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 E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vumbaca, F.
Right arrow Articles by Claffey, K. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vumbaca, F.
Right arrow Articles by Claffey, K. P.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, January 2008, p. 772-783, Vol. 28, No. 2
0270-7306/08/$08.00+0     doi:10.1128/MCB.02078-06
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Double-Stranded RNA-Binding Protein Regulates Vascular Endothelial Growth Factor mRNA Stability, Translation, and Breast Cancer Angiogenesis{triangledown}

Frank Vumbaca, Kathryn N. Phoenix, Daniel Rodriguez-Pinto,{dagger} David K. Han, and Kevin P. Claffey*

Center for Vascular Biology, University of Connecticut Health Center, Farmington, Connecticut 06030-3501

Received 7 November 2006/ Returned for modification 4 December 2006/ Accepted 1 November 2007


arrow
ABSTRACT
 
Vascular endothelial growth factor (VEGF) is a key angiogenic factor expressed under restricted nutrient and oxygen conditions in most solid tumors. The expression of VEGF under hypoxic conditions requires transcription through activated hypoxia-inducible factor 1 (HIF-1), increased mRNA stability, and facilitated translation. This study identified double-stranded RNA-binding protein 76/NF90 (DRBP76/NF90), a specific isoform of the DRBP family, as a VEGF mRNA-binding protein which plays a key role in VEGF mRNA stability and protein synthesis under hypoxia. The DRBP76/NF90 protein binds to a human VEGF 3' untranslated mRNA stability element. RNA interference targeting the DRBP76/NF90 isoform limited hypoxia-inducible VEGF mRNA and protein expression with no change in HIF-1-dependent transcriptional activity. Stable repression of DRBP76/NF90 in MDA-MB-435 breast cancer cells demonstrated reduced polysome-associated VEGF mRNA levels under hypoxic conditions and reduced mRNA stability. Transient overexpression of the DRBP76/NF90 protein increased both VEGF mRNA and protein levels synthesized under normoxic and hypoxic conditions. Cells with stable repression of the DRBP76/NF90 isoform showed reduced tumorigenic and angiogenic potential in an orthotopic breast tumor model. These data demonstrate that the DRBP76/NF90 isoform facilitates VEGF expression by promoting VEGF mRNA loading onto polysomes and translation under hypoxic conditions, thus promoting breast cancer growth and angiogenesis in vivo.


arrow
INTRODUCTION
 
Cellular adaptation to nutrient and metabolic stress is a critical mechanism for tumor cell survival and progression. Hypoxia induces cells to express genes that function to shift cellular metabolism from aerobic oxidative phosphorylation to anaerobic glycolytic pathways. One essential mechanism of this adaptive response is the transcription of target genes through activation of the hypoxia-inducible factor 1 (HIF-1) transcription factor. Genes transcribed by HIF-1 include those in the glycolytic pathway, including glucose transporter 1 (glut1), hexokinase 1 (HK), phosphoglycerate kinase 1 (PGK1), pyruvate kinase (PK), glyceraldehyde phosphate dehydrogenase (GAPDH), and lactate dehydrogenase (LDH) (14, 35). In addition to the genes necessary for glycolysis, survival factors such as vascular endothelial growth factor (VEGF), which can induce local vascular permeability and angiogenesis, are also effectively transcribed by HIF-1 (8, 25, 41).

VEGF expression under hypoxia also requires posttranscriptional mRNA stability and mRNA transport mechanisms. VEGF mRNA is highly labile under normal oxygen and nutrient conditions (3, 5, 17) and is mediated through AU-rich elements (AREs) in the 3' untranslated region (3'-UTR) (5). A consensus destabilization motif (AUUUA) occurs eight times in the human VEGF 3'-UTR, which is 1.6 kb in length (3). ARE-binding proteins such as AUF1 and tristetraprolin (TTP) have all been shown to destabilize mRNAs in various mammalian cell types (2, 4, 11, 40, 44). In addition to their destabilization effects, ARE elements can contribute to mRNA stabilization through interactions with the ELAV family of RNA-binding proteins, which includes Hel-N1, HuC, HuD, and HuR (7, 19, 22, 30). Interestingly, poly(A)-binding protein has been predominantly a stabilizing factor for polyadenylated mRNAs; however, recent investigations suggest that it may also have destabilizing effects (2, 11, 26). Hypoxia-induced mRNA stability has been shown to be a mechanism that can facilitate VEGF expression in tumors even without HIF-1 transcription (32). The identification of 3'-UTR elements in VEGF which promote mRNA stability has determined that AU-rich regions also confer hypoxia-dependent mRNA stability (3, 10, 24). The RNA-binding proteins that interact with these 3'-UTR elements include HuR, hnRNP A1, hnRNP L, poly(A)-binding protein, PAIP2, and TI5IId, according to literature reports (23). HuR and hnRNP L are predominantly nuclear proteins that have the capacity to shuttle between nuclear and cytoplasmic compartments, especially under hypoxic conditions (6, 13, 18, 31). Previous identification of a predicted stem-loop structure in the 3'-UTR of VEGF mRNA showed that this element can provide hypoxia-induced stability to a heterologous mRNA (3). Cross-linking and affinity purification experiments identified both HuR and hnRNP L as RNA-binding proteins for this hypoxia stability region (HSR) element (36). An additional protein with an apparent molecular mass of 90/88 kDa was also found to cross-link to the HSR 126-bp 3'-UTR stem-loop RNA under hypoxic conditions but has not been identified to date (3).

VEGF protein synthesis under hypoxic conditions also requires 5'-UTR mRNA elements to maximize expression. In the case of VEGF, the 5'-UTR contains predicted internal ribosome entry sites that facilitate mRNA loading onto ribosomes and efficient translation (15, 27, 39). These sequences are G/C-rich, have a predicted secondary structure that makes the translation start site accessible, and have been shown to confer increased expression of chloramphenicol acetyltransferase (43) reporter protein under hypoxia in HeLa cells (39). In a recent analysis of translational control mechanisms, eIF-4F initiation complexes were found to be disrupted under conditions of hypoxia (42). This would dictate that 5'-cap-dependent translation would be blocked and that mRNAs with internal ribosome entry site sequences would be preferentially translated under hypoxia. The increased association of carbonic anhydrase IX (CAIX) mRNA with polysomes was demonstrated in prolonged hypoxia up to 16 h, suggesting that mRNA shuttling and loading mechanisms are also important for hypoxia-dependent gene expression (20).

In this study, a 90- to 88-kDa protein complex that binds to the VEGF HSR 3'-UTR A/U-rich stem-loop element that confers hypoxia-dependent mRNA stability was identified. Affinity purification and proteomic analysis revealed that the characteristics of this protein complex were consistent with those of the double-stranded RNA-binding protein-interleukin enhancer binding protein factor-3/nuclear factor family of alternatively spliced DRBPs. One of these alternatively spliced proteins, double-stranded RNA-binding protein 76/NF90 (DRBP76/NF90), was found to contribute to VEGF expression under hypoxia, and silencing its expression reduced VEGF mRNA and protein levels. The role for DRBP76/NF90 in VEGF mRNA and protein synthesis under hypoxic conditions and in tumor progression was evaluated, and DRBP76/NF90 was shown to significantly contribute to hypoxia-induced VEGF expression and tumorigenesis in vivo in a breast carcinoma model.


arrow
MATERIALS AND METHODS
 
Cell lines and culture conditions. The human ductal carcinoma cell line MDA-MB-435S was obtained from ATCC (Bethesda, MD). Cells were cultured in Dulbecco's modified Eagle's medium (Gibco) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 10 U/ml penicillin, and 10 µg/ml streptomycin. Cell lines were cultured under either normoxic conditions (5% CO2, 21% O2, 74% air) or hypoxic conditions (5% CO2, 1.5% O2, 93.5% N2) in humidified triple gas incubators at 37°C.

Cell harvest and lysis. Following exposure to normoxia or hypoxia, conditioned media were collected and centrifuged at 14,000 x g for 10 min and supernatants stored at –80°C. Cells were washed three times with cold phosphate-buffered saline (PBS) and lysed in 1% Triton lysis buffer (1% Triton X-100, 50 mM HEPES [pH 7.5], 10 mM sodium pyrophosphate, 150 mM NaCl, 100 mM NaF, 0.2 mM sodium orthovanadate, 1 mM EGTA [pH 7.5], 1.5 mM MgCl2, 10% glycerol, and protease inhibitor cocktail [10 µl/ml] [Sigma]), disrupted by pipetting, and incubated on ice for 10 min. The lysates were separated by centrifugation at 14,000 x g for 10 min at 4°C and supernatants frozen at –80°C. Non-Triton-soluble fractions (crude nuclear pellets) were extracted with radioimmunoprecipitation assay (RIPA) buffer (1% NP-40, 0.1% sodium dodecyl sulfate [SDS], and 1% deoxycholate in PBS). Total cell lysates were provided by direct lysis with 5 volumes of RIPA buffer after three initial PBS washes.

Protein concentrations were determined using a Bio-Rad DC protein assay (Bio-Rad Inc.) according to the manufacturer's instructions, with bovine serum albumin as a standard protein.

