E1 bind to the differentiation control element in the 3' untranslated region (3' UTR) of reticulocyte-15-lipoxygenase (r15-LOX) mRNA, and silence translation of the mRNA (Ostareck et al. 1997, 2001; Ostareck-Lederer & Ostareck 2004). hnRNP K also acts as a scaffold protein that integrates signaling pathways (Bomsztyk et al. 2004; Adolph et al. 2007). hnRNP K binds to the SH3 domain of c-Src, and this binding leads to the activation of c-Src. c-Src, in turn, phosphorylates hnRNP K and this phosphorylation inhibits the translational repression of r15-LOX by hnRNP K (Ostareck-Lederer et al. 2002; Messias et al. 2006). Extracellular signal-regulated kinase (ERK) also phosphorylates hnRNP K, and this phosphorylation stimulates the translational repression of r15-LOX by hnRNP K through the accumulation of hnRNP K in the cytoplasm (Habelhah et al. 2001). hnRNP K is also methylated by protein arginine methyltransferase 1 (PRMT1) and this methylation inhibits the interaction with c-Src (Ostareck-Lederer et al. 2006). Thus, hnRNP K is considered to act as a docking platform to integrate signaling pathways by facilitating cross-talk between kinases and factors involved in the regulation of gene expression (Bomsztyk et al. 2004).

In response to environmental stresses such as oxidative, genotoxic, hyperosmotic, or heat shock, eukaryotic cells re-program RNA metabolism to repair stress-induced damage and adapt to changed conditions. During this process, the translation of mRNAs encoding ‘housekeeping’ proteins is aborted, whereas the translation of mRNAs encoding molecular chaperones and enzymes involved in damage repair is enhanced. mRNAs encoding ‘housekeeping’ proteins are re-directed from polysomes to discrete cytoplasmic foci known as stress granules (SGs) (Anderson & Kedersha 2002, 2006). SGs are considered to be sites of mRNA triage where ribonucleoprotein (mRNP) complexes are monitored for integrity and composition and are selected to re-initiation, degradation, or storage. SG assembly is initiated by the phosphorylation of translation initiation factor eIF2{alpha}, which reduces the availability of the eIF2-GTP-tRNAMet ternary complex that is necessary for translation initiation. SGs include stalled 48S pre-initiation complex, small ribosomal subunits, and translation initiation factors eIF2, eIF3, eIF4E and eIF4G (Anderson & Kedersha 2006). SGs also contain p54/Rck helicase, the 5'–3' exonuclease Xrn1, and many RNA-binding proteins. These RNA-binding proteins include poly(A) binding protein 1 (PABP1), T cell intracellular antigen-1 (TIA-1), TIA related (TIAR), HuR, Staufen, Smaug, tristetraprolin (TTP), Fragile X mental retardation protein, G3BP, CPEB, SMN, and one of hnRNPs, hnRNP A1 (Anderson & Kedersha 2006; Guil et al. 2006). Although hnRNP K has been reported to be involved in various processes of RNA metabolism and signal transduction (Bomsztyk et al. 2004), it remains unknown whether hnRNP K localized to SGs or whether hnRNP K is involved in the stress-induced regulation of RNA metabolism.

Here we show that hnRNP K and its new binding partner RNA binding motif protein 42 (RBM42) are components of SGs. RBM42 is previously uncharacterized protein that contains an RNA recognition motif (RRM) in its carboxy-terminal region. RBM42 directly bound to hnRNP K in vivo and in vitro. RBM42 also directly bound to the 3' untranslated region of p21 mRNA, one of the target mRNAs for hnRNP K. Under nonstress conditions, both hnRNP K and RBM42 predominantly localized in the nucleus. Under stress conditions, these proteins formed cytoplasmic foci where the SG marker TIAR localized. Depletion of hnRNP K by RNA interference decreased cellular ATP level following release from stress conditions, and simultaneous depletion of RBM42 with hnRNP K enhanced the effect of the hnRNP K depletion. Our results indicate that hnRNP K and its new binding partner RBM42 are components of SGs and suggest that hnRNP K and RBM42 have a role in the maintenance of cellular ATP level in the stress conditions possibly through protecting their target mRNAs.


   Results
Top
 Abstract
 Introduction
 Results
Discussion
 Experimental procedures
 References
 
Identification of RBM42 as a novel hnRNP K binding protein

To identify a new binding partner of hnRNP K, we performed the yeast two-hybrid screening using hnRNP K as a bait (Fig. 1A). We screened 2 x 105 clones of a rat lung library and 2 x 105 clones of a mouse embryo library, and obtained 129 and 13 positive clones, respectively. Thirteen rat clones and one mouse clone encoded the carboxyl-terminal portion (aa 30–474) of rat hypothetical protein LOC361545 (GENBANK/EMBL/DDBJ accession number NP_001014181) and the C-terminal portion (aa 33–474) of mouse hypothetical protein LOC68035 (GENBANK/EMBL/DDBJ accession number NP_598454 [GenBank] ), respectively. These rat and mouse proteins are highly similar to human hypothetical protein MGC10433/ RBM42 (GENBANK/EMBL/DDBJ accession number NM_024321 [GenBank] ). This human RBM42 protein contains one RRM (Fig. 1B), and RRM is highly conserved in its mouse and rat orthologs. RBM42 protein has been reported to associate hnRNP K in the genome-wide protein-protein network analysis (Rual et al. 2005). However, it has not been analyzed whether this hypothetical protein RBM42 is indeed expressed in cells or whether RBM42 interacts with hnRNP K in vivo. We isolated the full-length clones of two isoforms of human RBM42 cDNA (Fig. 1B). Longer isoform encoded the protein composed of 480 amino acids with a calculated relative molecular mass (Mr) of 50.4 kDa, and shorter isoform encoded the protein composed of 450 amino acids with a calculated Mr of 47.4 kDa. The shorter isoform lacked 30 amino acids of aa 147–186 of the longer isoform (Fig. 1B). We tentatively named these two isoforms, RBM42a and RBM42b. The amino acid sequences of mouse RBM42 and rat RBM42 were 92% identical to each other, and 88% and 87% identical to that of human RBM42, respectively. Two-hybrid analysis revealed that the carboxy-terminal region (aa 230–474) of rat RBM42 containing RRM is required for the binding of hnRNP K (Fig. 1C, upper panel). Conversely, the carboxy-terminal region (aa 220–464) of hnRNP K containing third KH domain is required for the binding of RBM42 (Fig. 1C, lower panel). This result suggested that these two RNA-binding proteins interact through their C-terminal regions containing RNA-binding domains.


Figure 1
View larger version (27K):
[in this window]
[in a new window]

  		Figure 1  Identification of RBM42 as an hnRNP K binding protein. (A) Yeast two-hybrid assay showing the specific binding of RBM42 to hnRNP K. Yeast transformants with the indicated plasmids were streaked on the synthetic complete medium lacking adenine to score the ADE2 reporter activity and incubated at 30 °C for 3 days. (B) Schematic structure of RBM42a and RBM42b. RRM; RNA recognition motif. (C) Association of RBM42 with hnRNP K was examined by yeast two-hybrid analysis. Yeast transformants with the indicated plasmids were streaked on the synthetic complete medium lacking adenine to score the ADE2 reporter activity and incubated at 30 °C for 3 days.

 
Expression of RBM42 mRNA and protein

We first examined the expression of RBM42 mRNA in mouse tissues and mouse and human cell lines. Northern blot analysis using the fragment of RBM42 cDNA (bp 193–1526) as a probe detected 1.6 kb mRNA in all the mouse tissues examined, including heart, brain, spleen, lung, liver, skeletal muscle, kidney and testis (Fig. 2A, RBM42). Tissue expression of RBM42 is similar to that of hnRNP K (Fig. 2A, RBM42, hnRNP K). Expression of RBM42 mRNA as well as that of hnRNP K was also detected in mouse and human cell lines, HeLa, HEK293, NIH3T3, and C2C12 (Fig. 2B).


Figure 2
View larger version (104K):
[in this window]
[in a new window]

  		Figure 2  Expression of RBM42 in different tissues and cell lines. (A) Expression of RBM42 mRNA in mouse tissues was examined by Northern blotting. Multiple tissue Northern Blots (Clontech) containing mouse poly(A) RNA was hybridized with a DIG-labeled RBM42 and hnRNP K probe. The hybridized DIG-labeled probe was detected using alkaline phosphatase-conjugated anti-DIG Ab. The lower panel shows same membrane hybridized with a GAPDH probe. (B) Total RNA was extracted from indicated cell lines and expression of RBM42 and hnRNP K mRNAs were examined by Northern blotting as described in (A). The lower panel shows same membrane hybridized with a GAPDH probe.

