hnRNP A1 is a member of the hnRNPs (heterogeneous nuclear ribonucleoproteins) family of proteins that play a central role in regulating genes responsible for cell proliferation, DNA repair, apoptosis, and telomere biogenesis. Previous studies have shown that hnRNPA1 had reduced protein levels and increased cytoplasmic accumulation in senescent human diploid fibroblasts. The consequence of reduced protein expression and altered cellular localization may account for the alterations in gene expression observed during senescence. There is limited information for gene targets of hnRNP A1 as well as its in vivo function. In these studies, we performed RNA co-immunoprecipitation experiments using hnRNP A1 as the target protein to identify potential mRNA species in ribonucleoprotein (RNP) complexes. Using this approach, we identified the human double minute 2 (HDM2) mRNA as a binding target for hnRNP A1 in young and senescent human diploid fibroblasts cells. It was also observed that alterations of hnRNP A1 expression modulate HDM2 mRNA levels in young IMR-90 cells. We also demonstrated that the levels of HDM2 mRNA increased with the downregulation of hnRNP A1 and decrease with the overexpression of hnRNP A1. Although we did not observe a significant decrease in HDM2 protein level, a concomitant increase in p53 protein level was detected with the overexpression of hnRNP A1. Our studies also show that hnRNP A1 directly interacts with HDM2 mRNA at a region corresponding to its 3′ UTR (untranslated region of a gene). The results from this study demonstrate that hnRNP A1 has a novel role in participating in the regulation of HDM2 gene expression.
- hnRNP A1
- RNP complexes
Cellular senescence is best described as an inevitable irreversible proliferation arrest the phase of primary human fibroblasts in culture [1, 2]. The senescent phenotype is characterized by distinctive changes in morphology to become enlarged, flattened, and granular . Multiple factors that activate senescence include various types of stress-related stimuli such as aberrant oncogenic signaling, oxidative stress, and DNA damage . Moreover, the onset of senescence can be regulated by events such as epigenetic regulation, chromosome dynamics, protein degradation, mitochondrial mechanisms, and metabolic pathways. The molecular pathways of senescence differ considerably among cell types as well as different species .
Alternative splicing of pre-mRNAs is a process in which varied mRNA transcripts are generated to provide a major source of protein diversity in higher eukaryotes. Pre-mRNA splicing is a nuclear process that can be constitutive or alternative [4, 5]. Constitutive splicing involves the removal of introns and the joining of adjacent exons in the order of their arrangement. One of the core proteins involved in splicing is hnRNPA1 [5, 6]. Consequently, a single protein may be produced from a single pre-mRNA in constitutive splicing . In contrast, in alternative splicing, the variable use of splice sites permits two or more mature mRNAs to be generated from the same pre-mRNA. Among the nuclear complexes primarily responsible for alternative splicing are heterogeneous nuclear ribonucleoproteins, small nuclear ribonucleoproteins snRNPs, and SR proteins [8, 9]. Splicing factors that play a crucial role through concentration changes or alterations of their expression patterns have significant impacts on mRNA alternative splicing [9, 10].
We have previously found that hnRNP A1 is significantly downregulated in cellular senescence  and can regulate the levels of the alternatively spliced
We initiated this study to identify novel targets of hnRNP A1 and to further explore the role of hnRNP A1 in the modulation of gene expression during cellular senescence. The experimental approach used in this study was to identify the
2.1 hnRNP A1-messenger ribonucleoprotein (mRNP) complexes
We sought to identify putative mRNA substrates for hnRNP A1 by identifying mRNA sequences that directly bind to the hnRNP A1 protein. We employed a modification of a procedure that had been used for the characterization of RNP complexes by Mili et al. . We isolated hnRNP A1 protein complexes bound to their RNA targets from total young and senescent fibroblast cell lysates. To demonstrate that the complexes represented the majority of hnRNP A1-mRNP complexes found in cellular pools, we measured the mRNA levels of hnRNP A1 and actin in the isolated complexes by RT-PCR. Actin has been previously reported to be in hnRNP A1 RNP complexes; thus, we used actin as a positive control. The results in Figure 1A, show that actin and hnRNP A1 protein were present in lysates isolated from young and senescent IMR-90 fibroblasts. We assessed the protein level of hnRNP A1 in 4B10 RNP complexes isolated from young and senescent protein lysates. 4B10 RNP complexes reflect the results in panel A (Figure 1B) as hnRNP A1 was lower in senescent IMR-90 protein lysates following immunoprecipitation when compared with young cell lysates. These observations indicate that hnRNP A1 was present in the isolated RNP complexes. One known mRNA target of hnRNP A1 is its own RNA; therefore, we measured the level of hnRNP A1 mRNA by RT-PCR (Figure 1
2.2 Analysis of the hnRNPA1-mRNA complex
RNA that was isolated from the hnRNP A1-complexes was reverse transcribed and then amplified by PCR using random decamers to amplify all cDNA sequences. We used two different concentrations of cDNA template for PCR as it was not always possible to visualize PCR products in the senescent samples because the cDNA abundance is typically lower in these cells. There was a correlative increase in PCR products as shown in Figure 2 with an increase in cDNA template. PCR products were then immediately ligated into a pCR2.1cloning vector.