Cell fractionation. Cells were washed twice in cold PBS and collected with a cell scraper in NP-40 lysis buffer (0.5% NP-40, 50 mM HEPES [pH 7.5], 10 mM sodium pyrophosphate, 150 mM NaCl, 100 mM NaF, 0.2 mM sodium orthovanadate, 1 mM EGTA [pH 7.5], 1.5 mM MgCl2, 10% glycerol, and protease inhibitor cocktail [Sigma] [10 µl/ml[). The NP-40 lysate was transferred to a Dounce homogenizer and further lysed using 10 strokes, incubated for 10 min on ice, and centrifuged at 3,000 x g for 10 min at 4°C. The crude nuclear pellets (after centrifugation at 3,000 x g) were salt extracted in lysis buffer (minus NP-40) with 0.25 M KCl by incubation for 30 min at 4°C and centrifugation at 13,000 x g for 20 min at 4°C, and supernatants were stored at –80°C (for nuclear extracts). The supernatant initially centrifuged at 3,000 x g was centrifuged at 20,000 x g for 30 min at 4°C and the supernatants (cytoplasmic and light membrane fractions) stored at –80°C. For affinity purification experiments, the supernatant initially centrifuged at 20,000 x g was centrifuged at 100,000 x g for 1.5 h at 4°C to obtain a primarily cytoplasmic soluble fraction without membranes.

RNA binding and UV cross-linking. RNA binding and UV cross-linking were performed as described previously (3). Briefly, 100 ng biotin-UTP-VEGF 3'-UTR 126-bp HSR RNA (5'-AGACACACCCACCCACATACATACATTTAT ATATATATATATTATATATATATAAAAATAAATATCTCTATTTATAT ATATAAAATATATATATTCTTTTTTTAAATTAACAGTGCTAATGTTA TT-3'), no RNA, or the HSR combined with the total VEGF 3'UTR or a control VEGF 3'-UTR sequence (CTRL1) (5'-GATGTATTTGACTGCTGTGGACT TGAGTTGGGAGGGGAATGTTCCCACTCAGATCCTGACAGGGAAGA GGAGGAGATGAGAGACTCTGGCATGATCTTTTTTTTGTCCCACTTG GTGGGGCCAGGGTCCTCTCCCCTGCCCAAGAATGTGCAAGGCCAG GGCATG-3') was incubated with 40 µg hypoxic 100,000 x g supernatant cell extract in up to 30 µl of RNA binding buffer, with or without 0.5% Triton X-100, containing 20 mM HEPES (pH 7.5), 5 mM MgCl2, 50 mM KCl, 0.5 mM EGTA (pH 7.5), 0.5 mM dithiothreitol, 10% glycerol, and fresh tRNA (0.067 mg/ml) (Sigma). The mixture was incubated for 20 min at 30°C and then UV cross-linked in a UV Stratalinker crosslinker (Stratagene, La Jolla, CA) for 15 min at room temperature. Streptavidin Magna beads were added and incubated on a rocker platform for 15 min at room temperature. Beads were magnetically separated and unbound protein was collected, the Magna beads were washed with PBS three times, and the bound protein was incubated with 1 µl of RNase T1 and 1 µg of RNase A for 10 min at room temperature. The sample was then denatured in SDS-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer under reducing conditions and separated with 10% SDS-PAGE and transferred to nitrocellulose. The blots were developed to detect biotin by use of streptavidin-horseradish peroxidase (streptavidin-HRP) conjugate (Sigma) and developed with Lumiglo substrate (K&P, Inc.) by use of BioMax LS film (Eastman Kodak) or imaged on a Kodak 2000MM multimodal imager.

Immunoblot analysis. Monoclonal antibody to human DRBP76 was used to detect DRBP76/NF90 and interleukin promoter enhancing factor 3 (ILF3)/NF110 (Transduction Laboratories). Immunoblottings were performed using a 1:1,000 dilution of anti-DRBP76 monoclonal antibody followed by a 1:3,000 dilution of HRP-conjugated goat anti-mouse immunoglobulin G (Amersham Pharmacia Biotech) in a blocking buffer containing either 5% bovine serum albumin or 5% nonfat milk and 0.1% Tween 20 in Tris-buffered saline. The blots were then developed using Lumiglo substrate (K&P, Inc.) on BioMax LS film (Eastman Kodak) or on a Kodak MM2000 multimodal imager.

Northern blot analysis. Total RNA was extracted using an RNeasy RNA extraction kit (Qiagen, Chatsworth, CA). Northern blot analysis was performed as described previously (3). Hybridization was carried out overnight at 65°C with [{alpha}32P]dCTP-labeled human VEGF 165 AccI/NcoI fragment (823 bp). A ribosome-associated protein cDNA probe, 36B4, was used as a loading control. Blots were washed at high stringency (1% SDS, 1x SSC [1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate] at 60°C) and exposed to Kodak MR film. Quantification was determined by densitometry using ImageQuant software.

Quantitative and qualitative RT-PCR. Reverse transcription (RT) was carried out with random hexamers as primers following the protocol of the manufacturer (Invitrogen). PCR amplification was carried out with Taq DNA polymerase or proofreading Platinum Pfx polymerase (Invitrogen). The qualitative PCR primers for three alternatively spliced VEGF isoforms as described were as follows: forward primer, 5'-ATGCGGATCAAACCTCACC-3'; reverse primer, 5'-ATCTGGTTCCCGAAACCCTG-3'. The GAPDH primers used were as follows: forward primer, 5'-CCACCCATGGCAAATTCCATGGCA-3'; reverse primer, 5'-TCTAGACGGCAGGTCAGGTCCACC-3'. All qualitative RT-PCR analyses were repeated at least twice to assure identical results. PCR products were resolved on a 0.75% agarose electrophoresis gel containing ethidium bromide and with Tris-acetate-EDTA buffer and imaged with the Kodak 2000MM multimodal imager. Quantitative RT-PCR (qRT-PCR) was performed as described previously (28). All quantitative analyses were performed in triplicate and normalized to human cyclophilin A or 36B4/RPLP0 (NCBI accession number NM_053275) values for each sample. The qRT-PCR primers used were as follows: cyclophilin A forward primer, 5'-CTGGACCCAACACAAATGGTT-3'; cyclophilin A reverse primer, 5'-CCACAATATTCATGCCTTCTTTCA-3'; DRBP76/NF90 forward primer, 5'-AACCATGGAGGCTACATGAAT-3'; DRBP76/NF90 reverse primer, 5'-CGCTCTAGGAAGACCCAAAATC-3'; ILF3/NF110 forward primer, 5'-GGCTCCTACTACCAAGGTGACAACT-3'; ILF3/NF110 reverse primer, 5'-TTATAGCCTTTCTGCTTGCCCT-3'; VEGF (all spliced forms) forward primer, 5'-CGAGGGCCTGGAGTGTGT-3'; VEGF reverse primer, 5'-GGCCTTGGTGAGGTTTGATC-3'; murine PECAM-1/CD31 forward primer, 5'-CAAGAGAAGCGGCCTGGGTAC-3'; murine PECAM-1/CD31 reverse primer, 5'-TTTTCGGACTGGCAGCTGAT-3'; RPLP0/36B4 (the ribosome-associated protein) forward primer, 5'-TCCTCGTGGAAGTGACATCGT-3'; RPLP0/36B4 reverse primer, 5'-CTGTCTTCCCTGGGCATCA-3'.

RNA interference. RNA interference oligonucleotides were designed using a gene-specific application (Design Center; Dharmacon) and synthesized as the following small interfering RNA (siRNA)-ready double-stranded RNA hybrids: ILF3 siRNA (5'-NNGGAUGUUGUCACAGCUAGU-3'), DRBP76 siRNA (5'-NNCAGCGUUGUUCGGCAUCAA-3'), and VEGF siRNA (5'-AACGAACGUACUUGCAGAUGU-3'). siRNAs were transfected into cells using Oligofectamine (Invitrogen) according to the manufacturer's protocol. Primary experiments with a dose-response curve evaluated optimal doses for transfection efficiency. Transiently transfected cells were allowed to recover for 24 h and were then incubated for 24 or 48 h under normoxic or hypoxic conditions. Cell-conditioned media were collected at the time of harvest, and cells were lysed in 1% Triton X-100 lysis buffer as described above. RNA was recovered from one-third of the Triton X-100 cellular lysate by use of an RNeasy kit (Qiagen). Protein (30 to 70 µg) was resolved on a 10% SDS-PAGE gel under reducing or nonreducing conditions and analyzed by immunoblotting for DRBP76NF90 and ILF3/NF110, with β-actin (Abcam, Cambridge, MA) as a loading control. The extracted RNA was run on a Northern blot and probed for VEGF mRNA by use of established protocols (3).

Stable small hairpin vector synthesis and cell selection. pSilencer 2.1-U6 neo vector (Ambion) was used as a base vector and modified to express green fluorescent protein (GFP) as follows: a fragment containing the cytomegalovirus (CMV) promoter and ZsGreen (Clontech) cDNA and an existing DRBP76/NF90 or ILF3/NF110 hairpin fragment were digested from pSirenRetroQ hairpin vector (Clontech) by use of BamHI and EcoRV digestion and ligated into the BamHI and BSAXI sites in pSilencer 2.1 U6-neo vector. The hairpin fragment sequence used for DRBP76/NF90 (developed using standard loop hairpin rules as suggested for pSirenRetroQ) was 5'-GATCCGCAGCGTTGTTCGGCATCAATTTTCAAGAGAAATTGAAGCCGAACAACGCTGCGGATCTTTTTT- 3'. Bacterial cells were transformed and selected with ampicillin. Vectors showing positive results were confirmed by direct sequencing, and MDA-MB-435 cells were transfected using Lipofectamine 2000 (Invitrogen). Cells were selected with 1 mg/ml Geneticin sulfate (G418) for at least four cell passages, and stable G418-resistant cells were then selected by cell sorting for maximal ZsGreen expression by use of a FACS 440 cell sorter.