 
We next examined the expression of RBM42 protein in mouse and human cell lines. For this purpose, two anti-RBM42 polyclonal antibodies (pAbs), N03 against the N-terminal portion of RBM42 (aa 30–133) and C01 against the C-terminal portion of RBM42 (aa 308–474) were generated. HeLa cells were transfected with pCMV-HA-RBM42a and pCMV-HA-RBM42b, which expressed HA-tagged full-length human RBM42a and RBM42b, respectively. The cell extracts were subjected to SDS-PAGE, followed by Western blotting with the anti-HA monoclonal Ab (mAb) and two anti-RBM42 pAbs. The exogenously expressed HA-tagged RBM42a and RBM42b proteins were detected by all three Abs (Fig. 3A and data not shown). In the long exposure, endogenous RBM42a and RBM42b proteins as well as the exogenously expressed proteins were detected by two anti-RBM42 pAbs, N03 and C01 (Fig. 3A, long exposure, and data not shown).


Figure 3
View larger version (56K):
[in this window]
[in a new window]

  		Figure 3  Expression of RBM42 protein in cultured cells. (A) HeLa cells were transiently transfected with pCMV-HA-hRBM42a or pCMV-HA-hRBM42b. Extracts prepared from transfected cells were subjected to SDS-PAGE, followed by Western blotting. Exogenously expressed HA-hRBM42a and HA-hRBM42b were detected with the anti-HA mAb (upper panel) or the anti-RBM42 pAb N03 (middle panel). Endogenous RBM42a and RBM42b as well as exogenous expressed HA-RBM42 were also detected with the anti-RBM42 pAb N03 in the long exposure (bottom panel). *Nonspecific band. Essentially same results were obtained with the anti-RBM42 pAb C01. (B) Expression of endogenous RBM42 protein in several cell lines. Extracts prepared from indicated cell lines were subjected to SDS-PAGE, followed by Western blotting. Endogenous RBM42 was detected with the anti-RBM42 pAb N03. (C) NIH3T3 cells were transfected with dsRNA designated for RBM42 (#1, #2) or control dsRNA (C) and harvested 24 h after transfection. Whole cell extracts were subjected to SDS-PAGE, followed by Western blotting analysis with the anti-RBM42 pAb N03 and anti-{alpha}-tubulin mAb. {alpha}-Tubulin was used as a loading control.

 
When the extracts of HeLa, HEK293, NIH3T3, and C2C12 cells were subjected to SDS-PAGE, followed by Western blotting with the two anti-RBM42 pAbs, two bands were detected (Fig. 3B and data not shown). Mr of the larger protein (c. 50 kDa) and the smaller protein (c. 47 kDa) was similar to that of HA-tagged RBM42a and HA-tagged RBM42b (Fig. 3A,B and data not shown). To confirm whether these two bands corresponded to RBM42a and RBM42b proteins, we generated RBM42 knockdown cells. NIH3T3 cells were transfected with dsRNA designated for RBM42 and then analyzed by Western blotting with the anti-RBM42 pAb. The levels of two bands corresponding to RBM42a and RBM42b were reduced in cells transfected with two different sequences of RBM42 dsRNA #1 and #2 as compared with cells transfected with control dsRNA (Fig. 3C). Same results were obtained with MTD-1 A cells (see Fig. 9A). Therefore, we concluded that two isoforms of RBM42 protein, RBM42a and RBM42b, were expressed in these human and mouse cell lines and that the isolated cDNAs encoded full-length cDNAs.


Figure 9
View larger version (36K):
[in this window]
[in a new window]

  		Figure 9  hnRNP K and RBM42 are not required for the formation of SGs. (A) MTD-1 A cells were transfected with either dsRNA designated for hnRNP K (#1, #2), RBM42 (#1, #2), or control RNA and harvested 48 h after transfection. Western blot analysis of the transfected cells shows specific depletion of the targeted proteins. (B) MTD-1 A cells were transfected with either dsRNA designated for hnRNP K (#1) or RBM42 (#1). Forty-eight hours after transfection, cells were stressed with final 0.25 mg/mL puromycin for 6 h to induce SGs and PBs formation. After treatments, cells were stained with anti-RBM42, anti-hnRNP K and anti-TIAR Abs as indicated. TIAR was used as a Stress Granules marker. Bars, 20 µm.

 
Co-immunoprecipitation of RBM42 with hnRNP K

We next examined the interaction of hnRNP K with RBM42 in cultured cells. First, we performed the immunoprecipitation analysis. HeLa cells were co-transfected with the FLAG-tagged full-length hnRNP K (FLAG-hnRNP K) and the HA-tagged full-length RBM42a (HA-RBM42a). When FLAG-hnRNP K was immunoprecipitated from the cell extract with the anti-FLAG mAb, HA-RBM42a was co-immunoprecipitated as detected by Western blotting with the anti-HA mAb (Fig. 4A, IP:Flag). Conversely, when HA-RBM42a was immunoprecipitated from the cell extract with the anti-HA mAb, FLAG-hnRNP K was co-immunoprecipitated as detected by Western blotting with the anti-FLAG mAb (Fig. 4A, IP:HA).


Figure 4
View larger version (35K):
[in this window]
[in a new window]

  		Figure 4  RBM42 interacts with hnRNP K. (A) HeLa cells were transiently transfected with pCMV-FLAG-hnRNP K and pCMV-HA-hRBM42a as indicated. Extracts prepared from transfected cells were subjected to immunoprecipitation with the anti-FLAG mAb or the anti-HA mAb. The immunocomplex was analyzed by Western blotting with the anti-HA and anti-FLAG mAbs. (B) Extracts prepared from HeLa cells were immunoprecipitated with the anti-RBM42 pAb or the anti-hnRNP K mAb in the presence (+) or absence (–) of RNase A. The immunocomplexes were analyzed by Western blotting with the anti-hnRNP K mAb and the anti-RBM42 pAb.

 
We next examined whether endogenous RBM42 protein interacted with endogenous hnRNP K protein. When endogenous hnRNP K was immunoprecipitated from the extract of HeLa cells with anti-hnRNP K mAb, endogenous RBM42a and RBM42b were co-immunoprecipitated as detected by Western blotting with the anti-RBM42 pAb (Fig. 4B, left panel, IP:hnRNP K). Conversely, when endogenous RBM42a and RBM42b were immunoprecipitated from the extract of HeLa cells with anti-RBM42 pAb, endogenous hnRNP K was co-immunoprecipitated as detected by Western blotting with the anti-hnRNP K mAb (Fig. 4B, right panel, IP:RBM42). hnRNP K, RBM42a, or RBM42b was not co-immunoprecipitated with control IgG. Same results were obtained with MTD-1 A cells (data not shown). Thus, our results indicate that endogenous RBM42 protein interacted with endogenous hnRNP K protein in these cell lines.

RBM42 and hnRNP K interacts through their C-terminal regions containing the RNA-binding domains (Fig. 1C). This result, together with the observation that both hnRNP K and RBM42 directly bind to RNA (see Fig. 6), implicated that RNA molecules would be involved in the association between RBM42 and hnRNP K. To test this possibility, we preformed an RNase A treatment of the cell extract in the co-immunoprecipitation analysis. The treatment of the cell extract with RNaseA abolished the co-immunoprecipitation of hnRNP K with RBM42 (Fig. 4B, IP:RBM42, RNase +). This result suggested that RNA molecules mediate the interaction between RBM42 and hnRNP K in cells.


Figure 6
View larger version (38K):
[in this window]
[in a new window]

  		Figure 6  RBM42 associates with 3' UTR of p21 mRNA through RNA recognition motif. (A) MTD-1 A cells extracts were incubated with in vitro transcribed p21 3' UTR (p21) biotinylated RNA. RNA-protein complexes were isolated with streptavidin beads, and the pulled-down proteins were detected by Western blot using anti-RBM42 or hnRNP K antibody. (B) NIH3T3 cells were transfected with HA-RBM42a, HA-RBM42a{Delta}RRM, HA-hnRNP K or HA-hnRNP K mut (HA-hnRNP K I403A G404A G407A) expressing plasmids. Whole cell extracts were incubated with in vitro transcribed p21 3' UTR biotinylated RNA. RNA-protein complexes were isolated with streptavidin beads, and the pulled-down proteins were detected by Western blot using anti-HA antibody. (C) GST-RRM and GST were purified from E. coli and then analyzed by a biotin pull-down assay with the p21 3' UTR biotinylated RNA (p21). The pulled down proteins were detected by Western blotting using anti-GST antibody.