There may also be a differential availability of hnRNP A1 protein in old cells as hnRNPA1 is modified by post-translational events, such as phosphorylation and methylation [13, 14]. There is an additional possibility that rearrangements of individual components in hnRNPA1-mRNA ribonucleoparticles may change during senescence, which could alter specific and non-specific mRNA sequences bound in the complex.
2.3 Identification of RNA species in hnRNP A1-mRNP complexes
We then determined the identity of the cloned inserts by sequencing. We found that there were partial mRNA sequences for four human genes bound in young and senescent RNP complexes. The identity of genes was identical for both young and old cells. Scores were considered to be positive if the similarity score was more than 200 . The genes identified were Homo sapiens
The identification of the human double minute 2 gene (
The human murine double minute 2 (
To determine whether the full-length
2.4 Identification of p16INK4a mRNA in hnRNPA1 complexes
We have previously shown that changes in the expression of hnRNP A1 regulate the alternative splicing and mRNA levels of two mRNA isoforms of the INK4a locus known as p14Arf and p16(INK4a) [ 10]. Both protein isoforms are growth suppressors and knockout of the INK4a gene allows cells to escape cellular senescence . Our previous studies have shown  that overexpression of hnRNA1 results in a preferential expression of the p14Arf mRNA isoform, and an increase in the mRNA levels of both isoforms, thus suggesting a role for hnRNP A1 in control of cell proliferation and senescence . In this study, we assessed the ability of hnRNP A1 to directly bind to INK4a transcripts in hnRNPA1 complexes. For this, we used AR5 cells and PRNS-1 (SV40-transformed clones of HS74 primary human bone marrow fibroblasts) since these cells express high levels of INK4a transcripts as compared with normal IMR-90 fibroblasts .
Figure 5 shows that the p16 transcript was amplified from hnRNPA1RNP complexes indicating that hnRNP A1 directly binds to p16 mRNA. We also measured the ability of hnRNP A1 to bind to actin mRNA as a positive control for our co-immunoprecipitation studies. Actin mRNA has been previously identified in hnRNPA1-RNP complexes . Figure 5 shows that in addition to p16, we were also able to detect actin mRNA in the RNP complexes.
2.5 Expression of
HDM2is modulated by hnRNPA1 expression levels
We next sought to determine whether hnRNPA1 modulated the expression of
We also investigated whether the protein level of HDM2 was modulated by the level of hnRNP A1 protein expression. To determine whether endogenous hnRNP A1 has an effect on HDM2 protein expression, scrambled siRNA or siA1 was transfected into IMR-90 fibroblast cells. We found that upon siRNA knockdown of hnRNP A1, the protein level of HDM2 was not altered as shown in Figure 7A. hnRNP A2, which has overlapping biochemical activity with hnRNP A1, when inhibited by siRNA interference, did not affect HDM2 expression. On the other hand, overexpression of hnRNP A1 in young cells transfected with GFP-A1 resulted in a slight decrease of HDM2 protein levels and an increase in p53 levels when compared with cells transfected with the GFP-Empty vector (Figure 7B). The increase in p53 protein levels may be a result of the decreased HDM2 expression. A direct correlation between protein and mRNA levels for any given gene is complicated by varying processes. For instance, studies conducted by various groups such as Vogel et al.  pointed out that transcription, mRNA export, decay, translation, and protein degradation are key processes in determining steady-state protein concentration .