Polysome isolation and analysis. Polysomes were isolated as previously described (21). MDA-MB-435 cells were grown to 90% confluence and subjected to 24 h of normoxia or hypoxia. Cells were treated with 0.1 mg/ml cycloheximide (CHX) for 3 min in normoxia or hypoxia and washed three times with PBS-CHX and then lysed with Triton X-100 (0.1%) buffer supplemented with 0.1 mg/ml CHX. Lysates were centrifuged at 10,000 x g, heparin (200 µg/ml) was added, and the lysates were recentrifuged. The lysate was layered on a 10 ml continuous sucrose gradient (20 to 50% sucrose in 15 mM MgCl-15 mM Tris [pH 7.4]-0.3 M NaCl) and centrifuged for 120 min at 190,000 x g in an SW41-Ti rotor at 4°C. Absorbance was read at 260 nm and 280 nm in 0.5 ml fractions. Fractions were processed for RNA isolation using a Qiagen RNeasy kit (Qiagen, Inc.) according to the manufacturer's protocol. Quantitative RT-PCR was performed as indicated above using human VEGF and ribosome-associated protein (36B4/RPL0) as a control. VEGF mRNA levels normalized to 36B4/RPL0 were determined for each fraction.

ELISA. A human VEGF sandwich enzyme-linked immunosorbent assay (ELISA) was performed according to a protocol used previously (37). VEGFs in conditioned media were normalized to the total protein extracted from wells by use of RIPA lysis buffer and quantification with a Bio-Rad DC protein assay kit (Bio-Rad).

Luciferase assay. MDA-MB-435S and MDA-MB-435 shDRBP76-GFP cells were transfected using a 5x HIF-1 promoter element luciferase reporter (3), luciferase control vector, or CMV-driven β-galactosidase vector and Lipofectamine 2000 (Invitrogen). Cells were recovered overnight and treated under either normoxia or hypoxia conditions for 24 h. Cells were harvested with a passive lysis buffer from an assay kit, and luciferase activity was measured according to the instructions of the supplier (Promega, Madison, WI). Luciferase activity was normalized to β-galactosidase expression as determined with a similar kit (Promega), and activity was normalized using total protein extracts.

Human orthotopic xenograft breast cancer model. Procedures for a xenograft tumor model were performed as previously described (1) using the MDA-MB-435-GFP cell line described previously (1, 38) and the shDRBP76-GFP described above. Cells (1 x 106) were injected subcutaneously in the mammary fat pad of female athymic (nu/nu) mice, and tumor growth was determined by external caliper measurements of tumor and calculated as width2 x length x 0.52 to approximate an ellipsoid structure. At harvest, tumors were divided into sections for immunohistochemical and molecular analysis. One third of the tumor was processed and embedded in paraffin. One third of the tumor was mechanically homogenized with a Tissuemiser (Fisher Scientific) in RNA lysis buffer and processed with an RNeasy kit (Qiagen). The remaining third of the tumor was embedded in optimal cutting medium and frozen at –80°C. Cryosectioning and hematoxylin and eosin staining were used to identify viable areas of the tumor, and those areas were excised from the frozen optimal cutting medium block using a 2 mm dermal biopsy core. Sample were processed for immunoblot analysis by homogenization with 1% Triton X-100 protein lysis buffer, incubated on ice for 10 min, and centrifuged at 14,000 x g for 10 min.

Immunohistochemistry. Paraffin-embedded tumor sections were postfixed and used for immunohistochemical analyses with antibodies to platelet endothelial cell adhesion molecule PECAM-1 (Pharmingen, San Diego, CA) and Ki67 (DakoCytomation) and for apoptosis using an ApopTag terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) kit (Intergen, Purchase, NY). Secondary detection was performed with appropriate biotinylated secondary antibodies and a VectaStain Elite kit (Vector, Inc., Burlingame, CA) with diaminobenzidine substrate. Counterstaining was performed with 1% methyl green. Negative-control slides were obtained by omitting the primary antibody. Sections from each tumor were also stained with hematoxylin and eosin. Images were quantified by counting the number of vessels per high-power field or viable numbers per total area by use of Image Pro Plus (Media Cybernetics, Silver Spring, MD) image analysis and recognition software.

Statistical analysis. Results from individual experiments are represented as means ± standard errors unless otherwise stated. Statistical comparison of groups was performed using two-tailed Student's t tests. Statistical significance was defined as P < 0.05 and P < 0.01 (see figures and legends).


arrow
RESULTS
 
UV cross-linking studies to assess protein binding to the VEGF 3'-UTR A/U-rich stem-loop element-HSR. Previous investigations have identified multiple proteins that can interact with the VEGF 3'-UTR to promote mRNA stability and intracellular transport under hypoxia conditions. These proteins have included HuR, AUF1, hnRNP A1, and hnRNP L, which are predominantly localized to the nucleus, although they appear to shuttle into the cytoplasm under stress conditions (6, 7, 9, 12, 33, 43). The A/U-rich stem-loop structure identified as the HSR was shown to confer mRNA stability to the VEGF mRNA under conditions of hypoxia and bound to hnRNP L directly (36). To address whether there might be additional RNA-binding proteins that could function to promote VEGF mRNA stability, we analyzed cytoplasmic and nuclear extracts from MDA-MB-435 human breast cancer cells for evidence of direct binding to the HSR element. Cell extracts were obtained from cultures exposed to normoxic or hypoxic (1.5% O2) conditions for 24 h and then UV cross-linked to a biotin-labeled VEGF 3'-UTR HSR element. Figure 1A shows the results of SDS-PAGE analysis of proteins that were cross-linked to biotinylated HSR RNA as detected by streptavidin-HRP activity. The extracts from the nuclear compartment of MDA-MB-435 cells cultured under normoxic or hypoxic conditions showed multiple proteins cross-linked to the HSR element. The most prominent of these were induced by hypoxia with apparent molecular masses of ~46 kDa and ~60 kDa. The 60-kDa protein has already been characterized as representing hnRNP L, a nuclear RNA-binding protein (36), and the 46-kDa band is likely to represent HuR, although this has not been confirmed directly. Interestingly, the light membrane-cytoplasm extracts demonstrated a strong doublet at an apparent molecular mass of 90 to 88 kDa that was described previously in a report of a study using total cell lysates and a complete VEGF 3'-UTR as a probe (3). The level of signal for this species was increased in hypoxic cell extracts to an extent similar to that seen for the nuclear proteins. To establish the specificity of protein binding to the HSR element, we performed a similar cross-linking experiment using biotinylated HSR RNA combined with a 10-fold molar excess of unlabeled competitor RNAs (Fig. 1B). The unlabeled HSR and full-length VEGF 3'-UTR RNAs both effectively competed for the 90- to 88-kDa UV-cross-linked signal in hypoxic cell extracts. A competitor RNA with a size similar to that of the HSR (126 bp), CTRL1, which is adjacent to the HSR element in the VEGF 3'-UTR, did not compete for the 90- to 88-kDa species. No UV cross-linking was observed when cell extracts were preheated to 100°C for 5 min, suggesting that the cross-linked signals are likely proteins that can be heat inactivated (data not shown). These data are consistent with previous observations indicating that the VEGF 3'UTR HSR element conferred mRNA stability to a heterologous luciferase reporter under hypoxia conditions whereas the control adjacent sequence, CTRL1, could not (3).


Figure 1
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 1. Identification of RNA-binding proteins that interact with human VEGF 3'-UTR stem-loop elements (HSR). (A) UV cross-linking of 24 h normoxic (Norm) or hypoxic (Hyp) MDA-MB-435 cell cytoplasmic/light-membrane extracts (C) or crude nuclear/heavy-membrane extracts (N) to the VEGF 125 bp 3'-UTR HSR element. Biotinylated signals were identified with streptavidin-HRP and are indicated with reference to the location of molecular mass markers. The arrow indicates the location of the predominant cytoplasmic binding protein. (B) Hypoxic extracts from MDA-MB-435 cells were used for UV cross-linking to the biotinylated HSR (HSR*) with a 10-fold molar excess of competitor RNAs and to unlabeled HSR (HSR*+HSR), the full-length VEGF 3'UTR (HSR*+3'UTR), and a 125-bp VEGF 3'-UTR control element which does not have A/U elements (HSR*+CTRL1). (C) Affinity purification of MDA-MB-435 hypoxia cell extracts with the VEGF 3'UTR 125-bp HSR element. Unbound protein and bound protein were analyzed by UV cross-linking to biotinylated HSR by use of SDS-PAGE/streptavidin-HRP detection; apparent molecular masses at 60 kDa, 90 kDa, and 100 kDa are indicated. A duplicate gel with silver-stained bands was run and the 90-kDa band excised in the pattern indicated (dashed lines). Note that visible silver-stained bands at 90 kDa and 60 kDa were detected.