 
Direct binding of RBM42 to hnRNP K in vitro

We next examined whether RBM42 directly binds to hnRNP K using purified proteins. For this purpose, we utilized an MBP-fusion protein of the carboxy-terminal region (aa 220–464) of hnRNP K containing third KH domain (MBP-hnRNP K-C), which is sufficient for binding to RBM42, and a GST-fusion protein of the carboxy-terminal region of RBM42a (aa 230–474) containing RRM (GST-RBM42-C), which is sufficient for binding to hnRNP K (Fig. 1C). MBP-hnRNP K-C bound to GST-RBM42-C immobilized on glutathione-sepharose beads (Fig. 5). MBP-hnRNP K-C bound to GST-RBM42-C, but not to GST alone (Fig. 5, middle panel, GST vs. GST-RBM42-C). MBP alone did not bind to GST-RBM42-C or GST alone (Fig. 5, right panel).


Figure 5
View larger version (39K):
[in this window]
[in a new window]

  		Figure 5  RBM42 binds to hnRNP K in GST-pull down assay. The purified MBP-hnRNP K-C (middle panel) or MBP (right panel) were incubated with GST-RBM42-C or GST immobilized on glutathione-sepharose beads in the presence (+) or absence (–) of RNase A. Proteins bound to glutathione-sepharose beads were washed and subjected to SDS-PAGE and stained with Coomassie Brilliant Blue.

 
Since the treatment of the cell extract with RNaseA abolished the co-immunoprecipitation of hnRNP K with RBM42 (Fig. 4B), we next examined whether an RNase A treatment also abolished the binding of these recombinant proteins. As shown in Fig. 5 (RNase +), the treatment with RNaseA did not affect the binding of MBP-hnRNP K-C to GST-RBM42-C. Thus, the recombinant GST-RBM42-C protein can directly bind to the recombinant MBP-hnRNP K-C protein in an RNA-independent manner, although the interaction of RBM42 and hnRNP K in cultured cells is likely to be mediated by RNA molecules.

RBM42 binds to 3' UTR of p21 mRNA through RRM.

As RBM42 contains RRM, we examined whether RBM42 directly bind to RNA (Fig. 6A–C). hnRNP K has been shown to bind to differentiation control element in the 3' UTR of r15-LOX mRNA and CU-rich element in the 3' UTR of p21 mRNA (Ostareck et al. 2001; Ostareck-Lederer & Ostareck 2004; Yano et al. 2005). To assess whether RBM42 binds to RNA, we performed a biotin pull-down assay with extracts from MTD-1 A cells using synthesized biotin-labeled RNA of 3' UTR of p21 mRNA (Fig. 6A). Endogenous RBM42 protein as well as hnRNP K protein were co-precipitated with biotinylated 3' UTR of p21 mRNA.

We next examined whether the RRM of RBM42 was required for its binding to the RNA. We generated HA-tagged version of a deletion mutant of RBM42a, which lacked the RRM (RBM42a{Delta}RRM). HA-RBM42a{Delta}RRM was not co-precipitated with the biotinylated 3' UTR of p21 mRNA (Fig. 6B), indicating that RBM42 binds to the RNA through its RRM. Finally, we examined whether the RRM of RBM42 directly binds to the RNA. GST-RRM were purified from Escherichia coli and then analyzed by a biotin pull-down assay. As shown in Fig. 6(C), GST-RRM was co-precipitated with the biotinylated 3' UTR of p21 mRNA, whereas GST did not. These results suggest that RBM42 directly binds to 3' UTR of p21 mRNA through its RRM.

Subcellular localization of RBM42

Previous study indicates that hnRNP K localizes in both nucleus and cytoplasm (Bomsztyk et al. 2004). To investigate RBM42 functions together with hnRNP K in mammalian cells, we examined by immunofluorescence microscopy whether RBM42 co-localizes with hnRNP K. As shown in Fig. 7(A), exogenously expressed Myc-RBM42a was localized in nucleus of most transfected NIH3T3 cells (approximately 70% of transfected cells) and co-localized with HA-hnRNP K there. Myc-RBM42b was also localized in nucleus (data not shown). Same results were obtained with MTD-1 A cells (see Fig. 10C). Exogenously expressed Myc-RBM42a and HA-hnRNP K were also found in both nucleus and cytoplasm of some transfected cells (approximately 30% of transfected cells, see Fig. 10B, upper panel, Myc-RBM42).


Figure 7
View larger version (39K):
[in this window]
[in a new window]

  		Figure 7  Subcellular localization of RBM42 and hnRNP K. (A) Localization of exogenously expressed Myc-RBM42 and HA-hnRNP K proteins in NIH3T3 cells. NIH3T3 cells were transfected with Myc-RBM42a and HA-hnRNP K. Cells were double stained with Rhodamine-conjugated anti-myc mAb and FITC-conjugated anti-FLAG mAb. The nuclei were counterstained with DAPI. Bars, 10 µm. (B) Localization of endogenous RBM42 and hnRNP K proteins in NIH3T3 cells. Fluorescent micrographs of fixed NIH3T3 cells as indicated. Cells were double stained with the anti-RBM42 pAb and the anti-hnRNP K mAb. The nuclei were counterstained with DAPI. Bars, 10 µm. (C) Subcellular localization of RBM42 and hnRNP K in NIH3T3 cells was determined by cellular fractionation. Cellular proteins were fractionated into cytoplasmic and nuclear fractions, which were then analyzed by Western blotting with the anti-hnRNP K mAb, the anti-RBM42 pAb, anti-NAP2 pAb, and anti-H2A, H2B pAb. Nucleosome assembly protein 2 (NAP2) was used as a cytoplasmic marker, and Histone 2A, 2B (H2A, H2B) were used as a nuclear marker.

 

Figure 10
View larger version (47K):
[in this window]
[in a new window]

  		Figure 10  hnRNP K and RBM42 independently localize to SGs. (A) MTD-1 A cells were transfected with either dsRNA designated for hnRNP K (#1), RBM42 (#1), or control RNA. Forty-eight hours after transfection, cells were stressed with final 0.25 mg/mL puromycin for 6 h to induce SGs and PBs formation. After treatments, cells were stained with anti-RBM42 and anti-hnRNP K Abs as indicated. Bars, 20 µm. (B) MTD-1 A cells were transfected with Myc-RBM42a (upper column) or HA-hnRNP K (bottom column). Twenty-four hours after transfection, cells were stained with anti-hnRNP K, anti-RBM42, anti-myc and anti-HA Abs as indicated. Bars, 20 µm. (C) MTD-1 A cells were transfected with HA-hnRNP K, HA-hnRNP K mut (HA K mut; HA-hnRNP K-I403A, G404A, G407A), HA-RBM42a, or HA-RBM42a{Delta}RRM (HA-{Delta}RRM). Twenty-four hours after transfection, cells were stressed with final 0.25 mg/mL puromycin for 6 h to induce SGs and PBs formation. After treatments, cells were stained with anti-HA and anti-TIAR Abs. Bars, 10 µm.

 
We then examined the localization of endogenous RBM42 and hnRNP K proteins using the anti-RBM42 pAb and anti-hnRNP K mAb in NIH3T3 cells. Endogenous RBM42 and hnRNP K were predominantly localized in the nucleus, and they were co-localized there (Fig. 7B). No clear cytoplasmic signal for endogenous RBM42 or hnRNP K was observed. Same results were obtained with MTD-1 A cells (see Fig. 8A, Control). Although hnRNP K seems to predominantly localize in the nucleus by immunofluorescence microscopy, hnRNP K has been reported to exist not only in the nuclear fraction but also in the cytoplasmic fraction by cell fractionation (Habelhah et al. 2001). Then we examined the localization of RBM42 by cell fractionation. Cellular proteins were fractionated into cytoplasmic and nuclear fractions, which were then analyzed by immunoblotting. Nucleosome assembly protein 2 (NAP2) were used as a cytoplasmic marker, and Histone 2 A, 2B (H2A, H2B) were used as a nuclear marker. As shown in Fig. 7(C), endogenous RBM42 and hnRNP K proteins were detected not only in the nuclear fraction but also in the cytoplasmic fraction. These results suggest that RBM42 may function as a complex with hnRNP K not only in the nucleus but also in the cytoplasm.