2.6 Identification of
HDM2RNA sequences that bind to hnRNP A1
Our PESX analysis revealed that hnRNP A1 has a putative binding site within the intronic region between intron 9 and exon 10 of HDM2 (Figure 4). We obtained HDM2 constructs from Dr. Meek (University of Dundee). We performed biotin pull-down assay using MP4 construct that is similar to the MDM2-B isoform lacking p53-binding region followed by Western immunoblotting. We performed the biotin pull-down assay by first incubating the hnRNP A1 antibody (4B10) with Dynabeads Myone streptavidin. We also incubated the Biotinlabeled
The role of hnRNP A1 during cellular senescence is unclear. Significant alterations in its levels, localization, and activity in senescent cells suggest that hnRNP A1 may contribute to the senescent phenotype . However, only a few gene targets are known for hnRNP A1 . This prompted us to search for additional mRNA targets for hnRNP A1 in young and senescent IMR-90 cells. We used an RNA co-immunoprecipitation protocol  to identify mRNA new targets for the hnRNP A1 protein. We found that hnRNP A1 is bound to several mRNAs not previously identified. Of particular interest to us was the observation that hnRNP A1 bound to
Posttranscriptional regulation of gene expression is important for the control of cellular processes such as cell proliferation, differentiation, development, and apoptosis . RNA-binding proteins are the main regulators of post-transcriptional regulation . hnRNP A1 is a multifunctional RNA-binding protein implicated in the regulation of major steps in posttranscriptional regulation of gene expression . Upon observation that hnRNP A1 binds to
In this study, our findings suggest that hnRNP A1 binds to the Biotinlabeled
4. Materials and methods
4.1 Cell culture and generation of senescent fibroblasts
The human lung fibroblast cell strain IMR90 from Coriell, NJ, was subcultured from early passage to terminal passage as previously described by Hubbard and Ozer  in Dulbecco’s modified Eagle’s medium and Ham’s F10 medium in a 1:1 mixture supplemented with 10% fetal bovine serum. IMR90 fibroblasts at population doubling <35 were used in all experiments and are considered comparable to young fibroblasts as determined by gene expression profiles previously performed . Senescent IMR90 in all experiments was at a population doubling of 62. For transfection experiments, once cells had reached 90% confluence, either the expression plasmid pEFGP (Control) or pEFGP-A1 was transfected into IMR-90 cells in DMEM/F10 media without FBS/penicillin using Lipofectamine 2000 (Invitrogen) and incubate at 37°C in a CO2 incubator for 6 h.
4.2 RNA isolation and RNA-PCR to check for genomic contamination
After RNA was isolated as detailed above, to ensure that there was no genomic contamination, an RNA–PCR procedure was performed. 2 μL of template RNA was added to a PCR mixture using β-actin primers, which would detect genomic sequences if present. The total PCR reaction was composed of 2 μl of template RNA, 5 μl of 10× RT-PCR Buffer (100 mM Tris–HCl, pH 8.3, 500 mM KCl, 15 mM MgCl2), 2.5 μl of dNTP mix (2.5 mM each dNTP), 0.25 μM each PCR primer, 0.5 unit of Thermostable DNA Polymerase (Novagen).
The primers for actin were: actin forward (5′CGCCGCCCTAGGCACCA3′) and actin reverse (5′TTGGCCTTAGGGTTCAGGGGGG 3′). For hnRNPA1, primer set were: hnRNP A1 forward (5′CTAAAGAGCCCGAACAGCTGAG 3′) and hnRNP A1 reverse (5′TCAGTGTCTTCTTTAATGCCACCA 3′). SYBR™ green stain. SYBR™ green can be visualized by blue fluorescence (Molecular Dynamics, Amersham) and quantified with ImageQuant software (Amersham).