RNA affinity chromatography purification and proteomic analysis of RNA-binding protein recognizing the VEGF 3'-UTR HSR element. To identify the 90- to 88-kDa HSR-binding protein(s) in hypoxic cell extracts, we employed RNA-affinity chromatography using the biotinylated-HSR element as bait. Nonionic detergent (Triton X-100) cell extracts were allowed to bind to the VEGF 3'-UTR HSR for 30 min prior to the addition of streptavidin-Magna beads. Unbound proteins were collected after magnetic separation, and the bound material was washed several times. Bound proteins were eluted from the beads and inserted into neutralization buffer by use of rapid acidification. Unbound and bound proteins were then tested for the presence of HSR-binding proteins by UV cross-linking to the biotin-labeled VEGF HSR element (Fig. 1C). The affinity-selected proteins were found to include several species similar to those observed in total cell lysates; these were the 60-kDa hnRNP L band (confirmed by direct immunoblot analysis of the affinity-bound proteins; data not shown) and a distinct 90-kDa protein as well as a minor band at apparent molecular mass of approximately 100 kDa. The signals were matched to their molecular mass locations on duplicate silver-stained gels and excised from the gel aligned with their respective UV cross-linking signals. This procedure was replicated using three separate MDA-MB-435 cell lysates, and the gel slices were pooled and analyzed by in-gel digestion with trypsin and liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis and peptide identification using human databases and the SEQUEST algorithm (16). The 90-kDa species was identified as ILF3/DRBP76 by matching 13 peptides within the predicted amino acid sequence in three different runs, with 5 of those peptides matched to both alternatively spliced isoforms of the DRBP family (Table 1). VEGF HSR affinity chromatography and LC-MS/MS identification of this 90-kDa species as DRBP was also confirmed using hypoxic extracts from the human glioma cell line U373 (data not shown). These data indicate that the DRBPs are primary candidates binding to the VEGF stem-loop HSR under conditions of hypoxia and represent an approximate molecular mass of 90 kDa, implying that DRBP76/NF90 may be the RNA-binding protein of interest in hypoxic cell extracts.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Proteomic LC-MS/MS analysis of 90-kDa banda

The DRBP76/NF90 and the ILF3/NF110 proteins differ in the C-terminal domain, where the ILF3/NF110 isoform has an extended C-terminal domain with 25 potential myristoylation sites which are present in one cluster, as determined by analysis with ProSite (Fig. 2A). Thus, it is likely that the proteins can selectively fractionate into cellular membranes, depending upon the posttranslational modification of these domains. In addition, modification of these sites may alter availability during hypoxic stress. To assess these possibilities, we analyzed cell extracts for the presence of the DRBP proteins by immunoblotting using nonionic detergent-soluble extracts (Triton X-100) or whole-cell extracts and RIPA lysis buffer (Fig. 2B). The nonionic detergent extraction was unable to effectively extract the ILF3/NF110 isoform under normoxic or hypoxic conditions. DRBP76/NF90, however, was able to be extracted into a soluble Triton X-100 fraction under these conditions, and its level was moderately increased under conditions of hypoxia. DRBP76/NF90 shows variable increases in Triton X-100 extracts under hypoxic conditions ranging from 30% up to a twofold increase compared to normoxic conditions, depending upon cell density, extraction procedure, and length of hypoxic exposure (data not shown). In contrast to the nonionic detergent extraction results, the RIPA whole-cell lysate demonstrated significant levels of ILF3/NF110 protein as well as of DRBP76/NF90, neither of which was modulated at the protein level under hypoxic conditions when normalized to the β-actin loading control results.


Figure 2
View larger version (27K):
[in this window]
[in a new window]

  		FIG. 2. DRBP protein structure and potential myristoylation sites demonstrate differential cell compartmentalization characteristics. (A) Structural locations of the predicted myristoylation sites in DRBP76 and ILF3 proteins as determined by ProSite analysis (vertical bars indicate potential sites). aa, amino acids. (B) Identification of ILF3 and DRBP proteins in MDA-MB-435 cell extracts obtained under normoxic or hypoxic conditions. Triton X-100 (Tx) and RIPA (R) extracts were assayed for the presence of DRBP proteins by immunoblotting. The locations of molecular mass markers are indicated. β-actin was detected after stripping the DRBP signal.

DRBP76/NF90 and ILF3/NF110 depletion by RNA interference affects VEGF expression under hypoxia. Since both the DRBP/NF90 and the larger alternatively spliced form ILF3/NF110 could potentially affect VEGF expression under hypoxia, we selectively depleted either isoform by use of RNA interference to determine their contributions to VEGF production. As shown in Fig. 3A, the targeted siRNA oligonucleotides selectively repressed the DRBP76/NF90 or the ILF3/NF110 isoforms, as determined by immunoblot analysis of total cell lysates obtained at 72 h posttransfection. Mock transfections or cells transfected with siRNA targeted to VEGF did not show alterations in the levels of the DRBP76/NF90 or ILF3/NF110 proteins.


Figure 3
View larger version (47K):
[in this window]
[in a new window]

  		FIG. 3. RNA interference with respect to DRBP isoforms affects hypoxia-induced VEGF expression in MDA-MB-435 cells. (A) MDA-MB-435 cells were treated with transfection reagent only (Mock) or with siRNA oligonucleotides (100 nM) for human VEGF, DRBP76/NF90, or ILF3/NF110. At 48 h posttransfection, cells were exposed to 24 h of hypoxia and cell extracts were analyzed by immunoblotting for the presence of DRBP protein isoforms and of β-actin as a control. The locations of molecular mass markers are indicated. (B) Hypoxia-induced VEGF secretion into media from the same experiment whose results are presented in panel A. Statistically significant differences from mock-treated hypoxia control results are indicated (n = 3; ** = P ≤ 0.01).

To determine whether the DRBP isoform-specific silencing affected VEGF expression under conditions of hypoxia, we evaluated levels of VEGF secreted into conditioned media of the treated cells (Fig. 3B). Hypoxia-induced VEGF was found to be significantly repressed by siRNA oligonucleotides targeting VEGF compared to the mock-transfected control results. Interestingly, the DRBP76/NF90-targeted cells also repressed VEGF production by approximately 60% whereas the ILF3/NF110-targeted cells only moderately reduced VEGF synthesis and secretion by only 15% of mock control results, although this was statistically significant. These data indicate that there was some selectivity with respect to the DRBP76/NF90 contribution to hypoxia-induced VEGF synthesis in MDA-MB-435 cells, most likely due to its abundance being greater than that of the ILF3/NF110 isoform.

Stable small hairpin interference of DRBP76/NF90 expression in MDA-MB-435 cells affects hypoxia-induced VEGF mRNA and protein expression. Since transient RNA interference of proteins has to be verified for effectiveness in each individual experiment and since the experimental time window is limited to less than 1 week, we employed a small hairpin expression system to repress DRBP76/NF90 protein expression in a stable manner. To facilitate screening of the stable integrated vector we employed a U6-promoter-driven small hairpin RNA (shRNA) system with neomycin resistance to which a GFP expression cassette was added (see Materials and Methods). Primary stable transfected pools of MDA-MB-435 cells were selected for drug resistance alone and demonstrated an approximately 50% repression of the DRBP76/NF90 protein isoform (data not shown). In an effort to improve upon the pool of cells with uniform DRBP76/NF90 repression, we performed double cell sorting to select cells with high-level GFP expression (shDRBP76-GFP; Fig. 4A).


Figure 4
View larger version (40K):
[in this window]
[in a new window]

  		FIG. 4. Characterization of stable repression of DRBP76/NF90 expression by shRNA and loss of HSR binding in MDA-MB-435 cells. (A) Cell sorting to select for GFP-expressing cells derived from stable transfectants with the shRNA expression vector for DRBP76 with a ZsGreen reporter as described in Materials and Methods. Cell sorter gating was defined by use of parental MDA-MB-435 cells (MDA-MB-435 pre-sort) and the presorted shRNA-expressing MDA-MB-435 cells (shDRBP76-GFP pre-sort) (upper panels). The results for cells negatively selected for GFP are shown in the bottom left panel [shDRBP76-GFP (–)] and for those positively selected [shDRBP76-GFP (+)] in the bottom right panel. Assessment of the P2 protein pool with respect to shDRBP76-GFP-positive cells showed a GFP expression level of 96%. (B) The expression of DRBPs in cell lysates from the stable selected cells was evaluated by immunoblotting cell extracts under normoxic or hypoxic conditions. The presence of the DRBP76/NF90 protein isoform (arrow) was not detectable in the Triton (Tx) lysates; minor amounts were present in the RIPA (R) extracts. However, the ILF3/NF110 isoform was unaffected and was present in the RIPA lysates in amounts similar to those shown for the parental MDA-MB-435 cells. (C) Cell extracts were obtained from MDA-MB-435 cells or the stable selected shDRBP76-GFP DRBP76-silenced cells cultured for 24 h under normoxic (N) or hypoxic (H) conditions. UV cross-linking (UV-X Link) to the biotin-labeled HSR element is shown in the top panel. The same extracts were analyzed for the presence of DRBP76 and β-actin protein by immunoblotting (IB).