Figure 8
View larger version (42K):
[in this window]
[in a new window]

  		Figure 8  RBM42 and hnRNP K accumulates in SGs, but not in PBs. (A) Localization of RBM42 and hnRNP K in SGs. MTD-1 A cells were untreated (Control) or treated with final 0.25 mg/mL puromycin for 6 h (Puro), 0.6 M sorbitol for 2.5 h (OSM), or 0.5 mM arsenite for 1 h (ARS) to induce SGs and PBs formation. After treatments, cells were stained with anti-RBM42 and anti-hnRNP K Abs as indicated. Bars, 10 µm. (B) Co-localization of RBM42, hnRNP K, and SG marker protein TIAR. MTD-1 A cells were treated with final 0.25 mg/mL puromycin for 6 h (Puro) to induce SGs and PBs formation. After treatments, cells were stained with anti-RBM42, anti-hnRNP K and anti-TIAR Abs as indicated. Bars, 10 µm. (C) No localization of RBM42 and hnRNP K in PBs. MTD-1 A cells were transfected with the indicated plasmids. Twenty-four hours after transfection, cells were stressed with final 0.25 mg/mL puromycin for 6 h to induce SGs and PBs formation. Bars, 10 µm.

 
RBM42 and hnRNP K accumulate in SGs

During the analysis of the cellular localization of RBM42 and hnRNP K, we found that, although RBM42 and hnRNP K predominantly localized in the nucleus, the signals for RBM42 and hnRNP K were also observed in the cytoplasm of the cells treated with a drug that destabilize polysomes, puromycin. When cells were treated with puromycin, the signals of RBM42 and hnRNP K appeared as cytoplasmic foci and the signals of RBM42 and hnRNP K co-localized at the foci (Fig. 8A, puro). It is well known that puromycin causes the formation of SGs (Anderson & Kedersha 2006). We next examined whether the cytoplasmic foci where RBM42 and hnRNP K localized were SGs using the well-known SG marker protein, TAIR (Anderson & Kedersha 2006). RBM42 and hnRNP K were co-localized with TIAR at the cytoplasmic foci when the cells were treated with puromycin (Fig. 8B, puro). Furthermore, when the cells were treated with arsenite or sorbitol, which are known to induce the formation of SGs, RBM42 and hnRNP K were also found in the cytplasmic foci (i.e. SGs) where TIAR localized (Fig. 8A, OSM, ARS, and data not shown).

Since it has been reported that some components of SGs such as TTP and HuR/D were also found in another cytoplasmic foci, processing bodies (PBs) (Kedersha et al. 2005; Anderson & Kedersha 2006), we next examined whether hnRNP K and RBM42 were also localized at PBs. PBs are distinct cytoplasmic sites of mRNA degradation that contain the decapping enzymes Dcp1 and Dcp2 and the exonuclease Xrn1. To distinguish PBs from SGs, we utilized EGFP-TIA1 and mRFP-DCP1a for SGs and PBs markers, respectively. EGFP-TIA-1-positive SGs were differently localized from DCP1a-positive PBs in the cells treated with puromycin (Fig. 8C). The signals for EGFP-RBM42 and EGFP-hnRNP K were also excluded from DCP1a-positive PBs (Fig. 8C). These results indicated that RBM42 and hnRNPK were components of SGs, but not PBs, and suggested that RBM42 and hnRNP K might function in the stress response.

hnRNP K and RBM42 independently localize to SGs

To examine a significance of the localization of hnRNP K and RBM42 at SGs, we first tested whether hnRNP K or RBM 42 was involved in the formation of SGs. For this purpose, we generated hnRNP K and RBM42 knockdown cells. MTD-1 A cells were transfected with two different sequences of dsRNA designated for hnRNP K or RBM42 and then analyzed by Western blotting. The levels of hnRNP K protein and RBM42 protein were reduced in cells transfected with hnRNP K dsRNA #1 or #2 and RBM42 dsRNA #1 or #2, respectively, as compared with cells transfected with control dsRNA (Fig. 9A). When hnRNP K knockdown cells and RBM42 knockdown cells were treated with puromycin and stained with anti-TIAR pAb, both cells formed SGs where TIAR localized similarly to control cells (Fig. 9B). Thus, neither hnRNP K nor RBM42 seems to be required for the formation of SGs.

Since hnRNP K interacted with RBM42, we next examined whether hnRNP K was involved in the localization of RBM42 at SGs or vice versa. RBM42 localized to both the nucleus and SGs in hnRNP K knockdown cells treated with puromycin (Fig. 10A, RNAi-hnRNP K). Similarly, hnRNP K localized to both nucleus and SGs in RBM42 knockdown cells treated with puromycin (Fig. 10A, RNAi-RBM42). These results suggest that hnRNP K and RBM42 independently localize to SGs.

It has been reported that transient expression of some SG components such as G3 BP induces the formation of SGs (Kedersha et al. 2005). Then we next examined whether transient over-expression of RBM42 or hnRNP K induces the formation of SGs. Neither RBM42 nor hnRNP K over-expression induced the formation of SGs (Fig. 10B). It should be noted that some fraction of endogenous hnRNP K moved to the cytoplasm in the cell over-expressing RBM42. While endogenous hnRNP K predominantly localized to nucleus of the untransfected cells, hnRNP K localized in both nucleus and cytoplasm of the cells over-expressing RBM42 (Fig. 10B, upper column, hnRNP K). In contrast, RBM42 predominantly localized in nucleus of both the untransfected cells and the cells over-expressing hnRNP K (Fig. 10B, bottom column, RBM42). Thus, RBM42 might be involved in the cytoplasmic localization of hnRNP K, although RBM42 knockdown did not affect the localization of hnRNP K at SGs.

Since hnRNP K and RBM42 have RNA-binding domains (KH domain and RRM) and bind to RNA (see Fig. 6), we next examined whether the RNA-binding domains of hnRNP K and RBM42 is required for the localization to SGs (Fig. 10C). For this purpose, we utilized mutant forms of hnRNP K and RBM42. hnRNP K mut (hnRNP K-I403A, G404A, G407A) has three mutations in the conserved domain of its third KH domain and has reduced RNA-binding activity (Siomi et al. 1994; see Fig. 6(B), hnRNP K mut). RBM42a {Delta}RRM lacked the RRM and RNA-binding activity (see Fig. 6B, RBM42a{Delta}RRM). As shown in Fig. 10(C), HA-hnRNP K mut and RBM42a {Delta}RRM still localized at SGs. It should be noted that RBM42a {Delta}RRM did not localize in the nucleus, while hnRNP K mut still localized in the nucleus. Thus, the localization of hnRNP K and RBM42 to SGs does not require their RNA-binding domains.

hnRNP K increases cellular ATP level following stress

Finally, we examined a possible role of hnRNP K and RBM42 in the stress response. For this purpose, we compared the recovery of cellular ATP levels from stress of control MTD-1 A cells with that of hnRNP K or RBM42 knockdown cells. Control, hnRNP K knockdown, and RBM42 knockdown cells were stressed with arsenate for 1 h (Fig. 11B–D) or sorbitol for 2.5 h (Fig. 11E–G), the stress stimulus was removed, and the amounts of ATP in the cells were monitored during and after stress release. We confirmed that the levels of hnRNP K protein and RBM42 protein were reduced in a dose-dependent manner in the cells transfected with hnRNP K dsRNA #1 or #2 and RBM42 dsRNA #1 or #2, respectively, as compared with cells transfected with control dsRNA (Fig. 11A). As shown in Fig. 11B,C,E,F, control cells and RBM42 knockdown cells recovered ATP levels well after a few hours. In contrast, even though down-regulation of hnRNP K had no measurable effect on the cellular ATP levels before the induction of stress, the ATP levels in hnRNP K knockdown cells recovered poorly after release (Fig. 11B,E, K#1, K#2). The results obtained by targeting hnRNP K for RNAi with two different sequences were similar (sequences hnRNP K#1 and hnRNP K#2 in Fig. 11B,E) and the effect of hnRNP K knockdown was a dose-dependent (Fig. 11B,E, 2 pmol, 4 pmol), indicating a specific effect of the knockdown of hnRNP K. To confirm further that this poor recovery of ATP levels in the hnRNP K knockdown cells was due to the knockdown of hnRNP K, we complemented the cells with siRNA-resistant hnRNP K. As shown in Fig. 12(A), expression of the siRNA-resistant hnRNP K rescued the cellular ATP levels of the hnRNP K knockdown cells (Fig. 12A. K#1 +resi-hnRNP K). Thus, hnRNP K has a role for the maintenance of cellular ATP levels following release from stress conditions.