5. Immunoblotting and protein analysis
Standard Western blotting protocols (Harlow et al. 1999) were to analyze specific proteins . Protein extracts isolated from young and senescent fibroblasts were generated by washing cells were washed 3 times with 1X cold PBS and then, cultures were placed on ice. Cold RIPA (radioimmunoassay buffer containing NP-40 at 1%, sodium deoxycholate at 1%, sodium dodecyl sulfate at 0.1%, NaCl at 150 mM and Tris-HCl at 10 mM with protease inhibitors leupeptin at 0.1 μg/ml, pepstatin at 0.1 μg/ml, and phenylmethylsulfonyl fluoride at 1 mM) was added to culture dishes followed by scraping cells into cold microfuge tubes. The lysate was passed through a 21-gauge syringe needle to ensure complete lysis. Lysates were centrifuged at 10,000×g for 10 minutes at 4°C. The cleared lysate was collected and aliquots were prepared to estimate the amount of protein by the Bradford protein assay (Bio-Rad). Lysates were run on 8–12% acrylamide gels and then transferred in an electroblotting apparatus. Membranes (PVDF, Osmonics) were blocked with 5% non-fat milk in PBS. Monoclonal antibody 4B10 (1:10,000) was used to detect hnRNPA1/A1. An anti-actin monoclonal antibody (1:5000) was obtained from Chemicon. Antibodies specific for HDM2 were graciously provided by Dr. Jill Bargonetti. Secondary antibody, goat IgG, or mouse IgG conjugated with HRP were used for visualization of bands using the ECL kit (Amersham).
6. Overexpression of hnRNP A1 by transient transfection
Young IMR-90 cells were cultured in 10-cm plates. After approximately 24 h of incubation when the cells reached 90% confluence, the expression plasmid pEFGP (control) or pEFGP-A1 was transfected into IMR-90 cells in DMEM/F10 media without FBS/penicillin using Lipofectamine 2000 (Invitrogen) and incubated at 37°C in CO2 incubator for 6 h. We changed the media to DMEM with FBS and without penicillin and incubated for 48 h at 37°C.
7. RNA co-immunoprecipitation protocol
The RNA co-immunoprecipitation protocol was a modified version published by Mili et al.  that included a short immunoprecipitation step that minimized degradation of protein-associated RNA.
8. Confirmation of gene expression using real-time PCR
Real-time PCR experiments on selected genes were performed using an Applied Biosystems 7500 real-time PCR system that utilizes TaqMan gene expression assays for the following genes: mdm2; Human GAPD (GAPDH) Endogenous Control FAM/MGB (4333764F). Reactions were performed according to standard methods using the universal 10X PCR TaqMan mix, at a final reaction volume of 25 μL (Applied Biosystems).
9. Cloning protocol and sequencing
PCR products were ligated into the pCR 2.1 (Invitrogen, TA cloning kit) cloning vector that utilizes the single dT overhangs that are a by-product of PCR reactions catalyzed by Taq polymerase. The ligation reaction was performed at 14°C overnight using T4 DNA ligase and 3 μL of fresh PCR product (Invitrogen protocols).
10. Sequence analysis using BLAST and PESX
Vector sequences were subtracted from the sequences obtained. The rest of the sequence was compared against known sequences using the BLAST tool (www.ncbi.nlm.nih.gov). Sequences were chosen based on being previously identified as human genes and either in the coding regions or in the flanking regions of the mRNA sequences of known genes.
11. Biotin pull-down assay
Biotinylated transcripts were obtained by reverse transcription with the Maxiscript Kit (Invitrogen) according to manufacturer instructions and as previously described above in Section 3.2. The biotin pull-down assay was performed by first incubating hnRNP A1 antibody (4B10) with Dynabeads Myone streptavidin (Invitrogen) for 1 hour at 4 C. Also, incubated Biotinlabeled
We are grateful to Alisa Chalmers for technical assistance. This research was supported by NIH/NCI U54CA137788/U54CA132378 to KH and to NIH/NIMHD 3G12MD007603-30S2 and NIH RISE R25GM056833 to HM. We also thank Dr. Serafin Pinol-Roma for hnRNP specific antibodies and consultation on the RNA co-immunoprecipitation assays. We also thank Dr. Bargonetti for the HDM2 antibody and Dr. Meek for providing HDM2 plasmid constructs. Heriberto Moran and Shanaz A. Ghandhi contributed equally to this work.
Conflict of interest and financial disclosures
There are no conflicts of interest nor financial interest or benefits.
|HDM2||Human Double Minute 2|
|hnRNPA1||Heterogeneous nuclear ribonucleoprotein A1|
|MP4||Biotin-labeled HDM2 RNA probe|