The effectiveness of the shDRBP76 inhibition was determined by immunoblot analysis of both Triton X-100 and RIPA cell extracts under conditions of normoxia and hypoxia. Figure 4B shows the analysis of extracts from the shDRBP76-GFP cells; there was no discernible DRBP76/NF90 protein present in the Triton X-100-soluble extracts obtained under normoxic or hypoxic conditions. However, with total cell RIPA lysates, the DRBP76/NF90 isoform was detected at a trace level in the shDRBP76-GFP cells, suggesting that only minor protein production was possible. Estimating protein abundance using normalized β-actin values obtained with the parental and the shDRBP76-GFP extracts showed the level of repression to be 90% or greater. In contrast to the DRBP76/NF90 isoform results, the ILF3/NF110 protein isoform was still present at appreciable levels, with a minor reduction of approximately 10% from the parental nontransfected cell levels. A similar analysis of MDA-MB-435 cells stably selected for Geneticin sulfate resistance and expressing GFP alone (MDA-MB-435-GFP) showed DRBP76/NF90 and ILF3/NF110 expression levels identical to those seen with those of the parental MDA-MB-435 cells whose results are shown here; thus, drug resistance and selection for GFP overexpression did not affect DRBP levels (data not shown).

To evaluate whether DRBP76 silencing affected VEGF mRNA binding capacity, we performed VEGF-HSR UV cross-linking experiments with cell extracts from parental MDA-MB-435 cells and the shDRBP76-GFP cells exposed to normoxic or hypoxic culture conditions. Figure 4C shows a loss of UV cross-linking signal for DRBP76/NF90 in the shDRBP76-GFP cells which correlates with a distinct band in the parental cell extracts. To control for the level of proteins in the gel, direct immunoblot analyses were subsequently performed using the UV-cross-linked blots and antibodies to DRBP/ILF3 and β-actin as internal controls. The DRBP76/NF90 isoform was clearly depleted from the shDRBP76-GFP cell extracts under both normoxic and hypoxic conditions, and the β-actin signal levels were comparable for all lanes. Interestingly, there was some increase in binding capacity for the ILF3/NF110 isoform, which may indicate an increase in available ILF3 protein in the shDRBP76-GFP cells.

DRBP76/NF90 silencing in MDA-MB-435 cells shows reduced VEGF mRNA expression, little difference in HIF-1-dependent transcriptional activity, and a reduced VEGF mRNA half-life (t1/2) under hypoxia conditions. Parental MDA-MB-435 cells and the shDRBP76-GFP cells with stable repression of the DRBP76/NF90 isoform were subjected to normoxic or hypoxic conditions for 20 h. Total RNA was isolated and Northern blot analysis performed in triplicate to assess VEGF mRNA levels. Figure 5A shows the induction of VEGF mRNA under hypoxia conditions for the MDA-MB-435 cells. The levels of VEGF mRNA under normoxic and hypoxic conditions in the shDRBP76-GFP cells were lower than those seen with the parental MDA-MB-435 cells. The hypoxia induction of VEGF mRNA was also muted in these cells, with a hypoxic increase of VEGF/36B4 signal of 0.091 compared to 0.175 for the parental cells. A ribosomal housekeeping gene, 36B4/RPLP0, was found not to be adversely affected in the shDRBP76-GFP cells and was thus used as an internal control.


Figure 5
View larger version (40K):
[in this window]
[in a new window]

  		FIG. 5. Repression of DRBP76/NF90 affects VEGF mRNA accumulation and stability under hypoxia conditions without altering HIF-1-dependent transcriptional activity. (A) Northern blot analysis of VEGF mRNA in parental MDA-MB-435 cells and shDRBP76-GFP cells. RNA isolation and Northern blotting were performed using triplicate plates, and the results obtained with each cell line were analyzed after 24 h of normoxia or hypoxia. Signals for VEGF and 36B4/RPL19 mRNAs are shown. Quantification of VEGF with respect to 36B4 signal levels for each lane was averaged for three lanes; the results are represented as averages ± standard deviations. *, statistically significant (P ≤ 0.05) differences from parental MDA-MB-435 cell results for each set of conditions. (B) Transient transfection analysis of HIF-1-dependent luciferase transcription in parental MDA-MB-435 or shDRBP76-GFP cells. Cell lines were cotransfected in triplicate with CMV-β-galactosidase (β-Gal) and HIF-1-driven luciferase as described in Materials and Methods. Cells were allowed to recover for 24 h and were placed under normoxic or hypoxic conditions for 24 h. Luciferase and β-Gal activity levels were analyzed using triplicate plates. The level of luciferase activity was normalized to that of β-Gal activity for each cell line (n = 3). (C) MDA-MB-435 or shDRBP76-GFP cells were exposed to hypoxia for 20 h, treated with actinomycin D, and maintained under hypoxia conditions. The VEGF/36B4 signal ratio was quantified by qRT-PCR at time 0 and at 1 and 2 h after actinomycin D treatment. VEGF stability was calculated by determination of the slope appearing over the 2-h time point at which the MDA-MB-435 cell line exhibited greater stability of VEGF mRNA than the shDRBP76-GFP cell line. *, P ≤ 0.05.

To address whether the DRBP76/NF90-depleted cells were deficient under hypoxia with respect to transcriptional pathways necessary for VEGF expression, we transiently transfected parental or shDRBP76-GFP cells with an HIF-1-dependent luciferase transcriptional reporter. Transfected cells were then exposed to normoxic or hypoxic conditions and luciferase values determined from cell lysates normalized to β-galactosidase and cotransfected as a control. The parental MDA-MB-435 cells demonstrated a threefold increase in HIF-1-dependent luciferase activity under hypoxia conditions (Fig. 5B). The shDRBP76-GFP cells showed a similar induction of HIF-1-specific reporter activity, with no observable differences under normoxic or hypoxic conditions.

To assess whether VEGF mRNA stability may be affected, we performed a transcriptional blockade and analysis of VEGF mRNA decay under conditions of hypoxia (Fig. 5C). Parental MDA-MB-435 cells demonstrated increased mRNA stability as expected under hypoxia at a t1/2 of 2.97 h, whereas the t1/2 of VEGF mRNA in the shDRBP76-GFP cells was determined to have been reduced in a statistically significant manner to 1.72 h. These data suggest that VEGF mRNA t1/2 was affected by the loss of DRBP76/NF90 function.

VEGF mRNA loading onto active polysomes is affected by loss of DRBP76/NF90 expression. To achieve effective VEGF protein synthesis under hypoxia, mRNAs must be loaded onto active polysomes for protein translation, as observed with several other hypoxia-induced proteins, especially during prolonged hypoxia (20). Despite the significant depletion of total RNAs present on polysomes under hypoxia, VEGF mRNA is selectively associated with polysomes. Thus, to determine whether VEGF mRNA loading onto polysomes is one potential function for DRBP76/NF90, we isolated polysome fractions from parental and shDRBP76-GFP cells under normoxic and hypoxic conditions. The polysome nucleic acid and protein profiles look similar for the parental MDA-MB-435 and the shDRBP76-GFP cells, as shown in Fig. 6A. The normoxic sucrose gradient profiles show significant nucleic acid and protein levels in the polysome region between fractions 5 and 10. In contrast, the hypoxia profiles show increased nucleic acid in the light regions (fractions 2 to 5) and a significant deprivation of nucleic acid in the polysome regions (fractions 5 to 10). The levels of polysome reduction in total nucleic acid in hypoxic cells appeared to be equivalent for the two cell lines; thus, there was no significant change in the overall polysome protein and RNA profiles in the shDRBP76-GFP cells.


Figure 6
View larger version (28K):
[in this window]
[in a new window]

  		FIG. 6. VEGF mRNA polysome loading is decreased in shDRBP76-GFP cells compared to parental MDA-MB-435 cell results under hypoxic conditions. (A) Cell polysome profiles for MDA-MB-435 parental and shDRBP76-GFP cells under normoxic and hypoxic conditions. Gradient fractions represent increasing depths; nucleic acid (260 nm) and protein (280 nm) values are provided for each sample. The locations of 80-s fractions are indicated by arrows. (B) Analysis by qRT-PCR of polysome fractions A and B for MDA-MB-435 and shDRBP76-GFP cells under normoxic (Norm) and hypoxic (Hyp) conditions. Equal amounts of RNA were reverse transcribed and quantified for all isoforms of VEGF by qRT-PCR and were compared to levels of 36B4/RPLP0 mRNA for each fraction. Statistically significant differences from parental MDA-MB-435 cell results are indicated. **, P ≤ 0.01.

To evaluate whether polysome-associated VEGF mRNA could be selectively affected in the shDRBP76-GFP cells, total RNAs from hypoxic polysomes were isolated from either the free RNA fractions or the separated polysomes in the gradients of both the parental and the shDRBP76-GFP cell lines. RNA was analyzed for total VEGF mRNA by use of quantitative RT-PCR. Figure 6B shows that VEGF mRNA levels were very low in the free RNA pool (A fractions) and were mostly associated with the polysome fractions (B fractions). Interestingly, under normoxic conditions VEGF mRNA was detected in the polysome fractions of both cell lines. However, under hypoxic conditions and as expected, the VEGF mRNA levels were elevated in the parental cells; however, the mRNA was found to be less associated with polysomes in the shDRBP76-GFP cells. Thus, it appears that stress-induced mRNA loading onto the polysomes was impaired in the shDRBP76-GFP cells.

Overexpression of DRBP76/NF90 by MDA-MB-435 cells promotes both constitutive and hypoxia-induced VEGF mRNA and protein synthesis. To assess whether providing additional DRBP76/NF90 protein to cells could affect the induction of VEGF in human breast carcinoma cells, we transiently transfected MDA-MB-435 cells to overexpress the DRBP76/NF90 protein and tested the expression of VEGF mRNA by quantitative RT-PCR analysis (Fig. 7A). The transient transfection and analysis of cells with either β-galactosidase-expressing vector or vector encoding DRBP76/NF90 showed that VEGF mRNA levels were increased in the DRBP76/NF90-expressing cells under both normoxic and hypoxic conditions. To confirm whether the mRNA expression would translate into increased protein production, we examined secreted VEGF protein from transient transfected cells; the results demonstrated that VEGF protein accumulated in the media of transfected cells and that levels increased under both normoxic and hypoxic conditions in cells expressing DRBP76/NF90 compared to control cell results (Fig. 7B).