Figure 11
View larger version (56K):
[in this window]
[in a new window]

  		Figure 11  The effect of hnRNP K and RBM42 depletions on cellular ATP level. (A) MTD-1 A cells were transfected with dsRNA designated for hnRNP K (K#1, K#2), RBM42 (RBM#1, RBM#2), or control RNA as indicated and harvested 48 h after transfection. Western blot analysis of the transfected cells shows specific depletion of the targeted proteins. (B–D) MTD-1 A cells were transfected with the indicated amount of dsRNA designated for hnRNP K (#1, #2), RBM42 (#1, #2), or control RNA. Forty-eight hours after transfection, cells were stressed with 0.5 mm arsenite for 1 h, followed by replacement of the medium with standard growth medium. Cellular ATP levels were assessed for each time point. The values represent the means ± SD of three independent experiments. (E–G) MTD-1 A cells were transfected with the indicated amount of dsRNA designated for hnRNP K (#1, #2), RBM42 (#1, #2), or control RNA. Forty-eight hours after transfection, cells were stressed with 0.6 M sorbitol for 2.5 h, followed by replacement of the medium with standard growth medium. Cellular ATP levels were assessed for each time point. The values represent the means ± SD of three independent experiments.

 

Figure 12
View larger version (32K):
[in this window]
[in a new window]

  		Figure 12  Complementation of si-RNA resistant forms of hnRNP K and RBM42 to the effect of the hnRNP K and RBM42 depletion. (A) Complementation of siRNA-resistant hnRNP K to the effect of the hnRNP K depletion. MTD-1 A cells were infected with control virus (Vector) or virus containing a cDNA for siRNA-resistant hnRNP K (resi-hnRNP K), siRNA-resistant hnRNP K mutant (resi-hnRNP K mut), or siRNA-resistant hnRNP K-N (resi-hnRNP K-N). Infected cells were transfected with dsRNA designated for hnRNP K (K#1). Forty-eight hours after transfection, cells were stressed with 0.6 M sorbitol for 2.5 h, followed by replacement of the medium with standard growth medium. Cellular ATP levels were assessed for each time point. The values represent the means ± SD of three independent experiments. **P < 0.01 for resi-hnRNP K-infected cells (resi-hnRNP K) vs. vector-infected cells (Vector). (B) Complementation of siRNA-resistant RBM42a to the effect of the RBM42 depletion. MTD-1 A cells were infected with control virus (Vector) or virus containing a cDNA for siRNA-resistant RBM42a (resi-RBM42) or siRNA-resistant RBM42{Delta}RRM (resi-RBM42{Delta}RRM). Infected cells were transfected with the indicated amount of dsRNA designated for hnRNP K (K#1) and RBM42 (#1). Forty-eight hours after transfection, cells were stressed with 0.6 M sorbitol for 2.5 h, followed by replacement of the medium with standard growth medium. Cellular ATP levels were assessed for each time point. The values represent the means ± SD of three experiments. **P < 0.01 for resi-RBM42-infected cells (resi-RBM42) vs. vector-infected cells (Vector).

 
Since the depletion of RBM42 alone did not affect the cellular ATP levels, we next tried simultaneous depletion of RBM42 and hnRNP K. The simultaneous depletion of RBM42 and hnRNP K resulted in more reduced cellular ATP levels than the hnRNP K depletion alone (Fig. 11D, arsenite, G, sorbitol). To confirm that this additive effect of RBM42 depletion was indeed due to the knockdown of RBM42, we complemented the cells with siRNA-resistant RBM42. Expression of the siRNA-resistant RBM42 rescued the ATP levels of the hnRNP K and RBM42 double knockdown cells (Fig. 12B, K#1; RBM#1 +resi-RBM42). Thus, RBM42 also have a role for the maintenance of cellular ATP levels following release from stress conditions.

Finally, we examined whether RNA-binding activity of hnRNP K and RBM42 is required for the maintenance of cellular ATP levels in the stress condition. Expression of the siRNA-resistant form of hnRNP K mut, which has a mutation in third KH domain and lacks RNA-binding activity, did not rescue the cellular ATP levels of the hnRNP K knockdown cells (Fig. 12A. K#1 +resi-hnRNP K mut). The siRNA-resistant hnRNP K-N construct (aa 1–220), which lacked the C-terminal region (aa 220–464) and did not bind to RBM42 (Fig. 1C), did not rescue the ATP level of the hnRNP K knockdown cells (Fig. 12A, resi-hnRNP K-N). Similarly, the siRNA-resistant form of RBM42{Delta}RRM, which lacks RRM domain, did not complement the effect of RBM42 depletion (Fig. 12B, K#1; RBM#1 +resi-RBM42{Delta}RRM). These results indicate that RNA-binding activity of hnRNP K and RBM42 is required for the maintenance of cellular ATP levels in the stress condition. Taken together with the observation that hnRNP K and RBM42 are components of SGs, our results suggest that hnRNP K and RBM42 have a role in the maintenance of cellular ATP level in the stress conditions possibly through protecting their bound target mRNAs.


   Discussion
Top
 Abstract
 Introduction
 Results
 Discussion
Experimental procedures
 References
 
hnRNP K is an RNA-binding protein which interacts with diverse group of molecular partners and is involved in multiple processes that compose gene expression together with these partners (Bomsztyk et al. 2004). We isolated here a previously uncharacterized protein RBM42, which contain one RRM in its C-terminal region, as a binding partner of hnRNP K. Although RBM42 protein has been reported to associate with hnRNP K in the genome-wide protein–protein network analysis (Rual et al. 2005), the expression of this hypothetical protein and the in vivo interaction of RBM42 with hnRNP K have not been studied. We first confirmed that two isoforms of this hypothetical protein RBM42, RBM42a and RBM42b, were indeed expressed in several mouse tissues and mammalian culture cell lines by Northern blot and Western blot analyses. Then we confirmed the binding of RBM42 with hnRNP K. We have also shown here that RBM42 directly bound to 3' UTR of p21 mRNA, one of the target mRNAs for hnRNP K (Yano et al. 2005), through the RRM. Several lines of evidence suggest that RBM42 directly binds to hnRNP K in vivo and in vitro: (i) RBM42 binds to hnRNP K as estimated by the yeast two-hybrid assay and co-immunoprecipitation from the extracts of cells exogenously expressing RBM42 and hnRNP K; (ii) purified recombinant protein of the C-terminal region of RBM42 directly binds to purified recombinant protein of the C-terminal region of hnRNP K in a cell-free system; (iii) Endogenous RBM42 and hnRNP K are co-immunoprecipitated from the extracts of HeLa cells; (iv) RBM42 co-localizes with hnRNP K at the nucleus and at the SGs; (v) Exogenously expressed RBM42 recruits endogenous hnRNP K to the cytoplasm. RBM42 contains one RRM in its C-terminal region and this RRM was required for binding to the C-terminal region of hnRNP K which includes proline-rich domains and third KH domain. It should be noted that the treatment of the cell extract with RNaseA abolished the co-immunoprecipitation of hnRNP K with RBM42, while the RNase A treatment did not affect the binding of recombinant proteins of the C-terminal regions of hnRNP K and RBM42 in vitro. Although the recombinant RBM42 protein directly binds to the recombinant hnRNP K protein, RNA molecules might mediate and strengthen the association between RBM42 and hnRNP K in cells. The binding of RNA molecules to N-terminal KH1 and KH2 domain of hnRNP K might affect the binding of C-terminal region of hnRNP K to RBM42. Although many binding partners of hnRNP K have been identified in yeast two-hybrid screening, pull down assays, and affinity capture-MS method (Shnyreva et al. 2000; Bomsztyk et al. 2004; Osterowski et al. 2002; Mikula et al. 2006; Adolph et al. 2007), only few factors such as c-Src have been shown to binds directly to hnRNP K and co-localize with hnRNP K (Adolph et al. 2007). Our results of the direct binding and co-localization of RBM42 with hnRNP K strongly suggest that RBM42 functions with hnRNP K in vivo. Recently, we performed affinity capture-MS analysis with RBM42, but no other hnRNP proteins were found (T. Fukuda, T. Naiki, K. Irie, unpublished observation). In addition, we identified hnRNP K in the two-hybrid screening using RBM42a as a bait (T. Fukuda, T. Naiki, K. Irie, unpublished observation). Thus, RBM42 specifically interacts with hnRNP K. Since exogenously expressed RBM42 has a potential to recruit endogenous hnRNP K to the cytoplasm, RBM42 might be involved in the nucleus-cytoplasmic shuttling of hnRNP K.

We show here that hnRNP K and RBM42 localize to SGs under stress conditions. Knockdown of hnRNP K or RBM42 does not affect the formation of SGs and over-expression of hnRNP K or RBM42 does not induce the formation of SGs. Thus, hnRNP K or RBM42 is not required for the formation of SGs. In addition, although hnRNP K interacts with RBM42, the recruitment of hnRNP K to SGs does not require RBM42 and vice versa. Since the RNA binding domains of hnRNP K and RBM42 is not required for the localization to SGs, hnRNP K and RBM42 might be recruited to SGs through the interaction with other factor (s).

SGs are cytoplasmic aggregates of stalled translational pre-initiation complexes (Anderson & Kedersha 2002, 2006). During stress, mRNAs encoding constitutively expressed ‘housekeeping’ proteins are re-directed from polysomes to SGs, and this process is synchronous with stress-induced translational arrest. SGs are considered to be sites of mRNA triage where mRNP complexes are monitored for integrity and composition and are selected to re-initiation, degradation, or storage (Anderson & Kedersha 2006). Selected mRNAs are delivered from SGs to PBs, which are the sites for mRNA decay, for degradation (Anderson & Kedersha 2006). SGs contain many different mRNAs and proteins including eIF2, eIF3, eIF4E, eIF4G, hnRNP A1, HuR, PABP1, Staufen, TIA1, TIAR, TTP, Xrn1, and Zip-code binding protein (ZBP1) (Kedersha et al. 2005; Anderson & Kedersha 2006; Guil et al. 2006; Stöhr et al. 2006). Some of SGs components, eIF4E, HuR, Staufen, TIA1, TTP, and Xrn1 are also found in PBs, distinct cytoplasmic sites of mRNA degradation (Anderson & Kedersha 2006). hnRNP K and RBM42 are restricted to SGs, similar to eIF3, eIF4G, hnRNP A1, PABP1 and ZBP1. ZBP1 is reported to stabilize specific mRNA targets to protect premature decay of these mRNAs during stress condition (Stöhr et al. 2006). Similar to ZBP1, hnRNP K and RBM42 might also function in the stabilization and protection of their target mRNAs.

We show here that depletion of hnRNP K by RNAi decreased cellular ATP levels following release from stress conditions. Although the depletion of RBM42 alone did not affect the ATP levels, the simultaneous depletion of RBM42 with hnRNP K enhanced the effect of the hnRNP K depletion on a decrease in ATP levels. The RNA-binding activity of hnRNP K and RBM42 is required for the maintenance of cellular ATP levels. Together with the observation that hnRNP K and RBM42 are found at SGs, the site for mRNA storage, but not to PBs, the site for mRNA degradation, these results suggest that hnRNP K and RBM42 protect their target mRNAs that are required for the maintenance of cellular ATP levels at SGs under stress conditions. The C-terminal portion of hnRNP K, which interacts with RBM42, was required for the maintenance of cellular ATP levels, but not for SGs localization. Similarly, RRM domain of RBM42 was required for the maintenance of cellular ATP levels, but not for SGs localization. Thus, hnRNP K and RBM42 independently localize to SGs, but hnRNP K and RBM42 might protect same target mRNAs by their interaction. Since the effect of RBM42 depletion on the ATP levels was weaker than that of hnRNP K, the C-terminal portion of hnRNP K might interact with other factor(s) which has a redundant function with RBM42. Interestingly, it has been reported that another hnRNP, hnRNP A1, also localizes to SGs under stress condition and that hnRNP A1 is required for the maintenance of cellular ATP levels during stress and following release from stress conditions (Guil et al. 2006). hnRNP A1 is a abundant nuclear protein which is involved in alternative splicing regulation. Similar to hnRNP K, hnRNP A1 is also implicated in other processes of RNA metabolism such as mRNA export, mRNA stability, and translation. Thus, hnRNP K together with hnRNP A1 might protect mRNAs that are involved in the maintenance of cellular ATP levels at SGs.

In summary, we have found that hnRNP K and its new binding partner RBM42 are components of SGs. Our results suggest that hnRNP K and RBM42 have a role in the maintenance of cellular ATP levels in the stress condition possibly by protecting their target mRNAs. The identification of target mRNAs for hnRNP K and RBM42 would reveal which mRNAs are required for the maintenance of cellular ATP levels during the stress conditions.


   Experimental procedures
Top
 Abstract
 Introduction
 Results
 Discussion
 Experimental procedures
References
 
Yeast two-hybrid screening

One bait vector, pGBD-hnRNP K was constructed by subcloning the insert encoding human hnRNP K into pGBD-C3 (James et al. 1996). Yeast two-hybrid libraries constructed from mouse 11-day embryo and rat lung cDNAs were purchased from Clontech. The plasmids, pGBD-RBM42a (aa 30–474), pGBD-RBM42a-N (aa 30–230), pGBD-RBM42a-C (aa 230–474), were constructed by subcloning the fragments of rat RBM42a (aa 30–474), RBM42a (aa 30–230) and RBM42a (aa 230–474) into pGBD-C1, respectively. The plasmids, pGAD-hnRNP K (aa 1–464), pGAD-hnRNP K-N (aa 1–220), pGAD-hnRNP K (aa 220–464), were constructed by subcloning the fragments of human hnRNP K (aa 1–464), hnRNP K (aa 1–220), and hnRNP K (aa 220–464), into pGAD-C1, respectively. Two-hybrid screening using the yeast strain PJ69–4 A (MATa trp1–901 leu2–3, 112 ura3–52 his3–200 gal4 gal80 GAL2-ADE2 LYS2::GAL1-HIS3 met2::GAL7-lacZ) was performed as described (James et al. 1996). Standard procedures for yeast manipulations were performed as described (Kaiser et al. 1994).

Northern blot analysis

Total RNAs were prepared from cells with TRIzol reagent (Invitrogen). An amount of 10 µg of each RNA samples was loaded on a 1% agarose gel containing 5.5% formaldehyde and resolved by electrophoresis. RNA was transferred to a nylon membrane and then hybridized with digoxigenin (DIG)-labeled anti-sense probe. After washing and blocking, membrane was incubated with alkaline phosphatase (AP)-conjugated anti-DIG Ab and the signal was detected by enhanced chemiluminescence. Multiple mouse tissue Northern blots containing mouse poly(A) +RNA were purchased from Clontech.

Cell culture and transfection

HeLa cells were cultured in Minimum Essential Eagle's Medium with 10% fetal bovine serum. HEK293 cells, C2C12 cells, and MTD-1 A cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. NIH3T3 cells were cultured in Dulbecco's modified Eagle's medium with 10% bovine serum. HeLa cells and NIH3T3 cells were transfected using the Lipofectamine reagent (Invitrogen).