Figure 7
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 7. Transient overexpression of the DRBP76 isoform increases VEGF mRNA and protein expression. Results represent transient transfection of MDA-MB-435 cells with a control β-galactosidase expression vector (β-gal control) or a DRBP76 expression vector (DRBP76 pcDNA3.1) in which cells were allowed to recover for 30 h and were then exposed to normoxic or hypoxic conditions for 24 h. (A) Total RNA was analyzed by qRT-PCR for VEGF mRNA levels normalized to ribosome-associated protein mRNA (36B4/RPL0) (n = 3). *, statistically significant (P ≤ 0.05) in comparison to β-gal control results. (B) Conditioned medium from each sample was analyzed for secreted VEGF protein by ELISA normalized to total cellular protein. **, statistically significant (P ≤ 0.01) in comparison to β-gal control results.

DRBP76/NF90 depletion affects tumor progression of MDA-MB-435 cells in orthotopic xenografts in athymic nude mice. The DRBP76/NF90-silenced cells demonstrated reduced hypoxia-induced VEGF mRNA levels and polysome-associated VEGF mRNAs in vitro; however, it is possible that reducing DRBP76/NF90 levels could affect multiple cell functions. In an effort to determine whether DRBP76/NF90 depletion affected cell viability and proliferation, we seeded parental MDA-MB-435 cells and the DRBP76/NF90-silenced pool at low density and assessed cell viability after seeding and culturing for 24 h and 48 h using a standard MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] activity assay. Table 2 shows that the increases in MTT activity were essentially identical at 24 h for the two cell lines and were slightly lower at 48 h. To assess whether the DRBP76/NF90 repression could affect cell viability under conditions of hypoxic stress alone, we performed a similar assay in which cells were exposed to hypoxia for 24 and 48 h. Again, there was no significant difference between the parental and the shDRBP76-GFP cell results, since the results for both showed significant reductions in viability over time (Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Cell proliferation as a measure of MTT activity under conditions of normoxia and hypoxia repression of DRBP76/NF90 does not dramatically affect cell proliferation or hypoxic repression of mitochondrial MTT activitya

Since the shDRBP76-GFP cells showed properties similar to those of the parental MDA-MB-435 cells transfected to express GFP alone in vitro, we were able to investigate whether depletion of DRBP76/NF90 protein could alter the tumorigenic and/or angiogenic properties in vivo by use of an orthotopic model of breast tumor growth by implanting tumor cells in the mammary fat pad of athymic mice. Tumor growth over time was assessed with external caliper measurements twice weekly (plotted in Fig. 8A). The MDA-MB-435-GFP and the shDRBP76-GFP tumors showed similar growth patterns during the first several weeks after implantation and for up to 28 days. Interestingly, the shDRBP76-GFP tumors reached a plateau at 35 days postimplantation and demonstrated little progression after that point. MDA-MB-435-GFP tumors grew to at least twice the size of the shDRBP76-GFP tumors, and this difference was statistically significant beginning at day 42. Histological analysis of tumors showed large viable tumors for the MDA-MB-435-GFP and small thin restricted tumors for the shDRBP76-GFP pools (Fig. 8B). To determine whether the DRBP76/NF90 protein was still repressed by the hairpin expression construct in vivo, we extracted a defined region of viable tumor from two tumors representing each group (Fig. 8C). Analysis of the protein lysates by immunoblotting demonstrated that the parental MDA-MB-435-GFP tumors had the DRBP76/NF90 and the ILF3/NF110 isoforms at levels approximately equivalent to those observed for cells in vitro (Fig. 8C). In contrast, the shDRBP76-GFP tumors demonstrated the presence of ILF3/NF110 and low levels of DRBP76/NF90 protein. Quantification of the viable tumor areas revealed that the shDRBP76-GFP tumors had increased necrosis per total area; however, there was enough variability between tumors so that the results were not quite statistically significant (data not shown). Analysis of microvascular density by use of PECAM-1/CD31 staining for endothelial cells in the four largest tumors from each group indicated that the shDRBP76-GFP tumors had significantly reduced levels of PECAM-1/CD31 cell staining and that vascular networks or luminal structures were limited or absent altogether (Fig. 8D).


Figure 8
View larger version (37K):
[in this window]
[in a new window]

  		FIG. 8. An orthotopic mammary tumor model using MDA-MB-435-GFP and shDRBP76-GFP cells demonstrated that loss of DRBP76 affected tumor growth, increased necrosis, and reduced microvascular density and levels of VEGF expression. (A) Tumor growth values determined by external caliper measurements are expressed as average tumor volume per group (n = 8). Note that the divergence in tumor growth did not occur until day 35. (B) Representative low-power micrographs of hematoxylin and eosin-stained sections of MDA-MB-435-GFP and shDRBP76-GFP tumors. Note the differences in tumor densities and relative sizes. Necrotic regions in these images are indicated (N). Bar, 500 µm. (C) Immunoblot analysis of tumor lysates for DRBP proteins. Two tumors representing each group were used to core only regions that were viable as indicated in representative images (dashed circles) (bottom panels). Triton X-100 protein lysates were prepared and evaluated by immunoblotting for DRBP76/NF90 and ILF3/NF110, with β-actin as a loading control. (D) Representative immunohistostaining for PECAM-1/CD31 in viable tumor areas for MDA-MB-435-GFP and shDRBP76-GFP tumors (positive staining is indicated by brown microvessels). Note the abundance of luminal vessels in the MDA-MB-435-GFP tumors (arrows) and nonluminal cells in shDRBP76-GFP tumors (arrowheads). Quantification of the results of PECAM-1/CD31 staining in the five largest tumors from each group is indicated as the average numbers of vessels per high-power field (HPF) ± standard deviations. Bar, 100 µm. *, P ≤ 0.05. (E) Quantitative RT-PCR analysis of the alternatively spliced DRBP76/NF90 (DRBP76) and ILF3/NF110 isoforms as well as total VEGF and murine PECAM-1/CD31 (mCD31) from total RNA isolated from viable tumor areas for each group (n = 4). Signals were normalized to human cyclophilin A signal to assure normalization to viable tumor levels. Statistically significant differences from MDA-MB-435-GFP tumor results are indicated by P values.

Since VEGF expression was repressed in the shDRBP76-GFP cells in vitro, the tumors were analyzed for mRNAs encoding the DRBP isoforms, VEGF, and murine PECAM-1/CD31 to obtain a quantitative assessment of their relative levels within the viable tissue for five representative samples for each group. Figure 8E shows that the DRBP76/NF90 isoform was significantly repressed in the shDRBP76-GFP tumors, thus confirming that in the tumors which did grow, albeit at reduced levels, the tumors did not revert to DRBP76-expressing clones. Interestingly, the ILF3/NF110 mRNA was aberrantly expressed at extremely high levels nearly eightfold greater than the level in parental MDA-MB-435-GFP tumors, suggesting that some alternative splicing may be forced upon the cells with respect to DRBP/ILF3 mRNA processing. The levels of the ILF3/NF110 protein did not reflect this high level of mRNA, as determined by immunoblotting. VEGF mRNA was also significantly repressed in the shDRBP76-GFP tumors compared to the control tumor results (Fig. 8E). As expected from the histological analysis, murine PECAM-1/CD31 levels were significantly repressed in the shDRBP76-GFP tumors, which confirms the angiogenic repression observed histologically.

In addition to these molecular and histological analyses, markers of cell proliferation and tumor cell apoptosis were evaluated using Ki67 and TUNEL staining, respectively. In situ quantification of these endpoints in the MDA-MB-435-GFP and shDRBP76-GFP tumors revealed significant variations within each tumor and between the tumors in each group. Thus, these analyses did not provide data that would indicate proliferation or apoptotic alterations in the shDRBP76-GFP group, although these possibilities cannot be ruled out completely (data not shown). Thus, the hypoxic response in these tumors with respect to VEGF expression appears to be significantly affected by loss of DRBP76/NF90 protein and likely contributes to reduced angiogenic response and tumor growth potential.


arrow
DISCUSSION
 
Cellular adaptation to hypoxia is a complex process that is initiated by a reduction in the ability to maintain ATP levels via oxidative phosphorylation. One of the crucial events necessary for cell survival under conditions of nutrient depletion is the preservation of energy consumption by limiting biosynthetic processes such as protein synthesis. In addition, the key survival genes that function to increase nutrient availability and facilitate glycolytic metabolism are actively expressed. The enzymes and nutrient transporters that make up the glycolytic pathway such as glucose transporter 1, hexokinase, and phosphoglycerate kinase are selectively expressed. VEGF, a vascular permeability and angiogenic factor, is expressed as an adaptive response to both acute and chronic hypoxia. Given the significant repression of total protein synthesis in hypoxic conditions, which is estimated to be between 15% and 30% of normoxic levels, the cellular mechanisms that promote transcription, mRNA splicing, mRNA nuclear export, and translation of these adaptive genes are likely to be highly selective (21).