Construction of expression vectors

Full-length cDNAs of human RBM42a and RBM42b were obtained from a human cDNA prepared from HeLa cells by PCR. The obtained full-length cDNAs encoded two splice variants of human RBM42, RBM42a (GENBANKTM accession number NM_024321 [GenBank] ) and RBM42b (GENBANKTM accession number BC002868 [GenBank] ), consisting of 480 and 450 amino acids, respectively. A cDNA of human hnRNP K was kindly supplied by Dr H. Siomi (Keio University). Plasmids pEGFP-TIA-1 and mRFP-Dcp1a were kindly supplied by Drs N. Kedersha and P. Anderson (Brigham and Women's Hospital). The mammalian expression vectors pCMV-FLAG, pCMV-HA, and pCMV-Myc were used to express N-terminal FLAG-, hemagglutinin (HA)-, and Myc-tagged proteins, respectively (Takaesu et al. 2000). The mammalian expression vector expressing hnRNP K was constructed with pCMV-FLAG and pCMV-HA. The mammalian expression vectors expressing RBM42a and RMB42b, pCMV-HA-RBM42a, pCMV-HA-RBM42b, pCMV-Myc-RMB42a, pCMV-Myc-RMB42b, were constructed with pCMV-HA and pCMV-Myc. The mammalian expression vector expressing RBM42a{Delta}RRM (aa 1–381) and hnRNP K mut (HA-hnRNP K-I403A, G404A, G407A) were constructed with pCMV-HA. The glutathione S-transferase (GST) fusion or maltose-binding protein (MBP) fusion vectors of rat RBM42a (RBM42a, GENBANKTM accession number BC079321 [GenBank] ) and hnRNP K, GST-RBM42a-C (aa 230–474), GST-RBM42a-RRM (aa 376–446), MBP-hnRNP K-C (aa 220–464) were constructed with pGEX4T-1 (Amersham Biosciences) and pMAL-C2 (New England Biolabs). The GST and MBP fusion proteins were purified using glutathione-Sepharose beads (Amersham Biosciences) and amylose resin beads (New England Biolabs), respectively. The plasmid pCMV-HA-resi-hnRNP K, encoding siRNA-resistant hnRNP K, was constructed with pCMV-HA-hnRNP K by PCR-mediated site-directed mutagenesis with primers 5'-CTAT TCCCAAAGACTTAGCAGGATCTATTATTG-3' and 5'-CAAT AATAGATCCTGCTAAGTCTTTGGGAATAG-3'. The plasmid pCMV-HA-hnRNP K mut (HA-hnRNP K-I403A, G404A, G407A) was constructed with pCMV-HA-hnRNP K by PCR-mediated site-directed mutagenesis with primers 5'-GGATCTAT TGCTGCCAAAGGTGCTCAGCGGATTAAAC-3' and 5'-TTT AATCCGCTGAGCACCTTTGGCAGCAATAGATCC-3'. The plasmid pCMV-HA-resi-RBM42a, encoding siRNA-resistant RBM42a, was constructed with pCMV-HA-RBM42a by PCR-mediated site-directed mutagenesis with primers 5'-CCTGG ACCACCAATGATGCTACCGCCAATGGCTC-3' and 5'- GAGCCATTGGCGGTAGCATCATTGGTGGTCCAGG-3'. The retroviral vector expressing siRNA-resistant hnRNP K, siRNA-resistant hnRNP K mut, siRNA-resistant hnRNP K-N, siRNA-resistant RBM42a, and siRNA-resistant RBM42a{Delta}RRM, were constructed with pMX vector.

Antibodies

The GST fusion proteins with fragments of rat RBM42a-N (aa 30–133) and RBM42a-C (aa 308–474) were produced in E. coli, purified, and used as each antigen to raise pAbs in rabbits. Two pAbs against RBM42a-N (aa 30–133) and RBM42a-C (aa 308–474), N03 and C01, were used. An anti-FLAG-M2 mAb was purchased from Sigma. An anti-HA mAb was purchased from Roche Molecular Biochemicals. An anti-hnRNP K mAb was purchased from ImmuQuest. An anti-GST mAb and goat anti-TIAR pAb were purchased from Santa Cruz Biotechnology. A rabbit anti-Nucleosome assembly protein 2 (NAP2) and rabbit anti-Histone 2A, 2B (H2A, H2B) pAbs were kindly supplied by Drs M. Okuwaki and K. Nagata (University of Tsukuba).

Immunoprecipitation and Western blot

HeLa cells were transfected with the expression plasmids in various combinations. Cells were washed with cold PBS and then solubilized in lysis buffer (50 mM Tris–HCl, 2 mM EDTA, 100 mM NaCl, 50 mM NaF, 30 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 1 mM PMSF and 10 µg/mL leupeptine, 0.075 U/mL aprotinin, 1 mM DTT and 1% Triton-X100). Cell extracts were centrifuged at 20 000 g. for 30 min at 4 °C. The supernatants were incubated at 4 °C overnight with protein A Sepharose beads bounded with anti-FLAG or anti-HA mAb in the presence (+) or absence (–) of RNase A. After the beads were washed with lysis buffer, the bound proteins were eluted by boiling the beads in an SDS sample buffer (60 mM Tris–HCl at pH 6.7, 3% SDS, 2% 2-mercaptoethanol, and 5% glycerol) for 5 min. The samples were then subjected to SDS-PAGE followed by Western blotting with the anti-HA and anti-FLAG mAbs. Co-immunoprecipitation of endogenous RBM42 and hnRNP K using HeLa cells and MTD-1 A cells were done as described as above except that endogenous RBM42 and hnRNP K were immunoprecipitated with anti-RBM42 pAb and anti-hnRNP K mAb. Immunoprecipitates were then subjected to SDS-PAGE followed by Western blotting with the anti-hnRNP K mAb and anti-RBM42 pAb.

GST pull-down assay

MBP or MBP-hnRNP K-C was incubated with GST or GST-rRBM42a-C immobilized on glutathione-Sepharose beads 4 °C for 1 h in the presence (+) or absence (–) of RNase A. After the beads were extensively washed with PBS, the beads were boiled in the SDS sample buffer. The samples were then subjected to SDS-PAGE, followed by staining with Coomassie Brilliant Blue.

Biotin pull-down assay

Cell extracts prepared from MTD-1 A cells, NIH3T3 cells, or GST-RRM protein were incubated with biotinylated and in vitro transcribed p21 3'-UTR RNA. RNA-protein complexes were isolated with streptavidin beads. After washing three times, the samples were subjected to Western blot analysis to detect endogenous RBM42, hnRNP K, HA-tagged protein and GST-RRM protein using anti-RBM42, anti-hnRNP K, anti-HA, or anti-GST Abs.

Immunofluorescence microscopy

Cells were cultured on a cover-glass and fixed with 10% formaldehyde in PBS. The fixed sample was treated with 0.1% Triton X-100 in PBS and washed three times with PBS. After the sample was soaked with PBS containing 3% BSA, the sample was treated with various pAbs or mAbs, and washed with PBS containing 3% BSA, followed by incubation with the FITC-conjugated anti-mouse IgG Abs and Rhodamine-conjugated anti-rabbit IgG or Rhodamine-conjugated anti-goat IgG Abs. After the incubation, the sample was washed with PBS, embedded in Fluoromount-G (Southern Biotech), and viewed with the confocal imaging system (Leica).

RNA interference (RNAi) for knockdown of RBM42 and hnRNP K

Double-stranded Stealth RNA duplexes (Invitrogen) corresponding to mouse RBM42 coding region (#1: 5'-AUAAUUGGGCGG AUCACAGGAGCUG-3', #2: 5'-AUUGGUGGCAGCAUC AUGGGUGGUC-3') or mouse hnRNP K coding region (#1: 5'-AAUAGAUCCAGCCAAAUCUUUGGGA-3', #2: 5'-UCA ACUCGCAAUCAAAGUCACUUCC-3') were transfected into NIH3T3 and MTD-1 A cells, respectively, using Lipofectamine RNAiMAX reagent. As a control, we used Stealth RNAi negative control duplexes (Invitrogen) which sequence has no significant homology to any mammalian gene. Cells were cultured for 48 h after transfection and then used for Western blot analysis or viability assay.

Cellular ATP level assay

MTD-1 A cells were transfected with dsRNA using Lipofectamine RNAiMAX reagent. At 48 h post-transfection, cells were stressed with 0.5 mM arsenite for 1 h or 0.6 M sorbitol for 2.5 h, and were left to recover in standard growth medium. At each time point indicated, cellular ATP level was determined using Cell Titer-Glo Luminescent Cell Viability Assay kit (Promega) according to the manufacturer's directions. The assay was performed in triplicate, and the results shown are the averages from three independent experiments.

Retroviral experiments

Plate cells were transfected with 4 µg of pMX vector at 50% confluence. After 2 days, the retroviral supernatant was centrifuged and re-suspended in fresh medium containing 8 µg of polybrene. MTD-1 A cells were plated prior to infection and then incubated with viral medium for 24 h.


   Acknowledgements
 
Authors thank Dr H. Siomi (Keio University) for providing us with the cDNA of hnRNP K, Drs N. Kedersha and P. Anderson (Brigham and Women's Hospital) for providing us with the plasmids pEGFP-TIA-1 and mRFP-Dcp1a and Drs M. Okuwaki and K. Nagata (University of Tsukuba) for providing us with anti-NAP2 and anti-H2AB pAbs. The investigation at University of Tsukuba was supported by grants-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (2005–2007) and by grant from Uehara Memorial Foundation (to K. I).


   Footnotes
 
Communicated by: Eisuke Nishida

{dagger}These two authors have contributed equally to this work. Back

* Correspondence: kirie{at}md.tsukuba.ac.jp


   References
Top
 Abstract
 Introduction
 Results
 Discussion
 Experimental procedures
 References
 
Adolph, D., Flach, N., Mueller, K., Ostareck, D.H. & Ostareck-Lederer, A. (2007) Deciphering the cross talk between hnRNP K and c-Src: the c-Src activation domain in hnRNP K is distinct from a second interaction site. Mol. Cell. Biol. 27, 1758–1770.[Abstract/Free Full Text]

Anderson, P. & Kedersha, N. (2002) Stressful initiations. J. Cell Sci. 15, 3227–3234.

Anderson, P. & Kedersha, N. (2006) RNA granules. J. Cell Biol. 172, 803–808.[Abstract/Free Full Text]

Bomsztyk, K., Denisenko, O. & Ostrowski, J. (2004) hnRNP K: one protein multiple processes. BioEssays 26, 629–638.[CrossRef][Medline]

Bomsztyk, K., Van Seuningen, I., Suzuki, H., Denisenko, O. & Ostrowski. J. (1997) Diverse molecular interactions of the hnRNP K protein. FEBS Lett. 403, 113–115.[CrossRef][Medline]

Braddock, D.T., Baber, J.L., Levens, D. & Clore, G.M. (2002) Molecular basis of sequence-specific single-stranded DNA recognition by KH domains: solution structure of a complex between hnRNP K KH3 and single-stranded DNA. EMBO J. 21, 3476–3485.[CrossRef][Medline]

Dreyfuss, G., Kim, V.N. & Kataoka, N. (2002) Messenger-RNA-binding proteins and the messages they carry. Nat. Rev. Mol. Cell Biol. 3, 195–205.[CrossRef][Medline]

Dreyfuss, G., Matunis, M.J., Pinol-Roma, S. & Burd, C.G. (1993) hnRNP proteins and the biogenesis of mRNA. Annu. Rev. Biochem. 62, 289–321.[CrossRef][Medline]

Guil, S., Long, J.C. & Caceres, J.F. (2006) hnRNP A1 relocalization to the stress granules reflects a role in the stress response. Mol. Cell. Biol. 26, 5744–5758.[Abstract/Free Full Text]

Habelhah, H., Shah, K., Huang, L., Ostareck-Lederer, A., Burlingame, A.L., Shokat, K.M., Hentze, M.W. & Ronai, Z. (2001) ERK phosphorylation drives cytoplasmic accumulation of hnRNP-K and inhibition of mRNA translation. Nat. Cell Biol. 3, 325–330.[CrossRef][Medline]

James, P., Hallady, J. & Craig, E.A. (1996) Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425–1436.[Abstract]

Kaiser, C.A., Adams, A. & Gottschling, D.E. (1994) Methods in Yeast Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

Kedersha, N., Stoecklin, G., Ayodele, M., Yacono, P., Lykke-Andersen, J., Fritzler, M.J., Scheuner, D., Kaufman, R.J., Golan, D.E. & Anderson, P. (2005) Stress granules and processing bodies are dynamically linked sites of mRNP remodeling. J. Cell Biol. 169, 871–884.[Abstract/Free Full Text]

Messias, A.C., Harnisch, C., Ostareck-Lederer, A., Sattler, M. & Ostareck, D.H. (2006) The DICE-binding activity of KH domain 3 of hnRNP K is affected by c-Src-mediated tyrosine phosphorylation. J. Mol. Biol. 361, 470–481.[CrossRef][Medline]

Mikula, M., Dzwonek, A., Karczmarski, J., Rubel, T., Dadlez, M., Wyrwicz, L.S., Bomsztyk, K. & Ostrowski, J. (2006) Landscape of the hnRNP K protein-protein interactome. Proteomics 6, 2395–2406.[CrossRef][Medline]

Ostareck, D.H., Ostareck-Lederer, A., Shatsky, I.N. & Hentze, M.W. (2001) Lipoxygenase mRNA silencing in erythroid differentiation: The 3' UTR regulatory complex controls 60S ribosomal subunit joining. Cell 104, 281–290.[CrossRef][Medline]

Ostareck, D.H., Ostareck-Lederer, A., Wilm, M., Thiele, B.J., Mann, M. & Hentze, M.W. (1997) mRNA silencing in erythroid differentiation: hnRNP K and hnRNP E1 regulate 15-lipoxygenase translation from the 3' end. Cell 89, 597–606.[CrossRef][Medline]

Ostareck-Lederer, A. & Ostareck, D.H. (2004) Control of mRNA translation and stability in haematopoietic cells: the function of hnRNPs K and E1/E2. Biol. Cell 96, 407–411.[CrossRef][Medline]

Ostareck-Lederer, A., Ostareck, D.H., Cans, C., Neubauer, G., Bomsztyk, K., Superti-Furga, G. & Hentze, M.W. (2002) c-Src-mediated phosphorylation of hnRNP K drives translational activation of specifically silenced mRNAs. Mol. Cell. Biol. 22, 4535–4543.[Abstract/Free Full Text]

Ostareck-Lederer, A., Ostareck, D.H., Rucknagel, K.P., Schierhorn, A., Moritz, B., Huttelmaier, S., Flach, N., Handoko, L. & Wahle, E. (2006) Asymmetric arginine dimethylation of heterogeneous nuclear ribonucleoprotein K by protein-arginine methyltransferase 1 inhibits its interaction with c-Src. J. Biol. Chem. 281, 11115–11125.[Abstract/Free Full Text]

Ostrowski, J., Wyrwicz, L., Rychlewski, L. & Bomsztyk, K. (2002) Heterogeneous nuclear ribonucleoprotein K protein associates with multiple mitochondrial transcripts within the organelle. J. Biol. Chem. 277, 6303–6310.[Abstract/Free Full Text]

Rual, J.F., Venkatesan, K., Hao, T., et al. (2005) Towards a proteome-scale map of the human protein–protein interaction network. Nature 437, 1173–1178.[CrossRef][Medline]

Shnyreva, M., Schullery, D.S., Suzuki, H., Higaki, Y. & Bomsztyk, K. (2000) Interaction of two multifunctional proteins. Heterogeneous nuclear ribonucleoprotein K and Y-box-binding protein. J. Biol. Chem. 275, 15498–15503.[Abstract/Free Full Text]

Siomi, H., Choi, M., Siomi, M.C., Nussbaum, R.L. & Dreyfuss, G. (1994) Essential role for KH domains in RNA binding: impaired RNA binding by a mutation in the KH domain of FMR2 that causes fragile X symdrome. Cell 77, 33–39.[CrossRef][Medline]

Skalweit, A., Doller, A., Huth, A., Kähne, T., Persson, P.B. & Thiele, B.J. (2003) Posttranscriptional control of renin synthesis: identification of proteins interacting with renin mRNA-3'-untranslated region. Circ. Res. 92, 419–427.[Abstract/Free Full Text]

Stöhr, N., Lederer, M., Reinke, C., Meyer, S., Hatzfeld, M., Singer, R.H. & Hüttelmaier, S. (2006) ZBP1 regulates mRNA stability during cellular stress. J. Cell Biol. 175, 527–534.[Abstract/Free Full Text]

Takaesu, G., Kishida, S., Hiyama, A., Yamaguchi, K., Shibuya, H., Irie, K., Ninomiya-Tsuji, J. & Matsumoto, K. (2000) TAB2, a novel adaptor protein, mediates activation of TAK1 MAPKKK by linking TAK1 to TRAF6 in the IL-1 signal transduction pathway. Mol. Cell 5, 649–658.[CrossRef][Medline]

Taylor, S.J. & Shalloway, D. (1994) An RNA-binding protein associated with Src through its SH2 and SH3 domains in mitosis. Nature 368, 867–871.[CrossRef][Medline]

Van Seuningen, I., Ostrowski, J., Bustelo, X.R., Sleath, P.R. & Bomsztyk, K. (1995) The K protein domain that recruits the interleukin 1-responsive K protein kinase lies adjacent to a cluster of c-Src and Vav SH3-binding sites. Implications that K protein acts as a docking platform. J. Biol. Chem. 270, 26976–26985.[Abstract/Free Full Text]

Weng, Z., Thomas, S.M., Rickles, R.J., Taylor, J.A., Brauer, A.W., Seidel-Dugan, C., Michael, W.M., Dreyfuss, G. & Brugge, J.S. (1994) Identification of Src, Fyn, and Lyn SH3-binding proteins: implications for a function of SH3 domains. Mol. Cell. Biol. 14, 4509–4521.[Abstract/Free Full Text]

Yano, M., Okano, H.J. & Okano, H. (2005) Involvement of Hu and heterogeneous nuclear ribonucleoprotein K in neuronal differentiation through p21 mRNA post-transcriptional regulation. J. Biol. Chem. 280, 12690–12699.[Abstract/Free Full Text]

Received: 15 October 2008
Accepted: 23 October 2008





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fukuda, T.
Right arrow Articles by Irie, K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Fukuda, T.
Right arrow Articles by Irie, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE ADVANCED SEARCH TABLE OF CONTENTS