VEGF expression is known to be facilitated by transcriptional and posttranscriptional mechanisms. Transcriptional activation through HIF-1-dependent pathways has been well established. In addition, it has been established that mRNA stabilization also contributes to VEGF expression under hypoxic conditions and VEGF mRNA stability has been shown to be provided for by a number of mRNA-binding proteins, including HuR, hnRNP L, and hnRNP A1 proteins, all of which are predominantly nuclear. Effective mRNA export from the nucleus and loading onto active polysomes under hypoxia has been attributed to increases in extranuclear shuttling of these nuclear mRNA-binding proteins. Previous studies investigating VEGF 3'-UTR structure have shown that a predicted stem-loop structure participates in VEGF mRNA stability and contributes to increased protein expression under hypoxia conditions, thus defining a stability element in the VEGF 3'-UTR (3).

In this report we have identified the double-stranded RNA-binding protein isoform DRBP76/NF90 as a DRBP that binds to the VEGF 3'-UTR stem-loop hypoxia stability region. Although the DRBP family of proteins have multiple alternatively spliced isoforms (see references 29 and 34 for reviews), we determined that the 90-kDa isoform is abundant and promotes hypoxia-induced VEGF mRNA levels and protein translation. Interestingly, this isoform was present in both nuclear and cytoplasmic compartments whereas the ILF3/NF110 isoform was retained in the nucleus and nuclear-associated membranes, as determined by cell fractionation studies. The observation that DRBP76/NF90 isoform repression did not dramatically affect HIF-1-dependent transcription indicates that DRBP76/NF90 functions as a posttranscriptional regulator.

In support of DRBP function in mRNA stability, the loss of DRBP76/NF90 reduced hypoxia-dependent VEGF mRNA stability, as determined by t1/2 studies, as well as the association of VEGF mRNA with active polysomes. In contrast, overexpression of DRP76/NF90 increased VEGF mRNA and protein levels in human breast cancer cells, suggesting that misregulation of the DRBP proteins could also promote VEGF synthesis in human breast cancer cells, although overexpression or misregulation of DRBPs has not been established as a mechanism of increased VEGF expression to date.

The ability to repress the expression of the DRBP76/NF90 isoform in stable cell lines was useful in determining its role in VEGF expression under metabolic stress conditions. These cells demonstrated similar proliferation rates under optimal growth conditions. However, hypoxia-dependent VEGF protein synthesis and secretion was significantly affected, suggesting that posttranscriptional and translational processes are facilitated by DRBP76/NF90 activity. Thus, the DRBP76/NF90 protein isoform is not essential to cell viability or proliferation.

Given the observation that DRBP76/NF90 repression did not dramatically affect cell growth under optimal conditions and despite the reduction in VEGF expression under hypoxia, the limitations in tumorigenic potential for the shDRBP76-GFP cells were indeed dramatic. The effects of DRBP76/NF90 repression on tumor expansion and viability were significant on several fronts. First, the tumors had a defined time point (about 4 weeks) at which growth was significantly affected, resulting in a tumor size of approximately 75 mm3. Second, at harvest, those shDRBP76-GFP tumors that persisted still demonstrated reduced DRBP76/NF90 expression, with a corresponding reduction in VEGF expression. Finally, histological and molecular analysis revealed that the tumors had reduced viability, which coincides with the reduced angiogenic potential. The effect of DRBP76/NF90 isoform repression on tumor growth indicated that although the shDRBP76-GFP cells were viable for proliferation under optimal growth conditions, their stress response, including the ability to actively produce VEGF, was a primary defect that prevented tumorigenic progression in vivo.

These data demonstrate that the DRBP family of proteins, and the DRBP76/NF90 isoform in particular, significantly contributes to the regulation of VEGF mRNA transport and stability and facilitates translation under hypoxic conditions. This mechanism appears to selectively target stress-induced pathways, since the DRBP76/NF90 isoform was not essential to normal cell proliferation in vitro. These data also support the hypothesis that selective mechanisms and processes exist to facilitate the expression of hypoxia-inducible genes under conditions of hypoxic stress. The contribution of VEGF 3' untranslated mRNA structure and its interaction with DRBP76/NF90 represent another example of complex RNA elements that facilitate gene expression during cellular stress or in response to nutrient imbalance.


arrow
ACKNOWLEDGMENTS
 
This work was supported in part by National Institutes of Health-National Cancer Institute grant CA064436 (K.P.C.) and the Patrick and Catherine Weldon Donaghue Foundation (K.P.C.).

We appreciate the efforts of Nancy Ryan, Susan Kavel, and Diane Gran for their technical assistance.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Center for Vascular Biology, EM028, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501. Phone: (860) 679-8713. Fax: (860) 679-1201. E-mail: claffey{at}nso2.uchc.edu Back

{triangledown} Published ahead of print on 26 November 2007. Back

{dagger} Present address: The Feinstein Institute for Medical Research, Autoimmune Disease Center, 350 Community Drive, Manhasset, NY 11030. Back


arrow
REFERENCES
 
    1
  1. Agarwal, A., U. Munoz-Najar, U. Klueh, S. C. Shih, and K. P. Claffey. 2004. N-acetyl-cysteine promotes angiostatin production and vascular collapse in an orthotopic model of breast cancer. Am. J. Pathol. 164:1683-1696.[Abstract/Free Full Text]
    2. 2
  2. Bernstein, P., and J. Ross. 1989. Poly(A), poly(A) binding protein and the regulation of mRNA stability. Trends Biochem. Sci. 14:373-377.[CrossRef][Medline]
    3. 3
  3. Claffey, K. P., S.-C. Shih, A. Mullen, S. Dziennis, J. L. Cusick, K. R. Abrams, S. W. Lee, and M. Detmar. 1998. Identification of a Human VPF/VEGF 3' untranslated region mediating hypoxia-induced mRNA stability. Mol. Biol. Cell 9:469-481.[Abstract/Free Full Text]
    4. 4
  4. DeMaria, C. T., and G. Brewer. 1996. AUF1 binding affinity to A+U-rich elements correlates with rapid mRNA degradation. J. Biol. Chem. 271:12179-12184.[Abstract/Free Full Text]
    5. 5
  5. Dibbens, J. A., D. L. Miller, A. Damert, W. Risau, M. A. Vadas, and G. J. Goodall. 1999. Hypoxic regulation of vascular endothelial growth factor mRNA stability requires the cooperation of multiple RNA elements. Mol. Biol. Cell 10:907-919.[Abstract/Free Full Text]
    6. 6
  6. Fan, X. C., and J. A. Steitz. 1998. HNS, a nuclear-cytoplasmic shuttling sequence in HuR. Proc. Natl. Acad. Sci. USA 95:15293-15298.[Abstract/Free Full Text]
    7. 7
  7. Fan, X. C., and J. A. Steitz. 1998. Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO J. 17:3448-3460.[CrossRef][Medline]
    8. 8
  8. Forsythe, J. A., B. H. Jiang, N. V. Iyer, F. Agani, S. W. Leung, R. D. Koos, and G. L. Semenza. 1996. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol. Cell. Biol. 16:4604-4613.[Abstract/Free Full Text]
    9. 9
  9. Gallouzi, I. E., C. M. Brennan, M. G. Stenberg, M. S. Swanson, A. Eversole, N. Maizels, and J. A. Steitz. 2000. HuR binding to cytoplasmic mRNA is perturbed by heat shock. Proc. Natl. Acad. Sci. USA 97:3073-3078.[Abstract/Free Full Text]
    10. 10
  10. Goldberg-Cohen, I., H. Furneauxb, and A. P. Levy. 2002. A 40-bp RNA element that mediates stabilization of vascular endothelial growth factor mRNA by HuR. J. Biol. Chem. 277:13635-13640.[Abstract/Free Full Text]
    11. 11
  11. Gorgoni, B., and N. K. Gray. 2004. The roles of cytoplasmic poly(A)-binding proteins in regulating gene expression: a developmental perspective. Brief. Funct. Genomics Proteomics 3:125-141.[Abstract/Free Full Text]
    12. 12
  12. Guil, S., J. C. Long, and J. F. Cáceres. 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]
    13. 13
  13. Guttinger, S., P. Muhlhausser, R. Koller-Eichhorn, J. Brennecke, and U. Kutay. 2004. Transportin2 functions as importin and mediates nuclear import of HuR. Proc. Natl. Acad. Sci. USA 101:2918-2923.[Abstract/Free Full Text]
    14. 14
  14. Hirota, K., and G. L. Semenza. 2006. Regulation of angiogenesis by hypoxia-inducible factor 1. Crit. Rev. Oncol. Hematol. 59:15-26.[Medline]
    15. 15
  15. Huez, I., L. Creancier, S. Audigier, M. C. Gensac, A. C. Prats, and H. Prats. 1998. Two independent internal ribosome entry sites are involved in translation initiation of vascular endothelial growth factor mRNA. Mol. Cell. Biol. 18:6178-6190.[Abstract/Free Full Text]
    16. 16
  16. Hwang, S. I., D. H. Lundgren, V. Mayya, K. Rezaul, A. E. Cowan, J. K. Eng, and D. K. Han. 2006. Systematic characterization of nuclear proteome during apoptosis: a quantitative proteomic study by differential extraction and stable isotope labeling. Mol. Cell. Proteomics 5:1131-1145.[Abstract/Free Full Text]
    17. 17
  17. Ikeda, E., M. G. Achen, G. Breier, and W. Risau. 1995. Hypoxia-induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J. Biol. Chem. 270:19761-19766.[Abstract/Free Full Text]
    18. 18
  18. Kim, J. H., B. Hahm, Y. K. Kim, M. Choi, and S. K. Jang. 2000. Protein-protein interaction among hnRNPs shuttling between nucleus and cytoplasm. J. Mol. Biol. 298:395-405.[CrossRef][Medline]
    19. 19
  19. King, P. H. 2000. RNA-binding analyses of HuC and HuD with the VEGF and c-myc 3'-untranslated regions using a novel ELISA-based assay. Nucleic Acids Res. 28:E20.[CrossRef][Medline]
    20. 20
  20. Koritzinsky, M., M. G. Magagnin, T. van den Beucken, R. Seigneuric, K. Savelkouls, J. Dostie, S. Pyronnet, R. J. Kaufman, S. A. Weppler, J. W. Voncken, P. Lambin, C. Koumenis, N. Sonenberg, and B. G. Wouters. 2006. Gene expression during acute and prolonged hypoxia is regulated by distinct mechanisms of translational control. EMBO J. 25:1114-1125.[CrossRef][Medline]
    21. 21
  21. Koritzinsky, M., R. Seigneuric, M. G. Magagnin, T. van den Beucken, P. Lambin, and B. G. Wouters. 2005. The hypoxic proteome is influenced by gene-specific changes in mRNA translation. Radiother. Oncol. 76:177-186.[CrossRef][Medline]
    22. 22
  22. Levine, T. D., F. Gao, P. H. King, L. G. Andrews, and J. D. Keene. 1993. Hel-N1: an autoimmune RNA-binding protein with specificity for 3' uridylate-rich untranslated regions of growth factor mRNAs. Mol. Cell. Biol. 13:3494-3504.[Abstract/Free Full Text]
    23. 23
  23. Levy, A. P. 1998. Hypoxic regulation of VEGF mRNA stability by RNA-binding proteins. Trends Cardiovasc. Med. 8:246-250.[CrossRef][Medline]
    24. 24
  24. Levy, N. S., S. Chung, H. Furneaux, and A. P. Levy. 1998. Hypoxic stabilization of vascular endothelial growth factor mRNA by the RNA-binding protein HuR. J. Biol. Chem. 273:6417-6423.[Abstract/Free Full Text]
    25. 25
  25. Liu, Y., S. R. Cox, T. Morita, and S. Kourembanas. 1995. Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5' enhancer. Circ. Res. 77:638-643.[Abstract/Free Full Text]
    26. 26
  26. Ma, S., T. Musa, and J. Bag. 2006. Reduced stability of mitogen-activated protein kinase kinase-2 mRNA and phosphorylation of poly(A)-binding protein (PABP) in cells overexpressing PABP. J. Biol. Chem. 281:3145-3156.[Abstract/Free Full Text]
    27. 27
  27. Miller, D. L., J. A. Dibbens, A. Damert, W. Risau, M. A. Vadas, and G. J. Goodall. 1998. The vascular endothelial growth factor mRNA contains an internal ribosome entry site. FEBS Lett. 434:417-420.[CrossRef][Medline]
    28. 28
  28. Neurath, K. M., M. P. Keough, T. Mikkelsen, and K. P. Claffey. 2006. AMP-dependent protein kinase alpha 2 isoform promotes hypoxia-induced VEGF expression in human glioblastoma. Glia 53:733-743.[CrossRef][Medline]
    29. 29
  29. Parker, L. M., I. Fierro-Monti, T. W. Reichman, S. Gunnery, and M. B. Mathews. 2001. Double-stranded RNA-binding proteins and the control of protein synthesis and cell growth. Cold Spring Harb. Symp. Quant. Biol. 66:485-497.[CrossRef][Medline]
    30. 30
  30. Peng, S. S., C. Y. Chen, N. Xu, and A. B. Shyu. 1998. RNA stabilization by the AU-rich element binding protein, HuR, an ELAV protein. EMBO J. 17:3461-3470.[CrossRef][Medline]
    31. 31
  31. Rebane, A., A. Aab, and J. A. Steitz. 2004. Transportins 1 and 2 are redundant nuclear import factors for hnRNP A1 and HuR. RNA 10:590-599.[Abstract/Free Full Text]
    32. 32
  32. Ryan, H. E., M. Poloni, W. McNulty, D. Elson, M. Gassmann, J. M. Arbeit, and R. S. Johnson. 2000. Hypoxia-inducible factor-1{alpha} is a positive factor in solid tumor growth. Cancer Res. 60:4010-4015.[Abstract/Free Full Text]
    33. 33
  33. Sarkar, B., J. Y. Lu, and R. J. Schneider. 2003. Nuclear import and export functions in the different isoforms of the AUF1/heterogeneous nuclear ribonucleoprotein protein family. J. Biol. Chem. 278:20700-20707.[Abstract/Free Full Text]
    34. 34
  34. Saunders, L. R., and G. N. Barber. 2003. The dsRNA binding protein family: critical roles, diverse cellular functions. FASEB J. 17:961-983.[Abstract/Free Full Text]
    35. 35
  35. Semenza, G. L. 2000. HIF-1: mediator of physiological and pathophysiological responses to hypoxia. J. Appl. Physiol. 88:1474-1480.[Abstract/Free Full Text]
    36. 36
  36. Shih, S. C., and K. P. Claffey. 1999. Regulation of human vascular endothelial growth factor mRNA stability in hypoxia by heterogeneous nuclear ribonucleoprotein L. J. Biol. Chem. 274:1359-1365.[Abstract/Free Full Text]
    37. 37
  37. Shih, S. C., A. Mullen, K. Abrams, D. Mukhopadhyay, and K. P. Claffey. 1999. Role of protein kinase C isoforms in phorbol ester-induced vascular endothelial growth factor expression in human glioblastoma cells. J. Biol. Chem. 274:15407-15414.[Abstract/Free Full Text]
    38. 38
  38. Skobe, M., T. Hawighorst, D. G. Jackson, R. Prevo, L. Janes, P. Velasco, L. Riccardi, K. Alitalo, K. Claffey, and M. Detmar. 2001. Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat. Med. 7:192-198.[CrossRef][Medline]
    39. 39
  39. Stein, I., A. Itin, P. Einat, R. Skaliter, Z. Grossman, and E. Keshet. 1998. Translation of vascular endothelial growth factor mRNA by internal ribosome entry: implications for translation under hypoxia. Mol. Cell. Biol. 18:3112-3119.[Abstract/Free Full Text]
    40. 40
  40. Stoecklin, G., B. Gross, X. F. Ming, and C. Moroni. 2003. A novel mechanism of tumor suppression by destabilizing AU-rich growth factor mRNA. Oncogene 22:3554-3561.[CrossRef][Medline]
    41. 41
  41. Tsuzuki, Y., D. Fukumura, B. Oosthuyse, C. Koike, P. Carmeliet, and R. K. Jain. 2000. Vascular endothelial growth factor (VEGF) modulation by targeting hypoxia-inducible factor-1{alpha}-> hypoxia response element-> VEGF cascade differentially regulates vascular response and growth rate in tumors. Cancer Res. 60:6248-6252.[Abstract/Free Full Text]
    42. 42
  42. van den Beucken, T., M. Koritzinsky, and B. G. Wouters. 2006. Translational control of gene expression during hypoxia. Cancer Biol. Ther. 5:749-755.[Medline]
    43. 43
  43. van der Houven van Oordt, W., M. T. Diaz-Meco, J. Lozano, A. R. Krainer, J. Moscat, and J. F. Cáceres. 2000. The MKK(3/6)-p38-signaling cascade alters the subcellular distribution of hnRNP A1 and modulates alternative splicing regulation. J. Cell Biol. 149:307-316.[Abstract/Free Full Text]
    44. 44
  44. Zhang, T., V. Kruys, G. Huez, and C. Gueydan. 2002. AU-rich element-mediated translational control: complexity and multiple activities of trans-activating factors. Biochem. Soc. Trans. 30:952-958.[CrossRef][Medline]

Molecular and Cellular Biology, January 2008, p. 772-783, Vol. 28, No. 2
0270-7306/08/$08.00+0     doi:10.1128/MCB.02078-06
Copyright © 2008, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Kuwano, Y., Pullmann, R. Jr., Marasa, B. S., Abdelmohsen, K., Lee, E. K., Yang, X., Martindale, J. L., Zhan, M., Gorospe, M. (2010). NF90 selectively represses the translation of target mRNAs bearing an AU-rich signature motif. Nucleic Acids Res 38: 225-238 [Abstract] [Full Text]  
  • Avet-Loiseau, H., Li, C., Magrangeas, F., Gouraud, W., Charbonnel, C., Harousseau, J.-L., Attal, M., Marit, G., Mathiot, C., Facon, T., Moreau, P., Anderson, K. C., Campion, L., Munshi, N. C., Minvielle, S. (2009). Prognostic Significance of Copy-Number Alterations in Multiple Myeloma. JCO 27: 4585-4590 [Abstract] [Full Text]  
  • Hundley, H. A., Krauchuk, A. A., Bass, B. L. (2008). C. elegans and H. sapiens mRNAs with edited 3' UTRs are present on polysomes. RNA 14: 2050-2060 [Abstract] [Full Text]  

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 E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vumbaca, F.
Right arrow Articles by Claffey, K. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vumbaca, F.
Right arrow Articles by Claffey, K. P.

Copyright © 2010 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»